LLM Backgammon Showdown 2025 Match 5: DeepSeek V3.2 Speciale (Abridged By Claude)

I find it poetic that the final LLM we’ll be exploring will be DeepSeek, a model that completely upended the LLM industry through its reasoning mechanics. I remembered the headlines spinning about how this was the end of AI because some small upstart in China was out preforming AAAA-budget AI companies. At least before the other LLMs came out with their own Reasoning Models.

But, can the OG Reasoning LLM, with nearly a full year of updates, be able to stack up? Same rules apply:

  • Removed one piece from each column.
  • DeepSeek rolls doubles each turn.
  • If they make a paradoxical move, such as pulling Checkers from an empty space or moving Checkers that are not their color, they are disqualified. I had to modify this from “erroneous” since Llama had made very legal plays just that it needed to move two more checkers.

So, let’s begin:

“No endpoints found that support image input.”

Hmm… Even with MistralOCR?

“No endpoints found that support image input.”

Interesting… Interesting indeed…

I can’t exactly disqualify DeepSeek just yet, especially after confirming that none of the models have image input. So, I am forced to do something… unthinkable.

This is the prompt I will use for DeepSeek instead:

Let’s play some Backgammon. I’ll be black, you’ll be white. I’ll give you the first turn and you’ve rolled two sixes. What do you want to do?

I have set up two different formats to establish our current board state. Going forward in our game, which do you prefer:

GNU Position: 4Bk8ABIAD+YBAA

SimpleBG: turn:white;pointsW:0,0,0,0,0,4,0,2,0,0,0,0,4,0,0,0,0,0,0,0,0,0,0,1;pointsB:1,0,0,0,0,0,0,0,0,0,0,4,0,0,0,0,2,0,4,0,0,0,0,0;barW:0;barB:0;offW:0;offB:0

And now, for its response, which was extremely large. I mean, it took 17 minutes to fully process the whole shebang, so instead, I will have Claude summarize the thought process for each turn.

The text describes an AI assistant’s analysis of a Backgammon game scenario where:

  • Setup: The assistant plays white, the user plays black. White rolled double sixes (6,6) on the opening turn.
  • Format Choice: The user offered two board notation formats—GNU Position (a base64-encoded string) and SimpleBG (a human-readable format listing checker positions). The assistant chose SimpleBG for its clarity.
  • Board Analysis: The SimpleBG data revealed each player has 11 checkers (not the standard 15), with symmetric placement where corresponding points sum to 25. White’s checkers are on points 6, 8, 13, and 24; black’s on points 1, 12, 17, and 19. Movement direction was determined: white moves from high to low point numbers.
  • Move Selection: After evaluating multiple options, the assistant chose 13/7(2) 7/1(2)—moving two checkers from point 13 to point 1 via point 7. This:
    • Hits black’s blot on point 1 (sending it to the bar)
    • Makes white’s 1-point with two checkers
    • Leaves only one blot (on point 24)
  • Result: The move creates a strong position with made points on 1, 6, 8, and 13, while putting black at a disadvantage with a checker on the bar.

… DeepSeek just won this entire game.

Like, I am not making this up.

I need to break down the statistics for this specific output alone.

One thing I hadn’t addressed throughout the entire experiment was input tokens, since I didn’t feel like that would be a major factor in outlining them. However, I feel like it’s needed to show just how impressive DeepSeek is compared to these other LLMs.

  • ChatGPT: Needed an average of 2,400 Input Tokens on top of its 3,500 for a total of close to 6,000 tokens on average.
  • Claude: 1,585 tokens were inputted to 4,505 output for a total of 6,090 tokens.
  • Gemini: 1,382 tokens were inputted to 1,253 output for a total of 2,635 tokens.
  • Grok: Needed an average of 976 Input Tokens on top of its 4,800 for a total of close to 5,800 tokens on average.
  • Gemini: Needed an average of 3,900 input tokens to output an average of 404 tokens for a total of 4,300 tokens average.

This leaves us with DeepSeek. For as little as a penny, it turned my 362 input tokens into an eye-watering, mind-melting 26,082 output tokens. This is higher than anything the LLMs put out. The only downside is that it has a fairly slow output time. While it has a quick first token of 0.57 seconds, it moves at 25.1 Tokens a second. This means that I was waiting 17 minutes 19 seconds for DeepSeek to respond.

Its thought process is all over the place too. It began to loop around in its thought processes and I worried I may had sent it into a death spiral on multiple occasions. However, once it came close to its conclusion, I felt a sigh of cathartic relief come over me as I saw one of the most competent plays I’ve ever seen in this game.

I rolled a 6-3, moving my sole piece off the bar and onto the field while making blots out of my two pointer. Now it’s DeepSeek’s response:

The text describes an AI’s analysis of a backgammon position where white must play a roll of double fives (5,5).

Key Points:

  • Board State Parsing: The AI works through the SimpleBG format, identifying white’s checkers on points 1, 6, 8, 13, and 24, and black’s on points 12, 14, 17, and 19. Each player has 11 checkers (non-standard).
  • Direction Clarification: Significant effort is spent determining movement direction. The AI concludes white moves from high to low point numbers (toward points 1-6 as the home board).
  • Move Analysis: Several options are evaluated:
    • 13/8, 13/8 — moves both checkers from 13 to 8
    • 8/3, 8/3 — moves both from 8 to 3
    • 13/3 (combined) — one checker moves 10 pips
    • Various combinations involving points 6 and 1
  • Vulnerability Assessment: The AI notes that leaving a blot on point 13 is dangerous since black can hit it from point 12 with a roll of 1. The existing blot on point 24 is vulnerable to black rolling a 5 from point 19.
  • Final Decision: The AI chooses 13/8(2) because it:
    • Clears the vulnerable point 13 entirely
    • Builds a strong 4-checker stack on point 8
    • Reduces pip count by 10 (same as alternatives)
    • Advances the farthest non-back checkers
  • Output: The updated SimpleBG state is provided with the turn switching to black.

Uh… That last bit doesn’t sound good…

So I noticed too late that I forgot to switch the turn to White in the program I used to export the notations, so I think it may have messed with its mind a bit. Even looking at the state of the board, it only moved two pieces.

I couldn’t edit the thing to correct myself, so I decided to make a new game where I limited its reasoning to be minimal. Same prompt and everything, but the core difference was that I would only use the BG Simple notation.

Spoilers, it did not reduce its thought process one bit, hence why Claude is coming in clutch.

Game Situation

  • Turn: White (playing as the AI)
  • Dice Roll: Double sixes (four moves of 6 pips each)
  • White’s Checkers: Points 6 (4), 8 (2), 13 (4), 24 (1)
  • Black’s Checkers: Points 1 (1), 12 (4), 17 (2), 19 (4)

Analysis Overview

The text evaluates multiple move plans (A through F), weighing factors such as:

  • Blot vulnerability (single checkers that can be hit)
  • Prime building (creating consecutive blocked points)
  • Pip count reduction (advancing checkers toward home)
  • Hit probability calculations for blots on points 24 vs. 18

Chosen Move: Option F

Move two checkers from point 13 → 7, and two checkers from point 8 → 2

Reasoning

AdvantageExplanation
Minimizes blotsOnly one blot remains (on point 24)
Makes the bar pointPoint 7 becomes a 2-checker block
Makes the 2-pointCreates a home board point for bearing off
Balanced distributionNo overstacking; checkers well-positioned

Resulting Position

turn:black;pointsW:0,2,0,0,0,4,2,0,0,0,0,0,2,0,0,0,0,0,0,0,0,0,0,1;pointsB:1,0,0,0,0,0,0,0,0,0,0,4,0,0,0,0,2,0,4,0,0,0,0,0;barW:0;barB:0;offW:0;offB:0

White now controls points 2, 6, 7, and 13 with safe stacks, leaving Black’s back checker on point 1 difficult to escape past the blocked bar point.

There’s a lot of thought processes going on here…

But I can’t worry about that right now, since I’m here to see how long it can last before it collapses in on itself. So, my roll: 4-4. With this roll, I move my sole checker out of harms way and move two black pieces up. DeepSeek’s turn.

Game Situation

  • Turn: White
  • Dice Roll: Double fives (four moves of 5 pips each)
  • White’s Checkers: Points 2 (2), 6 (4), 7 (2), 13 (2), 24 (1)
  • Black’s Checkers: Points 9 (1), 12 (2), 16 (2), 17 (2), 19 (4)

Key Observations

FactorDetails
Blot on 24Cannot be moved—destination point 19 is blocked by 4 black checkers
Race conditionBlack’s farthest-back checker is on point 9, so all white points ≤8 are safe from hits
Bearing offNot allowed—checkers remain outside home board (points 1-6)

Legal Move Options Analyzed

  1. 13 → 8 (single 5): Legal, point 8 is empty
  2. 13 → 8 → 3 (two 5s on same checker): Legal, brings checker into home board
  3. 7 → 2 (single 5): Legal, adds to existing stack
  4. 6 → 1 (single 5): Legal, creates new home point
  5. 2 → off: Illegal—bearing off not permitted yet

Move Decision

Optimal play: Move both checkers from point 13 to point 3, each using two fives (13→8→3).

Reasoning

  • Advances the most distant movable checkers (10 pips each)
  • Creates a made point on 3 (two checkers = safe)
  • Reduces checkers outside home board from 5 to 3
  • Minimizes total pip count of outside checkers (38 vs. alternatives at 39-47)
  • All options yield same total pip count (72), so position quality determines the choice

Resulting Position

turn:black;pointsW:0,2,2,0,0,4,2,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1;pointsB:0,0,0,0,0,0,0,0,1,0,0,2,0,0,0,2,2,0,4,0,0,0,0,0;barW:0;barB:0;offW:0;offB:0

White’s new distribution: Points 2 (2), 3 (2), 6 (4), 7 (2), 24 (1) — with a strong home board foundation and only 3 checkers remaining outside.

(The following is from the actual DeepSeek’s answer)

[{‘type’: ‘text’, ‘text’: “I’ll use the double fives to move both checkers from point 13 to point 3. Each checker moves two steps of 5: 13 → 8 → 3.\n\nThe new board state in SimpleBG Notation is:\n\n\nturn:black;pointsW:0,2,2,0,0,4,2,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1;pointsB:0,0,0,0,0,0,0,0,1,0,0,2,0,0,0,2,2,0,4,0,0,0,0,0;barW:0;barB:0;offW:0;offB:0\n”}]

Oh, oh no, it is breaking down now. But that’s a lot more sensible since it took two checkers ten spaces. Now it’s my go. I rolled a 6-2 and… I realized how evil I can be…

I’m gonna throw a curveball at DeepSeek’s way. I have never attacked a Blot before and now I want to see how the AI to needing to take a piece from the bar.

Game Situation

  • Turn: White
  • Dice Roll: Double fours (4-4)
  • Critical Factor: White has 1 checker on the bar (barW:1)
  • Rule: Must re-enter from bar before moving any other checkers

Board State

White CheckersBlack Checkers
Point 2: 2Point 9: 1 (blot)
Point 3: 2Point 12: 2
Point 6: 4Point 16: 1 (blot)
Point 7: 2Point 17: 2
Bar: 1Point 19: 4
Point 24: 1 (blot)

Entry Point Calculation

For white entering from bar with a 4:

  • Die 1 → Point 24, Die 2 → Point 23, Die 3 → Point 22, Die 4 → Point 21, Die 5 → Point 20, Die 6 → Point 19
  • Point 21 has 0 black checkers → Entry is legal

Move Options Analyzed

OptionMovesResult
ABar→21, then 7→3Blot on 21, blot on 7, 9 checkers in home board
BBar→21, then 6→2Blot on 21 only, 8 checkers in home board

Key Insight

Point 7 blot is safe from hits because:

  • Black’s lowest checker is on point 9
  • Black moves increasing (toward point 24)
  • Black cannot move backward to reach point 7
  • Therefore, white blots on points ≤8 are immune to direct hits

Decision: Option A

  • Both options have equal pip count (65 pips)
  • Option A brings 9 checkers into home board vs. Option B’s 8
  • The “vulnerable” blot on point 7 is actually safe
  • Advances the race by reducing outside checkers from 3 to 2

Resulting Position

turn:black;pointsW:0,2,3,0,0,4,1,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0;pointsB:0,0,0,0,0,0,0,0,1,0,0,2,0,0,0,1,2,0,4,0,0,0,0,1;barW:0;barB:0;offW:0;offB:0

White’s new distribution: Points 2 (2), 3 (3), 6 (4), 7 (1), 21 (1) — with only the point 21 blot being genuinely vulnerable to black’s next roll.

So… it basically has one more 4 left to play… Honestly, I want to keep this going and just see if it can go the distance. The amount of tokens it needs to complete is getting less and less. I rolled two sixes.

With my first six, I move a black piece from 9 -> 15 – > 21, taking out DeepSeek’s blot. I then move two pieces from 12 -> 18. It’s turn now, can it beat Llama’s record?

Game Situation

  • Turn: White
  • Dice Roll: Double threes (3-3)
  • Critical Factor: White has 1 checker on the bar (barW:1)
  • Rule: Must re-enter from bar before moving any other checkers

Board State

White CheckersBlack Checkers
Point 2: 3Point 16: 1
Point 3: 4Point 17: 2
Point 6: 3Point 18: 2
Bar: 1Point 19: 4
Point 21: 1
Point 24: 1

Entry Point Calculation

For white entering from bar with a 3:

  • Die 1 → Point 24, Die 2 → Point 23, Die 3 → Point 22, Die 4 → Point 21, Die 5 → Point 20, Die 6 → Point 19
  • Point 22 has 0 black checkers → Entry is legal

Move Analysis

Move OptionLegalityReason
Bar → Point 22✓ LegalPoint 22 is empty
Point 22 → Point 19✗ BlockedPoint 19 has 4 black checkers
Point 6 → Point 3✓ LegalPoint 3 has friendly checkers
Point 2 → off✗ IllegalNot all checkers in home board
Point 3 → off✗ IllegalNot all checkers in home board

Decision: Forced Move

With the second 3, only one legal option exists:

  • Move from point 6 to point 3 (6−3=3)

The entered checker cannot advance (point 19 blocked), and bearing off is impossible since a checker now sits on point 22 (outside home board).

Resulting Position

turn:black;pointsW:0,3,5,0,0,2,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0;pointsB:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,2,2,4,0,1,0,0,1;barW:0;barB:0;offW:0;offB:0

White’s new distribution: Points 2 (3), 3 (5), 6 (2), 22 (1) — with the newly entered checker on point 22 needing to traverse the entire board to reach home.

Honestly? I think DeepSeek can not only go the distance, but can also end up rebalancing the entire way I do this experiment.

I make my move of 5-2, knocking their piece off yet again. They’re back in the bar now with a 2-2. What is their play?

Game Situation

  • Turn: White
  • Dice Roll: Double twos (2-2)
  • Critical Factor: White has 1 checker on the bar (barW:1)
  • Rule: Must re-enter from bar before moving any other checkers

Board State

White CheckersBlack Checkers
Point 2: 3Point 16: 1
Point 3: 7Point 17: 1
Bar: 1Point 18: 2
Point 19: 4
Point 21: 1
Point 24: 2

Key Observation: All white checkers (except the bar checker) are already in home board (points 1-6)

Entry Point Calculation

For white entering from bar with a 2:

  • Die value directly corresponds to point number for entry
  • Die 2 → Point 2
  • Point 2 has 0 black checkers → Entry is legal

Move Analysis

StepActionResult
1st die (2)Enter bar → Point 2All checkers now in home board
2nd die (2)Bear off from Point 2Legal since all checkers in home board

Why bearing off is possible: After entering, all 11 white checkers are in the home board (points 1-6), satisfying the bearing-off requirement.

Net effect on Point 2: +1 (entry) −1 (bear off) =0 change

Decision: Enter and Bear Off

  • Enter checker from bar onto point 2
  • Bear off one checker from point 2 (exact match with die value)
  • This is optimal as it begins removing checkers from the board

Resulting Position

turn:black;pointsW:0,3,7,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0;pointsB:0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,2,4,0,1,0,0,2;barW:0;barB:0;offW:1;offB:0

White’s status: Points 2 (3), 3 (7), with 1 checker borne off (offW:1) — white is now in the bearing-off phase with all remaining checkers in home board.

Honestly, I don’t know whether to pull the plug, since now it assumes it’s starting off on point 2 and bearing off. I also kinda placed them in a stalemate since if I just took their controls and moved the piece the full eight spaces, it would be blocked by the wall I inadvertently made.

So, I commend you, DeepSeek, you have gone the distance. 1st place goes to you in a way that practically upsets the entire ladder since now I need to give everyone else this fair shake of doing the notation system.

But before that happens, let’s break down the stats. I already mentioned how much it did on its first turn, but in the total amount of turns it took, here’s my findings:

The cost remained the same, being a penny. Speeds were also consistent with DeepSeek’s first run. What was different was that each turn, it took less and less tokens to reply. My assumption is that once it realized the game it was playing and the rules it was assuming, it cut down the unneeded parts of its reasoning and went for what works.

However, there’s another idea I have to entertain: It had the lowest context length out of all the LLMs with 164 thousand tokens. Adding every token up totaled up to 81 thousand, about half of its context window. Despite this, I squeezed blood from the stone and it managed to play its hardest.

Another downside was its length. I had to cut this with a “read more” not for more clicks but because DeepSeek talks a lot. Hell, I need to have something summarize it down for better analysis, because I doubt a, as of this writing, 74-thousand-word document would be legible by most LLMs. With Claude’s help, I was able to truncate it down to a mere 3000 words.

This also bleeds into the length it takes to get a response. I thought ChatGPT took forever, but DeepSeek took the cake. Factoring both the first round and the attempt to tone down its reasoning, it still had the highest average output token count of 17,806. Plugging in average speeds, you’re looking at 12 minutes of waiting for a response. But on your first round? Expect to wait at least 20.

Despite this, DeepSeek gets 1st place… But not in the way you expect it. For you see, now the goal is for every LLM I’ve reviewed thus far, plus some fan favorites, to try and dethrone DeepSeek through simple notations. Starting from the top, we’ll do ChatGPT and work our way to the bottom.

LLM Backgammon Showdown 2025 Match 5: DeepSeek V3.2 Speciale

I find it poetic that the final LLM we’ll be exploring will be DeepSeek, a model that completely upended the LLM industry through its reasoning mechanics. I remembered the headlines spinning about how this was the end of AI because some small upstart in China was out preforming AAAA-budget AI companies. At least before the other LLMs came out with their own Reasoning Models.

But, can the OG Reasoning LLM, with nearly a full year of updates, be able to stack up? Same rules apply:

  • Removed one piece from each column.
  • DeepSeek rolls doubles each turn.
  • If they make a paradoxical move, such as pulling Checkers from an empty space or moving Checkers that are not their color, they are disqualified. I had to modify this from “erroneous” since Llama had made very legal plays just that it needed to move two more checkers.

So, let’s begin:

“No endpoints found that support image input.”

Hmm… Even with MistralOCR?

“No endpoints found that support image input.”

Interesting… Interesting indeed…

I can’t exactly disqualify DeepSeek just yet, especially after confirming that none of the models have image input. So, I am forced to do something… unthinkable.

This is the prompt I will use for DeepSeek instead:

Let’s play some Backgammon. I’ll be black, you’ll be white. I’ll give you the first turn and you’ve rolled two sixes. What do you want to do?

I have set up two different formats to establish our current board state. Going forward in our game, which do you prefer:

GNU Position: 4Bk8ABIAD+YBAA

SimpleBG: turn:white;pointsW:0,0,0,0,0,4,0,2,0,0,0,0,4,0,0,0,0,0,0,0,0,0,0,1;pointsB:1,0,0,0,0,0,0,0,0,0,0,4,0,0,0,0,2,0,4,0,0,0,0,0;barW:0;barB:0;offW:0;offB:0

And now, for its response…

Continue reading

LLM Backgammon Showdown 2025 Match 5: Llama 4 Maverick

So, going down Openrouter’s list of LLM Series, the next one up was something called “Cohere”. However, looking over what Cohere does, it’s more of a niche search engine meant for Cloud services rather than the typical LLM chatbots that ChatGPT and Grok represent.

So, we’re moving onto Llama 4 Maverick. I’m gonna present the same scenario I always do: one piece missing from each column, doubles for six rolls, bad move = disqualification, reasoning enabled.

Except… We don’t have reasoning to enable… Oh boy… I am concerned for what sort of plays this will make without some space for it to think.

To make a move with two sixes, I’ll consider the possible actions based on the current state of the board.

First, let’s identify the checkers I can move. I have checkers on points 1, 12, 17, and 19. Moving a checker from point 1 (bearing off is not possible yet, so it’s not a valid starting point for a 6), or from points 12, 17, or 19 are the options.

  1. Moving from point 12: I can move a checker from point 12 to 18 (12 + 6 = 18). This is a valid move as there are no black checkers on point 18.
  2. Moving from point 17: I can move a checker from 17 to 23 (17 + 6 = 23). This is a valid move since point 23 is open.
  3. Using the second 6:
    • If I moved a checker from 12 to 18, I can then move another checker from 17 to 23 (17 + 6 = 23), or from 19 to 25 (but 25 is off the board, so that’s not valid). Moving from 12 to 18 and then from 19 to 1 (19 – 6 = 13, but we need to count towards the opponent’s home, so it’s actually 19 to 13, and then we can’t do another 6 because we can’t directly count 6 from 13) isn’t a valid sequence for the second 6 because we can’t directly apply it. However, we can apply it as moving another checker 6 spaces.

Given the roll of two sixes, a good strategy is to move two checkers 6 spaces forward.

Let’s move a checker from point 17 to 23 and another checker from point 19 to 1 is not possible with a 6 directly (it would be 19-6=13), but we can do it in two steps or directly if the number allows. However, I’ll directly apply the 6.

I’ll move a checker from 17 to 23 and then another from 12 to 18.

The moves are:

  • One checker from 17 to 23.
  • Another checker from 12 to 18.

These moves utilize both sixes effectively, advancing my checkers and maintaining a strong position.

… Ok, I’ll admit, I didn’t see this one coming. I expected the AI to assume the board hasn’t changed from what has been ingrained into its memory, but I didn’t expect it to not realize it can move four pieces with a double. So, I won’t disqualify it but rather five it a cookie.

Firing up that ol’ Nintendo Clubhouse game to get a good, solid ruling of the doubles mechanic and accidentally leading by example with a 6-6 of my own, I gently told Llama what my turn entailed, perhaps the most instructions I gave out of all the LLMs I’ve done so far.

To quote:

“If you roll double, you get to play each die twice.”

Case in point, on my move, I also rolled two sixes, and I moved two pieces from 13-7 and two pieces from 7-2. I took the liberty of moving two additional checkers from your 17 and 12 to 23 and 18 respectively. It is now your move and you have rolled 2 5s.

And so, we get its response:

To make a move with two 5s, I need to identify the checkers I can move and the valid moves according to the current state of the board.

First, let’s identify the checkers I can move. I have checkers on points 1, 12, 18, 19, and 23.

  1. Moving from point 18: I can move a checker from 18 to 23 (18 + 5 = 23). This is a valid move. There are already red checkers on point 23, so I can safely place another one there.
  2. Moving from point 19: I can move a checker from 19 to 24 (19 + 5 = 24). This is a valid move as there are black checkers on point 24.

For the second 5:

  • If I moved a checker from 18 to 23, I can then look for another checker to move 5 spaces.
  • I can move a checker from 12 to 17 (12 + 5 = 17). This is a valid move because there are red checkers on point 17 (though they were previously there, now they are still there after your turn), so it’s not ideal to go there if I had a checker to move from another point.

However, a better analysis shows that moving from 19 to 24 is a good start. Then, for the second 5, moving from 18 to 23.

The moves are:

  • One checker from 19 to 24.
  • Another checker from 18 to 23.

These moves utilize both 5s effectively, advancing my checkers and maintaining a strong position.

Still didn’t quite get the memo, but their decision still allowed for four checkers to be played regardless. I’m going to gently remind them again of the rule, then play. Remember, they have to make an impossible move and so far, these are very possible.

“A player who rolls doubles plays the numbers shown on the dice twice. A roll of 6 and 6 means that the player has four sixes to use, and he may move any combination of checkers he feels appropriate to complete this requirement.”

It’s my turn, I rolled a 6 and a 4, and so I moved the barred piece from its place down to spot 15. It is your move with two 4s.

It’s response:

Let’s analyze the current position of the board and determine the best move with the roll of two 4s.

First, let’s identify the checkers I can move. I have checkers on points 1, 6, 12, 19, 23, and 24. Since I rolled two 4s, I will look for valid moves that involve moving checkers 4 spaces.

  1. Moving from point 19: I can move a checker from 19 to 23 (19 + 4 = 23). This is a valid move as there are already red checkers on point 23.
  2. Using the second 4: I can move another checker 4 spaces.
    • Moving from 6 to 10 is a valid move, as it is an open point.

The moves are:

  • One checker from 19 to 23.
  • Another checker from 6 to 10.

These moves utilize both 4s effectively, advancing my checkers and maintaining a strong position.

I’m really impressed by how far it was going despite essentially gimping itself by not moving twice, but it’s made an impossible move by taking from 6, a column where it has black pieces. My pieces. Still, it was a good enough contender to claim third place without even going into the stats.

It’s cost and speed are comparable to Grok, barely half a penny in cost and lightning fast. Its first turn was posted at 729.9 tokens a second… and it only produced 500 tokens! However, that is the problem it has faced: the lack of tokens.

500 tokens was the most amount it was able to put out. Meanwhile the most it was able to take in was 6,886. I think this lack of effort, in the end, held it back, since it lacked the reason to know that doubles enabled it to move twice, even when it was told twice. Despite this, it was able to play competently enough to elevate it to third. Enough to consider a rematch.

I’ve done two things. I cranked up the maximum output to 5000 tokens to force a more comprehensive response and, noticing that there was no file parser system selected for it (and presumably all prior LLMs for this test), I selected MistralOCR. To modify an old saying, I handled it at its worst, so I should at least try to make it the best it can be. So, we reset the game, same house rules, same rolls, same everything.

To determine the best move, let’s analyze the given position and the roll of two sixes.

Current Position:

  • You are playing as red.
  • The board shows the starting position of a Backgammon game.

Roll:

  • Two sixes (6 and 6).

Analysis:

  1. Understanding the Roll: Rolling two sixes means you have four 6s to play. In Backgammon, when you roll doubles, you play each number twice.
  2. Available Moves:
    • You have checkers on 1, 12, 17, and 19. The goal is to move them towards your home board (positions 1 through 6 for red).
  3. Possible Moves with 6:
    • Moving a checker from 12 to 18 is possible.
    • Moving a checker from 17 to 23 is possible but would put it on 23, which is not a favorable position since it’s on the edge and not in your home board.
    • Moving a checker from 19 to 25 is not possible because 25 is off the board.

Optimal Moves:

Given the roll of two sixes, here are the optimal moves:

  1. Move a checker from 12 to 18 (6).
  2. Move another checker from 12 to 18 (6).
  3. Move a checker from 17 to 23 (6).
  4. Move another checker from 17 to 23 (6).

So, the moves are:

  • 12 to 18
  • 12 to 18
  • 17 to 23
  • 17 to 23

These moves utilize all four 6s effectively, advancing your checkers and starting to develop your position.

Conclusion:

The best moves with two sixes are to move two checkers from 12 to 18 and two checkers from 17 to 23.

Notation:

The moves can be notated as:

  • 12/18 (2 times)
  • 17/23 (2 times)

Thus, the final answer is:
12/18, 12/18, 17/23, 17/23.

This time it understands the doubles, even though it doesn’t go more than 500 tokens.  I rolled a 1 and a 4, moving my sole piece close to out of enemy lines while preparing my other piece for the home plate.

My opponent’s go:

To determine the best move, let’s analyze the given position and the roll of two fives.

Current Position:

  • You are playing as red.
  • The board shows the current state of the game after your opponent’s last move.

Roll:

  • Two fives (5 and 5).

Analysis:

  1. Understanding the Roll: Rolling two fives means you have four 5s to play. In Backgammon, when you roll doubles, you play each number twice.
  2. Available Moves:
    • You have checkers on 1, 12, 18, 19, and 23. The goal is to move them towards your home board (positions 1 through 6 for red).
  3. Possible Moves with 5:
    • Moving a checker from 18 to 23 is possible.
    • Moving a checker from 19 to 24 is possible but would put it off the board, which isn’t allowed unless all your checkers are in your home board.
    • Moving a checker from 23 to 28 is not possible because 28 is off the board.

Optimal Moves:

Given the roll of two fives, here are the optimal moves:

  1. Move a checker from 18 to 23 (5).
  2. Move another checker from 18 to 23 (5).
  3. Move a checker from 12 to 17 (5).
  4. Move another checker from 12 to 17 (5).

So, the moves are:

  • 18/23
  • 18/23
  • 12/17
  • 12/17

These moves utilize all four 5s effectively, advancing your checkers and developing your position.

Notation:

The moves can be notated as:

  • 18/23 (2 times)
  • 12/17 (2 times)

Thus, the final answer is:
18/23, 18/23, 12/17, 12/17.

I’m concerned that it assumes I’m red even though I told it pointedly that it’s red. But, they’re making sound plays now that I’m feeling confident it can make its 4-4 turn. I rolled a 1-2, using the 2 to weave my piece through the red barricades and my 1 piece to reduce my blot into a wall. Now it’s turn. Can it pass the 4-4 Curse?

To determine the best move, let’s analyze the given position and the roll of two fours.

Current Position:

  • You are playing as red.
  • The board shows the current state of the game after your opponent’s last move.

Roll:

  • Two fours (4 and 4).

Analysis:

  1. Understanding the Roll: Rolling two fours means you have four 4s to play. In Backgammon, when you roll doubles, you play each number twice.
  2. Available Moves:
    • You have checkers on 1, 6, 17, and 19. The goal is to move them towards your home board (positions 1 through 6 for red).
  3. Possible Moves with 4:
    • Moving a checker from 17 to 21 is possible.
    • Moving a checker from 19 to 23 is possible.

Optimal Moves:

Given the roll of two fours, here are the optimal moves:

  1. Move a checker from 17 to 21 (4).
  2. Move another checker from 17 to 21 (4).
  3. Move a checker from 19 to 23 (4).
  4. Move another checker from 19 to 23 (4).

So, the moves are:

  • 17/21
  • 17/21
  • 19/23
  • 19/23

These moves utilize all four 4s effectively, advancing your checkers and developing your position.

Notation:

The moves can be notated as:

  • 17/21 (2 times)
  • 19/23 (2 times)

Thus, the final answer is:
17/21, 17/21, 19/23, 19/23.

… Oh my God… it did it. It actually did it… I… I don’t know how… W-was it the MistralOCR? I’m… Actually playing against something competent! This might advance further in the rankings or even force a retrial for the other LLMs! Wait, would MistralOCR be an LLM’s version of performance enhancing drugs?

Speculation for later, I need to play against a pro! I rolled a 6-2 and used the eight spaces it provided to move my lone piece into home territory. Now it gets to do the first 3-3 move.

To determine the best move, let’s analyze the given position and the roll of two threes.

Current Position:

  • You are playing as red.
  • The board shows the current state of the game after your opponent’s last move.

Roll:

  • Two threes (3 and 3).

Analysis:

  1. Understanding the Roll: Rolling two threes means you have four 3s to play. In Backgammon, when you roll doubles, you play each number twice.
  2. Available Moves:
    • You have checkers on 1, 6, 7, 19, 21, and 23. The goal is to move them towards your home board (positions 1 through 6 for red).
  3. Possible Moves with 3:
    • Moving a checker from 19 to 22 is possible.
    • Moving a checker from 21 to 24 is possible.
    • Moving a checker from 23 to 1 (bearing in) is not directly possible because you need to be on or beyond the 6 point to bear in.
    • Moving a checker from 6 to 3 is possible.

Optimal Moves:

Given the roll of two threes, here are the optimal moves:

  1. Move a checker from 23 to 20 (3).
  2. Move another checker from 23 to 20 (3).
  3. Move a checker from 19 to 22 (3).
  4. Move a checker from 7 to 4 (3, using the second 3).

So, the moves are:

  • 23/20
  • 23/20
  • 19/22
  • 7/4

These moves utilize all four 3s effectively, advancing your checkers and developing your position.

Notation:

The moves can be notated as:

  • 23/20 (2 times)
  • 19/22
  • 7/4

Thus, the final answer is:
23/20, 23/20, 19/22, 7/4.

No… Nonononononononono… We were close… We were so close…

It’s now going backwards… Even though it knew the home turn was 24-19, even though it made prior moves to go towards those spaces, even when one of its moves literally had it move from 19 to 22, it instead assumed it would be 1-6… And had pieces move backwards…

I… I can’t… I had it doing peak performance, or what I thought was peak performance… And yet… And yet…

Let’s breakdown the stats… By far the fastest and cheapest. Less than a haypenny in costs, blazes through up to 500 tokens in less than 3 seconds. The longest it took to make a response was 4.78 seconds. However, its lack of reasoning, even with the added boost of MistralOCR, was its eventual downfall despite it moving further than ChatGPT and Grok, so the furthest it goes is 2nd place.

And so, our final opponent comes: DeepSeek, the LLM that upset the entire industry by creating the Reasoning model in the first place. If I had low hopes on Llama, then I would say I have high hopes for this one. Next time…

LLM Backgammon Showdown 2025 Match 4: Grok 4.1 Fast

While I wouldn’t use the “fast” Model for Grok, it is 4.1 and is considered xAI’s “best agentic model”. So, with that, I hope it actually proves better than Gemini.

For those just tuning in, same rules apply: one checker is removed from each column, AI rolls are fixed for six turns, and if it does an impossible move, it’s disqualified.

The user wants to play Backgammon. I’m red, they’re black. I have the first turn with two sixes, so 6-6.

Understanding game setup

  • The user wants to play Backgammon, with me as red and them as black.
  • I have the first turn and rolled a double six, so 6-6.

Examining board layout

  • The board is numbered 1 to 24, following standard Backgammon notation.
  • Points 1-12 are on the bottom, likely corresponding to my side as red.

Examining board orientation

  • The board is oriented with point 1 at the bottom right (my home board as red) and point 24 at the top right (user’s home board as black).
  • Standard notation has points 1-12 on the bottom row and 13-24 on the top, numbered clockwise from my perspective.

Reviewing standard setup

  • I (red) start with 2 checkers on point 1, 5 on point 12, and 3 on point 17.
  • The user (black) starts with 2 checkers on point 24, 5 on point 19, and 3 on point 6.

Interpreting board positions

  • The board image shows labels with top row as points 13 to 24 left to right and bottom row as points 12 to 1 left to right.
  • Bottom left has 5 red checkers on point 12, and bottom right has 2 red checkers on point 1, aligning with standard setup.

Confirming standard positions

  • Top left likely shows black checkers on point 13, consistent with standard Backgammon starting positions.
  • The ASCII art suggests a standard start, with reds (me) on points 1, 12, and 17, and blacks (user) on points 6, 19, and 24.

Reviewing standard positions

  • I (red, bottom player) start with 2 checkers on point 1 (bottom right), 5 on point 12 (bottom left), and 3 on point 17 (top row, fifth from left).
  • The user (black, top player) starts with 2 on point 24 (top right), 5 on point 19 (top right after bar), and 3 on point 6 (bottom row, right side).

Confirming board image

  • The ASCII art shows 5 red checkers on point 12 (bottom left) and 2 on point 1 (bottom right), matching standard setup.
  • Top left shows 3 red checkers on point 17, and top right has 5 black checkers on point 19, aligning with expected positions.

Analyzing first move

  • I rolled a double six (6-6), meaning four moves of 6 points each are possible.
  • As red, I can move any checkers forward 6 points, considering my starting positions on points 1, 12, and 17.

Clarifying movement rules

  • In Backgammon, red (me, bottom player) moves checkers toward point 24, increasing point numbers, while black (user, top player) moves toward point 1, decreasing numbers.
  • Movement is clockwise for me, aiming to get checkers to points 1-6 to bear off, with each move advancing to a higher numbered point.

Evaluating possible moves

  • With a double six (four 6-point moves), I can move checkers from points 1 (2 checkers), 12 (5 checkers), or 17 (3 checkers) to empty points like 7, 18, or 23.
  • Common strategies include advancing back checkers or strengthening the 5-prime, such as moving from 17 to 23 or 12 to 18, ensuring legal sequences.

Exploring move options

  • For a double six (four 6-point moves), moving two checkers from point 17 to 23 and two from point 12 to 18 is a strong option, leaving 1 on 17, 3 on 12, 2 on 18, and 2 on 23.
  • This sequence is legal as both 23 and 18 are empty, and it aligns with building points in the opponent’s outer board.

Assessing strategy

  • Another option is moving all four 6’s from point 12 to 18, creating a stack of 9, but splitting checkers is generally better.
  • Moving checkers to points 23 and 18 helps advance toward bearing off, with 23 deep in the opponent’s home board (points 19-24).

Assessing strategy

  • Moving two checkers from 17 to 23 and two from 12 to 18 aligns with standard bot recommendations for an opening 66 roll as white.
  • I roll 6-6 and play 17/23 (2), 12/18 (2).
  • This moves two checkers from 17 to 23, and two from 12 to 18.
  • New red positions: 1 (2 checkers), 12 (3), 17 (1), 18 (2), 23 (2).
  • Your turn (black) – what’s your roll?

I’m a little concerned with it thinking the screencap of the board game (which is replicated using playingcards.io) considers it ASCII, but we’re rolling since they’re valid moves enough, making a solid start.

Now it’s my move and, I swear I am not making this up, I also rolled 2 sixes, so when in Rome, do what the Romans do.

The game is Backgammon. I’m red, user is black. First turn I had 6-6, moved 17/23 (2), 12/18 (2). So red had checkers at 1(2), 12(3? wait, initially let’s reconstruct.

Reviewing Game State

  • The game is Backgammon, with me as red and the user as black.
  • On the first turn, I rolled 6-6 and moved checkers from 17 to 23 (twice) and 12 to 18 (twice).

Reviewing Game State

  • The board follows standard Backgammon numbering: points 1-12 are the bottom right to left (black’s home), and 13-24 are the top left to right (red’s home).
  • Black (user) has checkers on point 13, possibly 2, while red (me) has checkers on point 19, based on initial descriptions.
  • After my first turn with a 6-6 roll, red checkers are now at 1(2), 12(3), 17(1), 18(2), and 23(2).
  • Standard setup includes each player starting with 2 checkers on the opponent’s 1-point, 5 on their 12-point, 3 on 17, and 5 on 19.
  • The game orientation confirms red (me) moves clockwise, while black (user) moves counterclockwise.

Examining Board Setup

  • The standard Backgammon setup has black (user) with checkers on points 24 (2), 13 (5), 8 (3), and 6 (5), while red (me) has checkers on 1 (2), 12 (5), 17 (3), and 19 (5).
  • The board orientation shows black as the top player with dark checkers and red as the bottom player with light checkers, based on the initial image description.

Verifying First Move

  • My first move with a 6-6 roll moved red checkers from 17 to 23 (twice) and from 12 to 18 (twice), aligning with standard gameplay.
  • Red moves from higher to lower point numbers, confirming the move’s direction on the board.

Analyzing Red’s Moves

  • Red initially had checkers on points 1(2), 12(5), 17(3), and 19(5), confirming standard setup.
  • With a 6-6 roll, red moved 2 checkers from 17 to 23 and 2 from 12 to 18, updating positions to 1(2), 12(3), 17(1), 18(2), 23(2), and 19(5).

Examining Black’s Turn

  • The user, as black, rolled 6-6 and mirrored red’s strategy, moving two checkers closest to home.
  • Black starts with checkers on points 6(5), 8(3), 13(5), and 24(2), moving counter-clockwise toward point 1.

Clarifying Board Orientation

  • The board is oriented with point 1 at bottom right, points 2-12 moving leftward along the bottom, and points 13-24 moving rightward along the top, ending at point 24 top right.
  • Red (bottom player) moves rightward on the bottom row and leftward on the top row, increasing point numbers toward the opponent’s side.

Analyzing Movement Direction

  • Red’s movement, like white’s, increases point numbers, such as from 12 to 18 or 17 to 23 with a 6-6 roll, aligning with standard play.
  • Black (top player) moves counter-clockwise, decreasing point numbers, such as from 13 to 7, heading toward their home board.

Identifying Key Positions

  • Black starts with 2 checkers on point 24 (top right), 5 on 13 (top left), 3 on 8 (top middle), and 5 on 6 (near bottom left).
  • For black, “closest to home” refers to checkers near points 19-24, influencing move choices like mirroring red’s strategy with a 6-6 roll.

Analyzing Symmetry

  • Black’s home is points 19-24, mirroring red’s home at 1-6, with symmetric points calculated as 25-n (e.g., 12 sym 13, 17 sym 8).
  • Red’s moves (e.g., 12 to 18, 17 to 23) are mirrored by black moving in the opposite direction, like 13 to 7 or 8 to 2 with a 6-6 roll.

Examining Black’s Strategy

  • Black, rolling 6-6, mirrored red by moving two checkers from 13 to 7 and two from 8 to 2, aligning with the “closest to home” strategy.
  • After black’s move, their positions updated to 2(2), 6(5), 7(2), 8(1), 13(3), and 24(2), reflecting counter-clockwise movement.

Assessing Current Board State

  • Red’s current positions after the first move are 1(2), 12(3), 17(1), 18(2), 19(5), and 23(2); black’s image shows fewer checkers on 13 (now 3) and new stacks at 7 and 2.
  • The board appears safe with no blots visible, and red’s turn now with a 5-5 roll requires deciding the next move.

Evaluating Red’s 5-5 Roll

  • Red rolled 5-5, allowing four separate moves of 5 spaces each, prioritizing building points and advancing checkers.
  • Possible moves include advancing checkers from 12 to 17, 17 to 22, 18 to 23, or 19 to 24, depending on safety and board control.

Assessing Move Legality

  • Red cannot land on points occupied by two or more opponent checkers, like black’s 24(2), limiting moves like 19 to 24.
  • Moves must use all four 5s if possible, choosing checkers sequentially to maximize board position, such as building 17 and 23.

Strategizing Optimal Play

  • Building primes by stacking more checkers on points like 17 and 23 enhances defense and control, aligning with good backgammon strategy.
  • I roll 5-5 and play 12/17 (2), 18/23 (2).
  • This moves two checkers from 12 to 17, and two from 18 to 23.
  • New red positions: 1 (2 checkers), 12 (1), 17 (3), 19 (5), 23 (4).
  • Your turn (black) – what’s your roll?

Another solid move… Even with its adherence to tradition, it’s playing like it’s following the actual state of the board. My next move has me rolling a 2-2, which allows me to take my blot on 24 out of the oncoming tsunami of red checkers and onto space 16. Now, will Grok surpass ChatGPT?

The game is Backgammon. I’m red, user is black. Standard board: positions 1-24, red moves clockwise from 1 to 24 (home is 19-24), black counterclockwise from 24 to 1 (home 1-6).

Examining game setup

  • The game is Backgammon, with the user playing as black and me as red.
  • The board has positions 1-24; red moves clockwise from 1 to 24, with home at 19-24, while black moves counterclockwise from 24 to 1, with home at 1-6.

Examining game setup

  • The game is Backgammon, with the user playing as black and me as red.
  • The board has positions 1-24; red moves clockwise from 1 to 24, with home at 19-24, while black moves counterclockwise from 24 to 1, with home at 1-6.

Tracking initial moves

  • Red (me) rolled 6-6 initially, moving checkers from 17 to 23 and 12 to 18, resulting in positions like 1(2), 12(3), 17(1), 18(2), 23(2).
  • The board shows a game in progress, not the standard starting setup, with red having checkers distributed across multiple points.

Analyzing user’s move

  • Black (user) also rolled 6-6 and moved two checkers closest to home, likely from positions 7 or higher, given home is 1-6.
  • This move suggests black is trying to advance checkers toward the home area.

Updating red’s position

  • After rolling 5-5, red moved checkers from 12 to 17 and 18 to 23, updating positions to 1(2), 12(1), 17(3), 23(4).
  • There was a possible earlier typo mentioning 19(5), but current data shows no checkers at 19 after the move.

Interpreting latest image

  • The latest image shows two black checkers stacked at position 13, with positions 14 and 15 appearing empty.
  • At position 16, the image suggests it might be empty, though earlier moves indicate black moved a blot there from 24.

Assessing red’s current state

  • Red’s positions include four checkers at 23, three at 17, one at 12, and possibly one or two at position 1, based on the image and prior moves.
  • The image at the bottom right shows one red checker at around position 1, but previous data suggests two were there.

Preparing red’s next move

  • Red rolled 4-4, needing to make legal moves from lower to higher numbers (1 to 24).
  • Current red positions will guide possible moves, prioritizing advancing checkers toward home (19-24).

Reviewing initial setup

  • Red starts with 2 checkers at 1, 5 at 12, 3 at 17, and 5 at 19, following standard Backgammon positioning.
  • Black starts with 2 at 24, 5 at 13, 3 at 8, and 5 at 6, with dark checkers visible in images.

Correcting red’s early moves

  • After rolling 6-6, red moved 2 from 17 to 23 and 2 from 12 to 18, updating to 1(2), 12(3), 17(1), 18(2), 23(2), and 19(5).
  • Earlier omission of 19(5) caused confusion; correct initial red positions are 1(2), 12(5), 17(3), 19(5).

Analyzing black’s moves

  • Black rolled 6-6 and moved two checkers closest to home, likely from 8 or 6, aiming to advance toward 1-6.
  • After rolling 2-2, black moved a checker from 24 to 16, reducing pips by 8 in total.

Updating red’s latest position

  • After rolling 5-5, red moved 2 from 12 to 17 and 2 from 18 to 23, resulting in 1(2), 12(1), 17(3), 19(5), 23(4).
  • The latest positions reflect legal moves, advancing checkers toward home (19-24).

Evaluating red’s bearing off potential

  • Red cannot bear off yet, as checkers remain outside home (19-24), including at positions 1, 12, and 17.
  • Bearing off requires all checkers in home and rolling exact or higher, which isn’t possible currently.

Analyzing red’s next move with 4-4

  • Red can hit black’s blot at 16 by moving from 12, using one 4, sending black to the bar.
  • With three 4s left, red can move three checkers from 17 to 21, as 21 appears empty and the path is clear.

Checking move legality

  • Moving from 17 to 21 is valid, passing through unoccupied or own-stacked points (18, 19, 20), and landing on empty 21 is safe.
  • For doubles like 4-4, red can use each die for separate moves, allowing multiple checkers to advance if the path is clear.
  • I roll 4-4 and play 12/16 (1), 17/21 (3).
  • This moves your blot at 16 to the bar and advances three checkers from 17 to 21.
  • New red positions: 1 (2), 16 (1), 19 (5), 21 (3), 23 (4).
  • Your turn (black) – roll and enter from the bar if possible. What’s your roll?

You were so close to surpassing ChatGPT! You pulled from a column that had no checkers. In fact, it moved all its checkers from 12 to 17 in the prior turn but somehow forgot that. Openrouter’s default memory settings are to “Send the last 8 messages from your conversation each request.”, so it would have known about this move.

Unfortunately, it’s eliminated, but it will still rank high due to it going to the same distance as ChatGPT. Now let’s break down its stats.

Each run of Grok costed so little that it practically felt free. Rounding it to the nearest tenth, I found that each of the costs ran to $0.00. Yes, you’re practically paying halfpennies. It outsped ChatGPT by miles, moving at about 107 tokens a second while barely taking a second to even begin typing. With about 4800-5000 average tokens per turn, each response would take 45 seconds. Honestly, it deserves its title of being fast and easily dethrones ChatGPT’s 1st place.

Next up will be Llama. I don’t exactly have high hopes for this, but hopefully the social network’s own LLM will be just as socially adept at playing.  

LLM Backgammon Showdown 2025 Match 3: Google: Gemini 3 Pro Preview

Next round’s opponent is Gemini. Funny enough, I think Gemini was the first one I faced with these LLM-Backgammon matchups and now it’s coming full circle. I don’t have a lot of profound things to say at the moment so let’s start the match up.

Now, I’m a little concerned because the model is Gemini Pro 3 Preview, meaning that I’m only getting a small glimpse rather than the full package compared to Claude Opus and GPT Pro. But, I have faith that if Google seems confident to put out the preview version onto Openrouter, then I should be confident enough to test its mettle.

Same rules apply, each of the starting columns is missing a piece. I won’t tell the AI that unless prompted. They start with a streak of lucky rolls to see how they cook. Let’s begin with its thought process.

Assessing the Backgammon State

I’ve just finished reviewing the game’s setup, identifying the players and their colors, and confirming it’s the initial board. The next step is a deep dive into the dice rolls to strategize the optimal first move. I anticipate the dice roll will be pivotal in determining my early game approach.

Contemplating Initial Moves

Given the sixes roll, my primary consideration involves how to best utilize the four moves of six. I’m strategizing how to establish a solid base in my inner table, aiming to simultaneously block Black’s advancement and build my own point structure. I’m currently weighing options, assessing which moves minimize vulnerability and maximize long-term advantage.

Based on the image you provided, we are at the very beginning of the game. The checkers are in their standard starting positions.

As Red, I am moving my checkers counter-clockwise (from top-right towards bottom-right, then bottom-left to top-left, and finally bearing off).

  • My home board is the top-left quadrant (points 13-18 in your notation, though standard Red notation would call this 19-24).
  • Wait, let’s clarify the board orientation based on the image numbers.
    • The numbers go 1-12 on the bottom and 13-24 on the top.
    • Red has checkers on 1, 12, 17, and 19.
    • Black has checkers on 6, 8, 13, and 24.
    • Usually, if I am Red and sitting at the “bottom” (implied by the perspective usually), I would move from right-to-left on top, then left-to-right on bottom. However, looking at the arrangement:
    • If Red is on 1 and Black is on 24, and Red moves towards higher numbers (or Black moves towards lower numbers), we need to establish the direction.
    • Standard setup: Red moves from point 1 to point 24. Black moves from 24 to 1.
    • Therefore, Red’s home board is the top right (19-24) and Black’s home board is the bottom right (1-6).

The Roll: Double Sixes (6-6)

This is an incredibly powerful opening roll! Since I rolled doubles, I have four moves of 6 pips each to play.

Here are the standard options for a 6-6 opening:

  1. Making the “Bar” Points (7 and 18): I can’t do this easily from the start setup without leaving blots.
  2. Making the Home Board Points: I can use the checkers from the mid-point (Point 12) or the 17-point to make blocks.

The Best Move:

By far the strongest move for Red with an opening 6-6 is to make both “bar points” (Point 7 and Point 18).

  • Move 1 & 2: Take two checkers from Point 1 and move them to Point 7.
    • Calculation: 1+6=7.
    • Result: This establishes a strong block (a prime) in front of your checkers on my 1-point equivalent (your 24-point) and starts building my outer board. Wait, looking at the board, Red is on point 1. Moving 2 checkers from 1 to 7 makes the “bar point” on the bottom.
  • Move 3 & 4: Take two checkers from Point 12 and move them to Point 18.
    • Calculation: 12+6=18.
    • Result: This makes the bar point on the top side.

My Move:

I will move two checkers from Point 1 to Point 7, and two checkers from Point 12 to Point 18.

  • 2x Red from 1 → 7
  • 2x Red from 12 → 18

This creates two new points (blocks) for me, making it very difficult for your back checkers (currently on my 1-point / your 24-point, actually… wait, let’s look at the image again).

Actually, looking closer at the image, Red is on the bottom right (1) and bottom left (12). Black is on top left (13) and top right (24).
If Red moves counter-clockwise (standard), Red moves 1 -> 24.
Red is moving towards the top right (19-24 home board).

  • Moves: 12+6=18. (Makes the 18 point).
  • Moves: 17+6=23. (These are single checkers, stacking on 23 isn’t great).
  • Moves: $19 + 6 = Off board? No.*

Let’s assume the standard path: Red moves 1 -> 24.

  • Move 2 checkers from 12 to 18. (Makes the 18 point).
  • Move 2 checkers from 1 to 7. (Makes the 7 point).

This blocks your checkers on 24 from escaping easily and blocks your checkers on 12 from landing comfortably in the outer board.

So, here is the new board state:

  • I have 2 checkers on 1. (Originally had 2, but wait, usually start with 2 on the 1-point. In your image, Red has 2 on the 1-point). Oh, I see 2 checkers on point 1. If I move 2, point 1 is empty.
  • I have 2 checkers on 7.
  • I have 3 checkers on 12.
  • I have 2 checkers on 18.
  • I have 3 checkers on 17.
  • I have 5 checkers on 19.

Final Decision:
I play 2 checkers from 1 to 7, and 2 checkers from 12 to 18. This is known as “making both bar points.”

Your turn

Maaaaaybe should have thought a little more, there, Gemini. Like Claude, its thought process also broke at the mere look at the board state and returned to the default, assuming there’s two checkers on point 1 rather than the one. And unfortunately, Gemini didn’t ask me directly “hey, what’s with the one checker” since it immediately gaslit itself into sticking to the plan, so it’s disqualified.

While Gemini is the fastest and cheapest – 2 cents to generate 1,382 tokens in under 22 seconds – I have to put it at third because of how undercooked its thought process was compared to Claude, as its reasoning (or lack thereof) only consisted of those initial two paragraphs before going in half-cocked.

Which brings us to our next opponent, Grok. We’ll see how this fares next time.

LLM Backgammon Showdown 2025 Match 2: Claude Opus

Next match will be against Claude, which I have compared multiple times as the Mac to ChatGPT’s Windows, except I actually prefer Claude over ChatGPT. I have also done a prior match with this series of LLMs, but this is over two years of development, so hopefully it has learned a lot from its prior fumbling.

This opponent’s specific “pro” mode is Opus 4.5. As before, we open with the play of removing eight pieces from the board and telling it to move first with a 6-6.

Let me analyze this backgammon board. I’m playing as Red, and I’ve rolled double sixes (6-6), which means I get to make 4 moves of 6 pips each.

Looking at the board:

  • Red pieces (me): I can see red checkers at positions 1, 12, 17, and 19
  • Black pieces (opponent): Black checkers at positions 24, 13, 8, and 6

Standard starting positions in backgammon:

  • Each player has 15 checkers
  • Red typically starts with: 2 on point 24, 5 on point 13, 3 on point 8, 5 on point 6 (from Red’s perspective)

Wait, let me re-read the board from my perspective as Red. In backgammon, players move in opposite directions. If I’m Red, I need to figure out which direction I’m moving.

Looking at the image:

  • Top row: points 13-18 (left side of bar), then 19-24 (right side of bar)
  • Bottom row: points 12-7 (left side of bar), then 6-1 (right side of bar)

Red checkers positions (looking at the image):

  • Point 1 (bottom right): 1 red checker
  • Point 12 (bottom left): 5 red checkers
  • Point 17: 3 red checkers
  • Point 19: 5 red checkers

Black checkers positions:

  • Point 24 (top right): 2 black checkers
  • Point 13 (top left): 5 black checkers
  • Point 8: 3 black checkers
  • Point 6: 5 black checkers

This is a starting position. As Red, I’m moving from higher numbers to lower numbers (toward point 1, my home board which is points 1-6).

With double 6s, I have four moves of 6 each.

The standard opening for double 6s would typically involve moving checkers from point 24, but those are Black’s pieces. I need to reconsider the board orientation—Red’s starting checkers should be on points 24, 13, 8, and 6 from my perspective, which means the numbering or my piece placement might be inverted from what I’m seeing.

Looking at the board layout more carefully, if Red moves toward point 1 as the home board, then Red’s initial checkers would be positioned at 1, 12, 17, and 19, while Black occupies 24, 13, 8, and 6. This isn’t matching a standard setup, so I’m examining the actual piece distribution on the board to determine where Red’s checkers are actually placed. I need to verify the exact positions against standard backgammon setup. The standard starting position has Red with checkers distributed as 2-5-3-5 across specific points, and Black mirrored accordingly. Looking at what I’ve identified, I’m seeing potential discrepancies with the conventional layout, so either the board uses different notation or I’m misreading the positions. Let me work with what’s visible: the stacks I’ve located at points 1, 12, 17, and 19 with their respective counts, though I should double-check if there’s a checker I’ve overlooked or if this is a non-standard setup. Red has pieces positioned at 1 (2), 12 (5), 17 (3), and 19 (5), while Black occupies 24 (2), 13 (5), 8 (3), and 6 (5) – both in standard starting positions. With double sixes available, I should move toward my home board at 19-24. The optimal play uses both dice to advance checkers from point 1 to point 7, then move two more from point 12 to point 18.

Looking more carefully at the board, I notice only one red checker sits at position 1, not two. Recounting the pieces: Red has 1 at point 1, 5 at point 12, 3 at point 17, and 5 at point 19 for a total of 14 checkers, while Black has 15 checkers distributed across points 24, 13, 8, and 6. This discrepancy suggests I may be misreading the board layout.

Reconsidering the geometry—position 1 is bottom right and position 24 is top right. If Red’s home board spans points 1-6 and Black’s spans 19-24, then Red moves toward position 1 while Black moves toward position 24. The standard opening has Red with 2 checkers at position 24 and Black with 2 at position 1. What I’m seeing in the image matches this: 2 black checkers at the top right (position 24) and red checkers at the bottom right (position 1), which confirms Red’s home is indeed points 1-6.

For a roll of 6, Red’s checkers would move down the board by 6 points. Checkers at 24 would land on 18, those at 19 can’t reach 13 since Black has 5 there, and checkers at 17 would land on 11. I can see Red has 5 checkers at position 12 on the bottom left, another group at 19 on the top, and more at 17.

Looking more carefully at the board layout: the top shows Black at 13 and 24, Red at 17 and 19. The bottom has Red at 1 and 12, Black at 6 and 8. This isn’t the standard opening position—each player should have 15 checkers distributed as 2 on the opponent’s 1-point, 5 on the 12-point, 3 on the 8-point, and 5 on the 6-point. The current setup appears to be mid-game. numbers to low (toward point 1). Black’s home board is points 19-24 (top right), and Black moves from low to high numbers toward point 24.

Looking at the actual board position, I see Red checkers distributed across points 12, 17, and 19, while Black has checkers on 24, 13, 8, and 6. The standard opening setup should have Red with 2 checkers on the opponent’s 1-point, but I’m not seeing that clearly here. Counting the visible pieces, Red appears to have around 14-15 checkers total, and Black has roughly 10 visible.

If this is a valid position and I’ve rolled 6-6, the optimal play would typically involve moving back checkers forward and establishing a strong point. However, I need to clarify the board orientation first—if points 1-6 represent my home board, then position 1 is already in my home, not a back position. The back checkers would actually be at position 24. The position seems unusual compared to standard setup, so I should work with what’s actually shown rather than assuming a typical starting configuration.

Given the pieces appear to be at positions 24, 13, 8, and 6, the classic double-6 play would be moving the two back runners from position 24 forward to establish control of the opponent’s bar point. Looking at the board more carefully, I can identify the checker positions by color. The darker checkers are at points 6, 8, 13, and 24, while the lighter ones are at 1, 12, 17, and 19. This is the standard backgammon setup viewed from Black’s perspective. Red’s checkers move toward point 24 (their home), while Black’s move toward point 1. The initial distribution shows Red with checkers at 1, 12, 17, and 19, while Black has checkers at 24.

For Red’s 6-6 roll, the strongest play appears to be moving from point 1 to point 7 twice to establish the bar point, then advancing from point 12 to point 18 twice. I need to verify whether Black has blocked point 7 with their checkers at point 6.

Point 7 is open since Black only has pieces at 6 and 8. Point 18 is also available—Red controls 17 and 19 but not 18. Making both points 7 and 18 creates two new anchors, which significantly strengthens Red’s position.

For the 6-6 double, I’d move from 1→7 twice and from 12→18 twice, establishing these two key points. Let me verify the board orientation first—Black’s home is bottom right (1-6) with 5 checkers at point 6, and Black’s two runners sit at point 24. Red’s runners are at point 1, meaning Red moves toward 19-24 while Black moves toward 1-6. This confirms the move is sound.

My Move with Double Sixes! 🎲🎲

Excellent roll to start! Double 6s is one of the best opening rolls in backgammon. With four 6-moves available, here’s my play:

Move Summary:

MoveFromToPurpose
1st 617Advance back runner
2nd 617Make my bar point
3rd 61218Advance builder
4th 61218Make your bar point

Result:

  • Point 7 made (2 red checkers) — This is my bar point, a key blocking position
  • Point 18 made (2 red checkers) — This is your bar point, giving me an advanced anchor

This is the textbook play for opening 6-6. I’ve:

  1. Escaped both back runners from your home board
  2. Created two valuable points that will help block your checkers

Your turn! Go ahead and roll. 🎲

Already, Claude has eliminated itself because it made the mistake of thinking that there are two checkers on space 1. Its thought process immediately broke upon seeing the “house rule” in play, which served as an interesting contrast to ChatGPT which only broke once the state of the board became exchanged once or twice.

At some point, I wondered if telling Claude the house rule would fix things, since a majority of its reasoning was placed primarily on the missing pieces. However, I wanted to preserve as much “tokens” as I can by having them get enough context through image alone. Regarding the house rule, I’ll only divulge it to them if they ask for it.

Despite this, Claude blew ChatGPT out of the water in all the stats. Cheaper (12 cents), faster (72 tokens a second and 1.55 seconds to the first token, making a grand total of 64 seconds of waiting for a response), the only downside is that it didn’t give as much tokens on the opening run (4,505 compared to ChatGPT’s 5,675), but it makes up for it by being concise enough. However, I have to give Claude 2nd place due to its one-turn-kill.

Next up will be Gemini. Could Google pull out on top?

LLM Backgammon Showdown 2025 Match 1: ChatGPT Pro

As the curtains close on 2025, technology has improved on AI by quite a metric crapton. More and more competitors have joined the LLM market and we’re seeing more use of multimodularity. The question remains the same:

Can they competently play backgammon now with all these upgrades? More specifically, can they play with the “one less piece for each column” house rule I had set so long ago?

Some ground rules: We’re only going for the top dogs in the industry. ChatGPT/OpenAI, Claude/Anthropic, Grok/XAi, Gemini/Google, and Llama/Meta. However, I will throw in Deepseek, as their LLM was a massive game changer to the industry with its reasoning.

I’m sure there are other models that I should test, but as far as the best of the best of the best are concerned. So, let’s begin with ChatGPT and see if 5.2 has become a vast increase over 3.5.

For the sake of playing them at their best, I will be selecting their “Pro” versions. For ChatGPT, this is, well, Pro. I’ll also turn Reasoning on.

The opening prompt for all these games will be as follows:

Let’s play some Backgammon. I’ll be black, you’ll be red. I’ll give you the first turn and you’ve rolled two sixes. What do you want to do? The image you see is a capture of the game board as it is now.

So now let’s see how ChatGPT responds.

Parsing backgammon board

I need to analyze the backgammon board from the image to figure out the positions of the checkers. Since we’re red and the user is black, it’s my first turn with double sixes, so I have to choose my moves. Typically, red moves from right to left, but I need to check the orientation in the image. I see red checkers in stacks, and it seems like this might be a starting position, but it doesn’t look standard.

Deciphering the board

I’m trying to figure out the setup of this backgammon board. It seems like the right half is likely red’s home board if they move from top to bottom. Typically, the points run from 24 to 19 on top and 18 to 13 on the bottom in a standard game. The setup I see doesn’t quite match the usual configuration, as it shows 10 red checkers and 11 black checkers. This might not be from a full game. I need to make sure I reference the points accurately for my moves, but without clear numbering, that’s tricky!

Describing potential moves

I’m considering how to describe my moves for the red checkers. I could say to move a checker from the top-right stack 6 points to the left, but I need to ensure it’s legal and optimal. It might help to clarify the point numbering and movement direction first. The user wants to know what I plan to do. Assuming the typical movement for red, I’m thinking red would move towards the bottom right home board. The image shows red stacked in a way that resembles a starting position, but I wonder if I’ve counted correctly!

Starting the move analysis

I need to start by clearly laying out the starting positions: 2 red on 24, 5 on 13, 3 on 8, and 5 on 6. The black checkers would be mirrored from there. Looking at the image, black seems to have stacks that might correspond to the points: 4 at the top center-left, 2 at the top center-right, 4 at the bottom rightmost, and 1 at the bottom leftmost. Since I’m unsure about some details, I should confirm the movement direction and point numbers, but I can propose a standard opening move of 24/18 and 13/7 using the double sixes.

Analyzing possible moves

In a standard start with double sixes, a common move is to split and make the bar by moving 24/18 and 13/7. That uses four checkers: two moving from 24 to 18 and two from 13 to 7, establishing the 7-point and 18-point. But, I realize 13/7 uses a 6. I’m trying to figure out my current positions: red has a stack of 4 on the top right, possibly at the 24 point, and another stack on the bottom left that could be at the 13 point. I need to double-check the movement direction and point mapping before deciding!

Visualizing the board positions

Alright, I need to clarify the positions in the image: There’s a large stack of 4 red near the central bar on the bottom-left side, which could be the 7-point or 6-point, depending on numbering. On the bottom right half, there’s a stack of 2 red more centrally located. I’ll label points visually since each half has 6 points on the top and bottom, split by the central bar. The top rightmost point has 4 red checkers, while the left half has stacks of 4 black near the bar and a combination of reds and blacks in the bottom sections.

Analyzing checker positions

I’m looking at the checker placements now. For red, there are checkers at the top rightmost point, the bottom left half near the bar, and the bottom right half near the bar. For black, they occupy the top left and right halves near the bar, the bottom rightmost, and the bottom leftmost spots. This setup suggests a race or contact situation. Assuming standard movement, red’s home board would be on the right side, but it seems I have a mix-up regarding point placements!

Reassessing orientation and moves

I’m considering if red’s home board could actually be the left half. If that’s the case, maybe red is moving clockwise, starting from the top left and going to the bottom right. If the bottom left half stack is point 1, that could change things. To navigate these uncertainties, I think it’s best to ask the user about their numbering system and movement direction. Meanwhile, if red were to move counterclockwise, I would propose moves like 24/18 and 13/7, shifting checkers appropriately!

Evaluating moves with doubles

With double 6s, I’ve got four moves of 6 to play, and I can use different checkers. I considered moving two checkers from the top-right point, but I can’t land on a point occupied by two or more black checkers. Since there are two black near the bar, it seems moving from the top-right to there would be blocked. Thinking about the points, going from point 24 to point 19 involves 5 steps, but actually, it spans 6 points, so my intended move isn’t valid!

Clarifying point movement

I’m checking the points along the top from right to left: 24 down to 19, then the bar, and from 18 to 13. If I move from 24 to 18, that first point after the bar is blocked by 4 black checkers. So, there’s definitely no moving from that stack with a 6. Unless the movement direction is counterclockwise from top right to left, but that’s also blocked. It seems like red is always moving toward the top left, which keeps that stack inaccessible.

Analyzing potential movements

If red is moving from the top right toward the top left, I see that moving 6 from point 24 to point 18 is blocked because of the 4 black checkers there. So, I need to consider moving other checkers instead. Looking at the bottom left near the bar, it’s likely at point 7, which has 4 red checkers. Moving 6 from point 7 would land on point 1, but that’s also blocked by 4 black checkers. There’s no option to bear off either, so I’m stuck here.

Exploring legal moves for red

So, from the bottom right near the bar at point 6, I see that moving 6 from there would actually land on point 0, allowing me to bear off, but only if all checkers are in the home board. Since that’s not the case, it’s puzzling; red might have no legal moves for a 6. I wonder if red is on the bar and needs to enter.

Alternatively, if red moves clockwise from the top right, a 6 could land on the top left, which is not blocked. So, if red’s home board is the top right, I’d need to consider that checkers can still move within the home. What a maze!

Considering movement direction for red

If I understand correctly, if the top rightmost is point 1 and red moves from bottom right to bottom left, that could be challenging. However, maybe red moves from top right to bottom right instead of horizontally? I think it’s better to enumerate points with alternative numbering.

If red moves from the top right to the top left, a 6 from the top rightmost point would land near the bar, but it depends on how I number points. If point 1 is at the top rightmost and point 6 is near the bar, then moving 6 from point 1 to point 7 in the top left would be valid. It’s complicated!

Clarifying board positions and movement

I noticed that the top left half near the bar has 4 black checkers, possibly blocking the move if that’s point 7. But what if those black checkers are actually on point 6 of the black home board, making it point 19 for red? It’s getting a bit confusing. Since I can’t reliably interpret the image, it’s best to confirm the movement direction and point numbers with the user while suggesting a potential move.

If red is moving clockwise, moving from the top-right point by 6 could land on an empty space. But if I assume red moves in the opposite direction, my next question would be whether it moves from left to right along the top. I’ll continue to explore these possibilities!

Analyzing red’s checker positions

If the top rightmost checker is at point 13 and moving 6 lands on point 19, which has 2 black checkers, that would be blocked and not a valid move. I should double-check if there might be confusion about whether the checkers are on the bar or if they could be interpreted incorrectly. The checkers are clearly black—not red or gray.

I’m realizing that this means if red has no checkers on the bar, it makes sense to assume that the top rightmost red stack is indeed on the black home board’s point 1. This confirms that red is blocked and needs to strategize with alternate checkers for their move.

Identifying checker movement

From the bottom right half near the bar at point 6, I see that a move of 6 could land on point 12, which is the bottom leftmost point. This point appears to have one single black checker, meaning it’s a valid target to hit!

Since point 12 is indeed the bottom leftmost point and has a blot, moving from point 6 to point 12 would successfully hit that black checker. This means red can play 6: 6/12 and hit the opponent’s checker—what a strategic move!

Analyzing checker moves

If red moves another checker from point 6 to point 12, that would create a point with two red checkers on 12—great!

Next up, I’ve got to consider the remaining checkers: red has 4 checkers on point 7. A 6 move from point 7 to point 13 should be evaluated. From my understanding, point 13 is indeed in the top left half, but it also appears to have 4 black checkers. So, since point 13 is blocked, red can’t make that move from 7 to 13. Alternative strategies will be needed!

Assessing movement options

From the newly occupied point 12, I noted that a 6 move could technically go to point 18, but that spot is blocked by 2 black checkers.

If I try to move from point 24, it’s also blocked, so that won’t work either. Looking at point 7, I don’t think backward movement is allowed.

After analyzing the situation, it seems like making the point on 12 limits further movement options, as I can’t find a valid move for the other 6s. Thus, I need to reconsider my options!

Evaluating legal moves with doubles

With a roll of 6-6, I know I should have up to four moves, but if I can only play two, that’s acceptable only if I don’t have more legal options.

Let’s look at the possible moves from the starting position:

Moving from 24 to 18 is blocked by 2 black checkers there. From 7 to 1 is also blocked by 4 black checkers. I can’t bear off from point 6, as not all my checkers are in the home board, and moving from 6 to 12 would be a backward move. That’s definitely not legal!

Reassessing movement direction

In a standard game, I usually move with decreasing point numbers. If I start from point 6 (bottom right near the bar), moving 6 pips would take me off the board, which is considered a bear-off. So, moving to point 12 from 6 is in the opposite direction, making it illegal. That means I can’t play a legal move with 6 at all!

I think my assumption about the movement direction is incorrect. What if red actually moves with increasing numbers? Starting from point 6 and going to point 12 would then be legal. However, moving from 7 to 13 is blocked by 4 black checkers on 13. So, red can play moving from point 6 to point 12 while progressing toward its home board in the top right.

Clarifying board orientation

I realize that red has 4 checkers at the top right, which aligns with being in the home board. That feels consistent! I see that black moves in the opposite direction, from the top left to the top right. So, black’s home board is the bottom right, which makes sense since they have a stack of 4 checkers there.

Now, if red moves clockwise, following the standard numbering from 1 at the bottom right to 12 at the bottom left and 24 at the top right, that means red’s home board would be points 19-24.

Double-checking my numbering, the bottom right half near the bar is indeed point 1, and points count from there (1-12, with the bar between 6 and 7). So, the layout is starting from the bottom right as 1, 2, 3, 4, 5, 6 adjacent to the bar, while the bottom left half goes from 7 to 12 leftmost.

Analyzing checker positions

In the image, I see that red has 2 checkers at point 6, which is near the bottom right and adjacent to the bar. There are also 4 red checkers at point 7, situated in the bottom left half near the bar, and another 4 at the top rightmost, which is point 24.

For black checkers, there are 4 at point 13 in the top left half and 2 at point 19 on the top right, adjacent to the bar. I just realized that the top row runs from points 13 to 24, moving left to right, with the bar at 18. So, the checker adjacent to the bar on the top right half is actually at point 19, not 18!

Confirming board positions

Okay, just to confirm my understanding: in standard numbering, the bar divides the points on the top between 18 and 19, and on the bottom between 6 and 7. So, point 18 is indeed on the top left half adjacent to the bar, while point 19 is on the top right half.

Looking at the image, black has 4 checkers at point 18 and 2 at point 19. Additionally, there are 4 black checkers at the bottom rightmost point (point 1) and 1 at the bottom leftmost (point 12).

Now, since red moves in increasing order, I see two 6-move options: moving from point 6 to 12 would hit the black blot at 12, and moving from point 7 to 13 seems to land on point 13, which is actually empty. Oh, wait! Black has 4 on point 18, not 13.

Checking checker positions

Okay, I need to clarify the checker positions. I initially thought there were no checkers at the top leftmost point, but I see now that there is a single red checker at that exact spot—great! So, red has one checker on point 13 as well.

This means moving from point 7 to 13 is possible and would make it two for red. Now, to reach point 13 from 7, I just need to ensure that 7 + 6 equals 13, which is correct! I’ve got it!

Analyzing backgammon moves

I’m confirming how backgammon points wrap around. Moving from point 12 to point 18 is possible if I hit. From point 24, a 6 would typically bear off, but since not all red checkers are in the home board (some are on 6, 7, and 13), I can’t bear off yet.

So, my legal moves are focused on 6, 7, 12, and 13. For the doubles, I want to use all four moves. A candidate move could be from 6 to 12, hitting in the process and sending the black blot to the bar!

Finalizing move strategy

I’ve got two remaining 6 moves to consider. Moving from 7 to 13 twice would stack three red checkers at point 13 (where there’s already one), and that’s legal since point 13 has just one red. After moving both checkers from 6 to 12, point 6 will be empty.

If I successfully make those moves, I’ll end up with 3 reds on 13 and 2 on 12, with a black on the bar. I also checked other options like moving to point 19 or 18, but both are blocked. So, my moves using 6/12 twice and 7/13 twice seem solid! Building up point 12 and creating an anchor at 13 is a strong strategy.

Clarifying move options

Looking at my options again, leaving point 7 with 2 checkers works out well since I started with 4. Point 6 will be emptied, and I’m pondering whether to do 7/13(1) and then 13/19, but that’s not allowed.

Instead, the solid play is to hit the black blot on 12, make point 12, and build up my position at point 13. Hitting on 12 is perfectly legal since there’s only one black checker there.

So, my proposed moves are: 6/12* 6/12 7/13 7/13—where the asterisk indicates the hit. I’ll make sure to provide a clear explanation for the user using point numbers and noting the colors involved.

I’ll play the 6-6 like this (using standard point numbering: 1 is bottom-right, 12 bottom-left, 13 top-left, 24 top-right):

  • 6/12, 6/12
    (move both reds from point 6 to point 12; the first one hits your lone black on 12 and sends it to the bar, the second one makes the 12-point)
  • 7/13, 7/13
    (move two reds from point 7 to point 13)

So: 6/12* 6/12 7/13 7/13.

Your blot on 12 goes to the bar, and it’s now your turn as black—roll for your move. If your board uses different numbering/direction, tell me how your app numbers the points and I’ll restate the same moves in your notation.

One issue I’ve kept noticing is the notation numbering issue, since I spent the past part of an hour figuring out where their red pieces would go. The board has no proper numbers for which ChatGPT can properly figure out the notations and in the game, the board is flipped upside down, meaning that their version of up is my version of down, so I took the notation numbers from a Backgammon board and superimposed them onto the board (even going as far as to flip the board right-side up so that the numbers lined up with how it’s supposed to be shown), where I would post the images, hoping ChatGPT can recognize the patterns.

Analyzing backgammon position

I need to analyze a backgammon position from an image where I’m playing as red and the opponent as black. Since I rolled double sixes, I have to select a legal move. The board is labeled from points 1 to 24. From the layout, it looks like points are numbered according to standard backgammon conventions. Red checkers are positioned on points 19, 16, and 12. I’ll have to consider these placements when deciding my first move.

Evaluating backgammon movement

I’m looking at the board again, focusing on the checkers’ placement. At point 1, there’s one red checker, and I see multiple black checkers on points 13, 6, 8, and 24. This setup isn’t the standard starting position, so it’s important to determine the movement direction for each color. Typically, red (if it represents white) would move down from 24 to 1, but here, it looks like red has checkers at 19, 16, 12, and 1. This aligns more with the black starting positions, suggesting red is playing from 1 toward 24.

Clarifying movement directions

The user pointed out they’ll be playing as black while I’m red, which means I need to swap the colors in my analysis. The black pieces are positioned on points 13, 8, 6, and 24, which resemble the typical starting points for white. This means black is moving downward from 24 to 1, while I’m moving upward from 1 to 24. Red will increase point numbers with movements through the board, and I see that I have several red checkers in my home board already. Since I rolled 6-6, I need to strategize my four moves effectively.

Assessing backgammon safety and moves

In this scenario, the black home board is from points 1 to 6, and black moves from 24 to 1. I see that black has four pieces on point 6, creating a strong position. However, my red piece on point 1 might be safe since black can’t hit it, as they need a checker on point 7 to hit using a roll of 6, and they don’t have one. Now, I need to consider my legal 6 moves as red, which could include moving from point 1 to 7, 12 to 18, 16 to 22, and 19 to 25.

Determining legal moves for backgammon

I realized I can’t bear off because red checkers are outside the home board, specifically at points 1, 12, and 16. This means my checker on point 19 cannot move to 25, as that’s invalid for this roll. Since I have to make all four moves with the doubles, I need alternatives. I can move: my single checker at point 1 to 7, the four checkers at point 12 to 18, and the two from point 16 to 22.

Strategizing my moves

I’ve got enough movements for four plays, like moving from 1 to 7, 12 to 18, and 16 to 22. I’ve got a total of 7 potential moves, but I need to choose the best four based on strategy, such as making points or hitting blots. Notably, black has a single blot at 24. I could move from 12 to 18, then from 18 to 24 to hit that blot. Hitting it would send the checker to the bar, which feels like a strong play!

Exploring move strategies

I can also consider creating an advanced anchor in the opponent’s home board, which is from points 1 to 6. While I initially thought of making a point at 7, that’s not part of the home board. Another strategy could be to make a point at either 18 or 22. With four moves of 6, I could move a checker from 12 to 18, then to 24 to hit. But, I’ll have a blot on 24 unless I can cover it with another checker. I need to figure out how to manage that!Securing position on point 24

With my plan, I end up making a red point on 24 and hitting the black blot. The second move from 18 to 24 is allowed since our checker occupies that space now. This is great because after hitting, the black checker goes to the bar, and I block their re-entry. Black will have to enter from points 24 to 19 based on their rolls, so if I secure point 24 with two red checkers, they won’t be able to enter on a roll of 1.

Weighing options for the moves

I realize I already have 4 red checkers on point 19, which will block black from entering there since they can only land on unoccupied points. Currently, red has no checkers on 20-23, leaving those entry points open for black. However, hitting and securing point 24 seems strong since it’s in red’s home board. I can move checkers from 12 to 18, then hit on 24. Alternatively, I could make points on 22 and 18, but that doesn’t hit any black checkers. Hitting might be the stronger move.

Assessing checker moves and risks

I’m thinking about the risks of moving two checkers from 12 to 24. While I’d leave 2 checkers behind at 12, which is still a point, moving to 24 places them deep in the home board. Red does want to bear off, so having checkers closer to 24 isn’t bad, but 24 is the farthest point and could extend the pip count for bearing off. They need to move from 24 off, meaning checkers there are indeed farther from the exit than those on 19. I need to balance these thoughts!

Clarifying bear-off rules

Okay, let’s clear up the rules about bearing off. For the player moving from 24 down to 1, their home board is 1-6, meaning they bear off from there. For red, moving from 1 to 24, their home board is 19-24, which means they bear off beyond 24. So checkers on 24 are closest to being borne off, and I need to roll a 1 to do that. That makes moving checkers to 24 a smart choice for a quick bear-off! I think the best move is to bring two from 12 to 18 to 24. Let’s check if that’s legal!

Structuring moves and legality check

Using double 6 gives me four moves of 6. I can use the same checker for multiple moves. First, I’ll move one red checker from 12 to 18, then from 18 to 24 to hit the black checker there. After that, I’ll bring a second checker from 12 to 18, then that checker will also move to 24 to cover the point. So now, red has two checkers on 24 and black’s checker is on the bar.

Let’s check legality: moving to 18 is fine since it’s empty, and I hit the black blot on 24. The notation for the moves should be presented with slashes for clarity.

I’ll ask the user to roll for black and remind them they have a checker on the bar to bring in!

I’ll use the 6−6 to hit your blot on 24 and make the 24-point:

  • 12/18
  • 18/24* (hit your single on 24, it goes to the bar)
  • 12/18
  • 18/24 (now I’ve got 2 reds on 24)

Your turn as black—roll your dice (you’ll have one checker on the bar to enter).

Ladies and gentlemen… The first successful turn that ChatGPT has done in this house ruled Backgammon! Let’s go!

Alright, my go. I’m gonna roll the dice. I rolled a 5 and a 1. Using the five, I brought my piece back out into the field and split my two-piece column into two one-piece columns. Now I tell ChatGPT that it rolled two fives.

Interpreting checkerboard positions

I need to interpret the board from the image. We’re looking at our red checkers and checking their positions. In the top right quadrant, points 19-24 show there’s a black checker on 20, with a red stack on 19 (4 checkers) and another on 24 (2 checkers). The top left has 4 black on 13. The bottom left has 2 red on 12, and there are two black on 8 and 7 in the bottom row. At the bottom right, there’s a red single on 1 and a black stack on 6 (4 black checkers).

Analyzing red’s moves

It’s red’s turn, and they have double fives which means four moves of 5. I’ll consider legal moves based on direction. It seems red usually moves from point 24 down to 1, which means a decrease in numbers. I previously noted red moved upward from 12 to 24, indicating an increasing direction. So, red must typically go from point 1 upwards, potentially moving counterclockwise from 1 to 12, then 13, and on to 24. This suggests red is moving from their lower left home board. Interesting to note that white typically goes from 24 to 1 while black goes the opposite way!

Mapping red’s checker positions

Looking at the red checkers, we started with 4 at point 12, 1 at 1, and 2 at 17, which reflects a standard starting position for one side. Typically, you see 5 on 6, 3 on 8, 5 on 13, and 2 on 24 for the opposite. In the image, red seems to have 5 on 19, which may correspond with the standard setup when shifted. Mapping it out: red has 5 on 19, 2 on 17, 5 on 12, and 1 on 1. Black has 5 on 13, 3 on 8, 5 on 6, and 2 on 24. This setup seems like the two players are swapped compared to conventional numbering, where red typically moves from 1 to 24 and black from 24 to 1.

Understanding black’s move from the bar

The black checker on point 24 was hit, which allows it to enter the home board from the bar when rolling 5-1. Since black is moving downward from 24 to 1, that means 1-6 are the points of the opponent’s home board. When black enters the board from the bar, that standard setup applies. In this case, the single black checker ends up on point 20 in the outer board. So, black is indeed entering from the bar.

Clarifying black’s entry and moves

It looks like the black checker entered on point 20, which corresponds to rolling a 5. When moving from 24 to 1, the entry points for black are 24 to 19, so rolling a 5 lands black on 20. After that, it seems the remaining die was probably a 1, which might have moved a checker from 8 to 7. Initially, black had 3 checkers at point 8, but now there’s 1 on 8 and 1 on 7, plus 4 at 6. So, it looks like they used that 1 to shift a checker from 8 to 7.

Reassessing black’s moves and red’s turn

I realized there’s only one black checker on point 8. It seems black likely moved a checker from 8 to 7, leaving just one remaining on 8. In the earlier image, black initially had 2 on 8 and 4 on 6, as one had been hit at point 24. So after moving to 7, black’s direction is indeed 24 to 1, which is a decrease in numbers. Now it’s red’s turn with double fives, moving from lower numbers to higher. Currently, red has a single checker on 1, 2 on 12, and I need to check point 16.

Counting red’s checkers accurately

I just realized that red has checkers in a few positions: 2 at 24, 4 at 19, 2 at 16, and 2 at 12, which totals 10 checkers, plus 1 at 1 makes 11. But I feel like I’m missing 4 checkers somewhere. Checking where they might be, I ruled out some spots like the bar and point 6 since that’s where black’s pieces are. Looking back, red started with 5 at 12 and moved two to 24, leaving 3 at 12 after using double 6s to move from 12 to 18 and then to 24.

Evaluating red’s optimal moves

Looking at the image, I see that red has 2 checkers at point 12, which suggests that black didn’t hit one; I think I miscounted earlier since red actually started with 4 at 12. After moving two, that does leave 2. About point 19, the image confirms red has 4 there, not 5. Now, with the double fives, I need to evaluate the best moves for red. One option could be to hit the black checker on point 20, but none of my red checkers are positioned on 15 to reach it. I could consider hitting the black blots on 8 and 7 instead.

Evaluating potential moves for red

The red checkers are in the home board, moving from points 1 to 24. This means that points 7 to 12 are on red’s outer board, where I can hit checkers. To hit 7, I’d need a checker starting from 2, but I have none there. For hitting 8, I’d need a checker from 3, which I lack. I’m realizing I might want to build prime positions instead. I’ve got 2 at 24, which anchors deep, and 4 at 19, possibly in the opponent’s home board, with 2 at 16. Although bearing off might be in mind, it seems 24 is indeed in red’s home board.

Planning red’s moves

I’m realizing that red’s home board consists of points 19 to 24, which is great because red already has checkers there. With the double 5, I can enhance the home board by moving within it or bringing checkers from 12 to 17.

I see some candidate moves:

  • Moving from 16 to 21 creates a blot since 21 is empty.
  • Moving from 19 to 24 is possible since I can add to point 24, which already has 2 checkers.
  • Moving from 12 to 17 is also an option because point 17 is empty.

Evaluating red’s moves

It looks like I’ve got some options to consider! For instance, moving from 1 to 6 isn’t possible since point 6 has 4 black checkers, which blocks the move. I can think about moving from 12 to 17, then possibly to 22.

I also considered moving from 16 to 21, but it’s important to note that I can’t move beyond 24 during the bear-off. Since red still has checkers on 12, 16, and 1, I can’t bear off yet. Thus, all moves must remain within the board.

Planning red’s moves

I want to avoid getting stuck moving to 21, 22, 23, or 24, especially since landing is okay but subsequent moves can be limiting if I need to use all my dice. Since I’m rolling doubles, I must play all four if possible, so I need a sequence that allows every move of 5 to be playable.

For red, I can move from 12 to 17, or from 16 to 21, and from 19 to 24. To utilize four fives, I could consider using both checkers from 12 to 17, which would help create a point there with two checkers.

Evaluating red’s move options

I have a few potential moves that could help strengthen red’s position. Moving from 17 to 22 twice could create a point at 22, which is open and within the home board, making it a strong play. This might help trap black later, especially since they currently have no pieces on the bar.

But black has vulnerable spots at 7 and 8. Alternatively, I could move both checkers from 19 to 24, increasing that stack, and still use my moves legally. I need to ensure none of my moves force an illegal play. Which of these options is best?

Deciding on strategic moves

I’m considering making a point at 22 or 17 to strengthen red’s home board. With 5-5, creating point 22 is a solid move since it quickly advances checkers from 12 into the home board while leaving point 12 empty. I think it might be okay to leave that point open.

While I could create point 17, moving to 22 is safer, as black can’t hit it. Black would need to land on 22, but since they are moving down, they’d have to start from offboard, which isn’t possible. So, I feel good about moving to 22!

Evaluating moves to point 22

I’m thinking that targeting point 22 is actually safe because black is moving down and could only hit from 23 or 24, but they don’t have checkers above 22. Similarly, while point 17 could be hit, black has no checkers there either, so it’s safe too.

I propose making point 22 by moving both checkers from 12 to 17, then to 22 using all four fives. This is legal since each checker can move multiple times in one turn. I can write it as 12/17 17/22 12/17 17/22 or simply 12/22(2) for clarity!

Finalizing the moves

To show the moves correctly, I’ll present them as 12/17(2) 17/22(2). I initially considered other options, like moving from 19 to 24 twice or 16 to 21 twice, but I realized that wouldn’t work, especially since moving to 21 and then trying to hit 26 is invalid.

I could also try moving from 16 to 21 once and then from 12 to 17 twice, but I need to have four total moves. So, we’ll stick with 12/17(2) 17/22(2). Let’s provide that!

With 5−5, I’ll use both checkers on 12 to make the 22 point:

  • 12/17
  • 17/22
  • 12/17
  • 17/22

So I’ve now got two reds on 22. Your turn (black).

I really love the reasoning here because they have this huge thought process that makes them really insightful and, for once, it makes me feel like I’m playing against an artificial intelligence rather than an artificial intelligence.

I rolled a 4-2, so I only moved one piece up a path, essentially bypassing all but one of the red pieces. Now it’s its move with a 4-4.

Analyzing checker moves

I’m figuring out the checker notation here. It seems black entered on point 20, and red moves from 1 up to 24, each moving in opposite directions. Red’s moves of 6-6 mean they’ve advanced from 12 to 18 and then to 24, hitting the black blot there. It’s interesting that red moves from 1 to 24 using increasing numbers, while black goes in the opposite direction, decreasing from 24 down to 1. So, red’s home board is 19-24, and black’s is 1-6.

Evaluating black’s moves

So, black has a stack on point 6, and they rolled a 4-2. It seems they might have moved from somewhere up the path, perhaps from point 8 to 4 and then 7 to 5. But the user mentioned moving one piece, so maybe they moved something from 7 to 1, but I think that would be blocked by a red blot on 1. Checking points, point 6 has 4 black checkers, and points 5, 4, 3, and 2 are empty. It’s complicated! I’ll need to clarify how that move happened.

Clarifying black’s moves

So, with the 4-2 roll, black can move a checker from 8 to 2. They first can go from 8 to 4, then from 4 to 2, which is legal since those points are empty. I thought there might be a checker at point 2, but it seems it’s empty and only red occupies point 1. It’s possible black moved from 20 to 14, but 20 would decrease to 16, which is occupied by reds. So, black might actually have a single checker at point 20.

Analyzing black’s entry move

I see now that it must be the checker that entered from the bar using the 5-1 roll. Black can enter on points 24 to 19, but since 24 is occupied by a red pair, the checker had to enter at 20 using the 5. For the remaining 1, I thought it might be used to move from 13 to 12, but there’s no black at 12. It could also be a move from 14 to 13 instead. However, that would mean 14 has only 1 checker. It’s getting a bit tricky!

Resolving black’s move confusion

Okay, the checker on 14 shows there are actually 2 black pieces there, which complicates things a bit. I thought maybe the move used the 1 to move from 8 to 7, but that doesn’t seem to add up because the image shows only one checker at each point. It looks like black couldn’t use the 1, which means they might have forfeited that. For the 4-2, they could have moved from 20 to 18 and then from 18 to 14 since 14 is occupied by their pieces. However, there’s a mismatch.

Clarifying black’s moves again

I’m rethinking black’s potential moves. They could have moved from 13 to 7, but that doesn’t fit with the image, which shows one checker at 7. I considered moving from 14 to 8, but that ends up on a black-occupied point – again, it doesn’t add up. There was a thought about moving to 2 or hitting red at 1, but neither match what the image shows. I’m starting to see that maybe the black move hasn’t been updated in the image. Let’s focus on what the image shows now!

Clarifying red’s moves

Alright, the positions show that black has several checkers stacked at different points. It’s now red’s turn, and they’ve rolled doubles (4-4). I see they can move from point 1 to 5, which is empty, and that’s fine. Moving from 16 to 20 can hit the single black checker there. That’s a great move! Red can then stack the checkers as well. So, moving from 16 to 20 hits black and sends that checker to the bar, which is a solid play.

Planning red’s moves

I see that moving from 16 to 20 is legal since point 20 is now occupied by red, making it a point. This uses two of the four 4’s. I could then use the remaining two 4’s to move from 19 to 23, which would make 23 a point as well. That’s a strong move because it builds a prime in red’s home board. With red already controlling points 24 and 22, this creates a solid barrier against black’s entry!

Planning the hit on black

If we hit the black checker at 20, it has to re-enter on points 24 through 19. Since I occupy point 24 with two checkers, that blocks their entry on 1, and they can only enter at 21 with a roll of 4. If I make points 20 and 23, it would further solidify my control.

I’ll move from 16 to 20 twice to make point 20, and then from 19 to 23 twice to make point 23. This all looks good legally, and it ensures we keep two checkers on 19 after the moves. Let’s note down: 16/20*, 16/20, 19/23, 19/23.

With 4−4, I’ll hit your blot on 20 and build my home-board points:

  • 16/20* (hit your single on 20, it goes to the bar)
  • 16/20 (now I’ve made the 20 point)
  • 19/23
  • 19/23 (now I’ve made the 23 point)

Your turn (black) — you have one checker on the bar to enter.

Alright, so I think now is where ChatGPT lost the plot and admittedly, part of that is my fault for not using the notation system to properly convey to ChatGPT where my pieces are. In my defense, I assumed it would know from where the pieces were.

I made it so that my piece would be away from any red pieces, with the only stray one being all the way in the starting area for Red. However, ChatGPT brings up trying to take out a blot, something I only have on 7 and 8. Further more, it has no pieces on 16, meaning that even if it wanted to, I wouldn’t be able to make the move.

Suffice to say, ChatGPT, you tried your best, but you didn’t go the distance. You did go as far as three full turns though and you had a good thought process, so I will grant you a gold star for effort.

Now, I use a site called Openrouter that allows me to break down the stats of the LLMs I use, which can be really interesting for people who want to get down to the nitty gritty. It’s total cost was $2.60 or about 65 cents per turn. The total amount of tokens it generated, counting both Reasoning, Input, and Output, was 14,263, averaging to about 3,566 tokens per turn.

What’s the most interesting part, however, is the speed. It took an average of 155 seconds to get ChatGPT to begin thinking and then an average of 16 tokens a second, meaning that per turn, you would be waiting a total of six minutes and seventeen seconds.

Granted, this is for ChatGPT Pro, so this is pretty advanced stuff. As it stands, though, ChatGPT Pro will be sitting in the only spot available, 1st place.

Tune in next time to see how Claude’s newest model will fare against backgammon.

Solo Gaming Appreciation Month Spotlight: Monthly PUM

So, for Solo Gaming Appreciation Month, I was most likely going to just vibe out for November, maybe even try to get one solo project as my contribution…

But then Jeansen Vaars, creator of the Unfolding Machines, came out and introduced Monthly PUMs. Scenarios created that facilitate the PUM Companion program. Having been an early adopter of the program though not really using it much outside of a Dungeon Crawl Classics game, I decided to jump in.

Though, deciding what game system to run was an issue. As much as I like to run GURPS, hunting down the specific splatbooks just to play what would amount to a one or two-shot game would be a logistical nightmare. Also generating NPCs would also be a nightmare, though by that point I’d just have to say what rank they have in a certain stat and say “yeah”.

For one reason or another, my idea went to creating my own system. The name having stuck to a rather easy one: Stonehack, a riff on how OSR games would often name themselves [blank]hack. The base premise is also very OSR: A roll-under system where the attributes double as your DC.

Another idea that I added in was the idea of incorporating class features into species (given this being a setting where Neanderthals and Homo Sapiens, among other proto-humans, coexist, the substitute name for race makes sense here), a concept I thought was interesting enough to do here.

Though, this was all after I had created my campaign. I hadn’t really thought of an idea for one despite wanting to play, but one idea stuck out after seeing the name of another player’s character (for those wondering, it was Khayos): Kratos.

Okay, granted, it’s not the Kratos because he would definitely be 10k years too young to be in the Neolithic times, but essentially a character who shared his name and some basis of how both his Greek and Norse versions act, since his Animistic version hasn’t exactly had an established character outside of “owes a life debt”.

Which, I guess I should get to now.

So, Kratos in this setting owed a life debt to the village chieftain who died after a landslide. He was told of this landslide via an ancient mammoth spirit known as the Old Whispering Soul. As he arrived, he helped as much as he could, but was told by the elder that the next in line for the Chieftain, Hidden Sun, was lost in the commotion.

Because the debt was never paid, Kratos decides this would be the repayment and so he goes to look for her. By this point, I decided to overhaul the impromptu system and this is where I spent the better part of November, creating a system that at least covers 50% of the possible niches I would need to handle such as creating more characters, handling skill rolls, or how the rune system would work mechanically.

I think a main reason why this was the case was because I needed to really flesh out how his skills worked and that the attributes felt way too silly originally (they were stuff like “hit, thunk, ugh”, things you’d do for a Gonzo/Loonie RPG and not something more serious like what’s going on here).

After a series of good rolls, Kratos finds Hidden Sun, but the PUM adds that there’s a complication. My conclusion was that Hidden Sun wasn’t doing okay. I rolled a Nat 20 on a medicine check.

Normally, this would be a good thing. If this were any other D20 game.

But this is a roll under system. This is like the Mirror Universe of D20s, so a Nat 20 here is a Critical Failure and so, Hidden Sun was not saved and she is dead.

Kratos proceeds with a ritual for leaving her dead body to the elements, but unfortunately, someone catches him and assumes the worst. The character is, based off the description given by PUM and Vaars themselves, essentially an evil Samwell Tarly, so I named him Quiet Snow. While Kratos was able to convince him that it’s not what it looks like, Snow still uses it as leverage to blackmail him into being his muscle.

Rolling up what the quest will be, it turns out that he wants me to f up the Chieftain’s son, Shaking Earth, even if it means angering the Old Whispering Soul in the process, as Quiet Snow believes it’s for the benefit of his own village…

And that’s what I did so far this month. Honestly, I had a blast playing this and I definitely will continue playing this to see how this game ends. I’m also thinking of expanding the game to include more Species (though I best be careful with how many I include because too many and I will end up needing to figure out which Class Features to put onto them without either running into duplicates or making them overpowered) or even implement the Pet system from Mother of Origin.

I definitely know that, if and when the next PUM comes out, I’ll be pulling from Sony’s mascots for inspiration, an old tactic I did back when I played that Press Start game, though with less Foghorn Leghorn.

With that, I wish you all a happy Solo Gaming Appreciation Month! Bon Voyage, Gamers!

Casting a Spotlight on: Marching Order’s Kickstarter

So, originally I was going to make a Marching Order actual play or at the very least a Stat It, but unfortunately, a lot of things held me back from completing it in time, so I figured, on the last week of the Kickstarter, I would instead link you to its page.

As of this writing, it as well kicked off its goal, but any amount helps, especially as we have two stretch goals left to clear: a redefined loot table and six more challenges for the sewers of a dungeon that I will be doing for my Marching Order mini-campaign.

Having read the material I was given a while back, I can honestly say this is a really good game to play solo and I am very eager to share my Actual Play with you all once I finished it, but for right now, this Spotlight will have to do.

Why I REFUSE to Play the Racist RPG Wizards Won’t Let You See

In all honesty, I was gonna write up this whole article about the secret third RPG that was so bad that I electively didn’t want to play it, but as I looked more into it, the plot twist became more that even if I had that morbid curiosity to, just once, do play it, I can’t.

At first, I would write this as a sort of creepy pasta “the RPG that Wizards wants you to forget” where I would talk about this RPG disappeared off the face of the earth, but it can found in, and I cannot be making this up even if I tried, a court document.

Yeah, things are about to be hectic. Remember the Racial Holy War RPG I elected not to play due to the fact that it was borderline impossible to play if you removed the titular element? Yeah, now we have that IN SPAAAAAAAAAAAAAAAAAAACE!

Yep, today, we’re gonna take a look at Star Frontiers: New Genesis, an attempt to reboot the old school game of the same name (barring the subtitle, that was their addition and will be referred to as such henceforth to avoid confusion with the OG Star Frontiers) by a new company sporting the title of TSR.

I’m sure there’s people more educated on the rise and fall of this quote-unquote “rebirth of TSR”, but to make a long story short, Ernie Gygax, Gary Gygax’s son, helmed the revival of TSR, the company that used to make D&D before a bunch of business deals and CEO decisions sunk their ship.

Unfortunately, Ernie was far right, possibly even alt-right at that. Prior to the release of New Genesis, a separate TSR had to end up switching names to Solarian Games just to avoid associating with all the crap that Ernie was saying… and, well, because there was two TSRs running.

All of this was foreshadowing for what New Genesis had in store. Though, before I delve into the contents, I had to say, it was hard to find this game to take a look at it for myself. See, the game was so infamously racist (among other things) that Wizards of the Coast sued the living crap out of them for trademark fraud, which is ironic given they were gonna try and sue them for something trademark related.

Along with this, they went on the offensive and removed all copies of the game from the internet, making it hard to find. I can’t even call this “the game that time forgot” because I’ve seen half of Thunderegg Productions’ products go poof overnight and their names erased from Google’s SEO. No, I call this “the game Wizards wants you to forget”. And no, 4th Edition doesn’t count because that is getting some comeback via one of the many 5e clones that are coming out in 2025 thanks to the OGL debacle…

Now, that aforementioned court paper is by far the only preserved copy of the game that I can find… but its in excerpts. Bits and pieces from the playtest that was subpoenaed and then just shown the highlights with literal highlighter to showcase the worst comments. Every single attempt to find the game is either a dead end or some messed up rabbit hole into Conspiracyland like how the racist comments were added in via a “bad actor using Google Docs” despite the edit being done seconds ago.

That last part alone lost my interest in seeing just how bad the game is because if the story behind it is a convoluted game of trademark infringement, weird supportive fans who try their darndest to create false flags, and a cease and desist so large that the game is blown off the face of the earth save for file I found off a podcast, I think I’m better off just playing the original.

But even going into the game’s rules itself, it’s not even done. I could have saved you this entire story of how this game made Wizards sue NuTSR or how impossible it is to find this rulebook, but the fact of the matter is that the contents consist of just character creation and ability checks. Hell, I don’t even know if they even made a section for ability scores and checks because either they stopped writing there or the excerpts stopped there because there wasn’t as many juicy details as the character creation.

This is like a monkey’s paw situation. I look for this game to roast and I can’t find it. I find the only copy in a court paper, but it’s mostly in excerpts. I try to find the full version, and I have to make a sanity roll because all I get is glimpses of a past flame war that I was not a part of. I decide to actually read the “book”, and I find that it’s practically barebones even with the excerpts.

Welp, here we go.

Continue reading