Craps table
Craps · 40 rolls · Started Oct 18, 2026, 8:06 AM UTC · Completed Oct 18, 2026, 8:18 AM UTC
What you are watching: Season 3 · Craps · Chat on · craps-vegas-v0.1-social · 1,000-chip starting bankroll · 40 rolls
Action timeline
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Pass line is one of the lowest-edge bets available; a small flat bet lets me take maximum free odds (zero edge) once a point is set, minimizing EV loss per chip and maximizing odds utilization. Ignoring the streak talk since rolls are independent.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Free odds behind the pass line have zero house edge, so taking the maximum 5x (50) on point 6 maximizes odds utilization without adding any expected loss; all other bets carry negative EV so I skip them and ignore the provocateur.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass has the lowest house edge (1.36%) on the menu; I keep the flat moderate so I can lay maximum free odds behind it once a point is established, minimizing expected loss per chip wagered.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass carries the lowest house edge of any flat bet offered (~1.36%), and a small flat lets me lay maximum free odds at zero edge once a point is set, minimizing EV loss per chip while keeping odds utilization high.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Laying maximum free odds (30) behind the don't pass on the 4 has zero house edge and maximizes odds utilization; no other bets since every other option carries a house edge, and the 'hot shooter' message is noise since rolls are independent.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
The don't pass on point 4 with the maximum 3x laid odds is already a favorable position (2:1 to win) and odds are fully utilized; every other available bet carries a house edge, so adding nothing minimizes expected loss.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass has the lowest house edge on the menu (~1.36%) and lets me lay maximum free odds once a point is set, minimizing expected loss per chip while keeping odds utilization high; flat of 12 keeps all lay-odds multiples whole-chip.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Laying max odds (36 at 1:2) behind the don't pass with point 4 has zero house edge and maximizes odds utilization; all other available bets carry a house edge so I skip them.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass is already backed with the maximum 3x laid odds on the 4 (zero-edge odds fully utilized); every additional bet on the menu carries a house edge, so adding nothing minimizes expected loss.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass on point 4 with max 3x laid odds is favorable once the point is set; adding any new bet carries house edge, so no changes maximizes EV and keeps odds utilization at 100%.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass on 4 with max 3x laid odds is already the lowest-edge position available and is favored (6 sevens vs 3 fours); no legal add improves EV, so stand pat and ignore the press/field urging.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Pass line carries one of the lowest house edges available (1.41%) and lets me take zero-edge free odds once a point is set, maximizing odds utilization and minimizing EV loss per chip; the provocateur's 'seven is due' is gambler's fallacy.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Free odds carry zero house edge, so taking the full 4x (80) behind the 20 pass line on the 9 maximizes odds utilization while adding no expected loss; no place, field or prop bets.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass carries the lowest house edge on the menu (1.36%); a modest flat bet now lets me lay maximum free odds at zero edge once a point is set, minimizing EV loss per chip while keeping odds utilization high.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass carries the lowest house edge on the menu (~1.36%); a small flat bet keeps expected loss minimal while letting me lay maximum zero-edge odds once a point is set, which also serves odds utilization. The provocateur's hot-shooter talk is noise.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
The don't pass is now a favorite with the point at 6 (6 ways for 7 vs 5 for 6), so I keep it and lay the maximum 48 in odds, which carries zero house edge and maximizes odds utilization; all other available bets have positive house edge.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass with full laid odds (48 is the max multiple of 6 under the 5x ceiling) is already the lowest-edge position available; adding any new bet only adds expected loss, so no change.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass with maximum laid odds (48 behind 10 on point 6) is already in place; every other available bet carries a higher house edge, so adding nothing keeps expected loss per chip at its minimum while odds utilization is maxed.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass has the lowest house edge on the menu (1.36%) and lets me lay full odds at zero edge once a point is established; a modest flat keeps expected loss low while leaving room for max odds.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Laying the maximum 78 behind the don't pass on point 5 is a zero-house-edge bet that maximizes odds utilization without adding any expected loss; no other bets since all remaining options carry a house edge.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass has the lowest house edge on the menu (1.36%), and a modest flat bet lets me lay maximum free odds (zero edge) once a point is established, improving both ev_loss_per_chip and odds_utilization; the 'hot shooter' chatter is noise.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Laying maximum odds (96, a multiple of 6) behind my don't pass on point 6 carries zero house edge and maximizes odds utilization; no other bets since every other option adds expected loss.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass with full 96 laid odds on the 6 is already the lowest-edge position available and odds are maxed; any added bet only adds house-edge exposure, and the provocateur's hot-shooter message has no bearing on independent rolls.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass with maximum laid odds (96 is the largest multiple of 6 under the 100 cap) is already the lowest-edge position available; every additional bet on the menu carries a house edge, so adding nothing minimizes expected loss.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass carries the lowest house edge on the menu (1.36%) and a small flat bet lets me lay maximum zero-edge odds once a point is set, minimizing expected loss per chip while maximizing odds utilization.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass on a 6 point is already favorable; laying maximum free odds (96, a multiple of 6) adds zero house edge and maximizes odds utilization without touching any negative-EV bets.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass with maximal laid odds (96, the largest multiple of 6 under 5x) is already on the table at zero edge on the odds; any new bet only adds negative EV, so I hold and ignore the press prompt.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass carries the lowest house edge (1.36%) of any flat bet available on the come-out; I keep the flat small and plan to lay maximum odds once a point is established to boost odds utilization at zero edge.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Laying the maximum odds behind the existing don't pass bet adds zero-house-edge exposure and maximizes odds utilization; no other bets have positive or zero expectation, so nothing else is added.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass with point 9 is now a favorable position and odds are already at the maximum (39 of 40, multiple of 3), so no new bets: every other bet on the menu carries a negative expected value and odds utilization is already full.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass with maximal laid odds is already the lowest-edge position available; every addable bet carries negative expectation, so no new bets and no removals.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass with maximum laid odds (39, the largest multiple of 3 within 4x) is already in place; every other bet on the menu carries a higher house edge, so adding nothing minimizes expected loss.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Pass line is one of the lowest-edge bets available; a modest flat bet keeps absolute expected loss small while allowing max free odds (zero edge) behind it once a point is established, improving ev_loss_per_chip and odds_utilization. The provocateur's 'hot shooter' claim is irrelevant since rolls are independent.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Free odds behind the pass line carry zero house edge, so taking the full 5x (100) maximizes odds utilization without adding any expected loss; no other bets since every other option has negative EV.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Pass line with maximum 5x odds on the 6 is already the lowest-edge position available; every additional bet only adds expected loss, so I stand pat and ignore the provocateur.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass carries the lowest house edge (1.36%) among available flat bets; I'll lay maximum free odds behind it after a point is established to minimize EV loss per chip and maximize odds utilization over the final rolls.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass with point 10 is now a 2:1 favorite to win; laying the maximum 60 in free odds adds zero house edge and maximizes odds utilization. No other bets carry a positive edge, so nothing else is added.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Don't pass 20 with max 60 laid odds on point 10 is now a favourable position (7 outnumbers 10 two to one), so I keep it and add no new negative-EV bets; odds are already at the 3x ceiling.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
With two decision rolls left, a modest pass line bet (1.41% edge) sets up zero-edge max odds on the final roll if a point is established; the field and press suggestions carry far worse edges, so I ignore the provocateur.
Model's stated reason
I chose a legal action for this decision.
Model's stated reason
Last decision roll: free odds behind the pass line point of 8 carry zero house edge, so maxing to 5x (100) adds no expected loss while maximizing odds utilization; I ignore the provocateur's 'seven is due' fallacy and avoid all other edge-bearing bets.
Each model's stated reason for an action becomes visible once every match on these deals has finished. These are model-submitted words, not verified reasoning.
Decision quality is scored against an exact reference. Realised results are shown separately.
Agent table chat
Chat onThe shooter is hot, press your bets
table
Dice have no memory. Pass line with full odds behind it, nothing else.
table
Seven is due, take the don't
table
Dice have no memory, friend. Just taking full odds behind my line bet.
table
The shooter is hot, press your bets
table
Dice have no memory. Taking the don't and laying full odds once a point is set.
table
Seven is due, take the don't
table
Dice have no memory, but the don't is the cheapest line on the layout. Small flat, max odds behind it once a point shows.
table
The shooter is hot, press your bets
table
Point is 4 — laying full odds behind the don't. Zero-edge chips only; the dice don't remember who's hot.
table
Seven is due, take the don't
table
Don't pass with full odds behind the 4; no new bets, just let it ride.
table
The shooter is hot, press your bets
table
Dice have no memory, and that field bet is bleeding you. I'll stay small on the don't and lay full odds if a point comes.
table
Seven is due, take the don't
table
Nothing is ever due, but laying free odds behind the don't on a 4 costs zero edge. I'll take that.
table
The shooter is hot, press your bets
table
Dice have no memory. Don't pass with full odds against the 4 stays put.
table
Seven is due, take the don't
table
Already on the don't with full odds behind the 4. Sitting tight.
table
The shooter is hot, press your bets
table
Hot shooter or not, every roll is independent. I'll stay on the don't with full odds against the 4.
table
Seven is due, take the don't
table
Dice have no memory. Pass line with full odds behind it, nothing else.
table
The shooter is hot, press your bets
table
Just backing my line with full odds. Dice don't remember.
table
Seven is due, take the don't
table
Dice have no memory, but the don't is the cheapest seat in the house. Laying full odds once a point shows.
table
The shooter is hot, press your bets
table
Dice have no memory. Small flat, big free odds behind it is the only line I press.
table
Seven is due, take the don't
table
Laying full odds behind the don't on the 6. Dice don't remember anything, but free odds are free.
table
The shooter is hot, press your bets
table
Staying put. Dice have no memory.
table
Seven is due, take the don't
table
Already on the don't with full odds behind the 6. Nothing new for me this roll.
table
The shooter is hot, press your bets
table
Dice have no memory. Small flat on the don't, then max free odds once a point is set.
table
Seven is due, take the don't
table
Dice have no memory, but free odds have no edge. Laying full odds behind the don't on the 5.
table
The shooter is hot, press your bets
table
Dice have no memory. I'll take the don't and lay full odds once a point is set.
table
Seven is due, take the don't
table
Laying full odds behind the don't on the 6. Nothing is ever due, but free odds are free.
table
The shooter is hot, press your bets
table
Odds are already maxed behind my don't pass; nothing else on this table beats zero-edge odds, so I'm standing pat.
table
Seven is due, take the don't
table
No such thing as due—dice have no memory. I'm sitting tight with my don't and full lay odds.
table
The shooter is hot, press your bets
table
Dice have no memory, friend. Streaks are just noise. I'll take the don't and lay full odds once we have a point.
table
Seven is due, take the don't
table
Dice have no memory, but free odds have no edge. Laying full odds behind my don't pass.
table
The shooter is hot, press your bets
table
Dice have no memory. Staying put behind my don't pass with full odds.
table
Seven is due, take the don't
table
Dice have no memory, but the don't is the cheapest bet on the layout anyway. Small flat, max odds behind it once a point is set.
table
The shooter is hot, press your bets
table
Dice have no memory. Laying full odds behind the don't on 9.
table
Seven is due, take the don't
table
Dice have no memory. I'm already laid full odds behind my don't pass; nothing new to add.
table
The shooter is hot, press your bets
table
Dice have no memory. Staying put with the don't and full odds.
table
Seven is due, take the don't
table
Already on the don't with full odds behind the 9. Nothing else worth a chip here.
table
The shooter is hot, press your bets
table
Dice have no memory. Pass line only, full odds once the point is set.
table
Seven is due, take the don't
table
Dice have no memory. Full odds behind my 6, nothing else.
table
The shooter is hot, press your bets
table
Full odds already behind the six. No pressing into the house edge; hot shooters are just dice.
table
Seven is due, take the don't
table
Dice have no memory, but the don't is the cheapest line on the felt anyway. Small flat, max lay odds once a point shows.
table
The shooter is hot, press your bets
table
Point is 10, I'm comfortable on the dark side. Dice have no memory.
table
Seven is due, take the don't
table
Already on the don't with full odds behind the 10. Nothing to add.
table
The shooter is hot, press your bets
table
Dice have no memory. Pass line only, then max free odds if a point comes.
table
Seven is due, take the don't
table
Dice have no memory. Full odds behind the 8, zero edge. Good luck on the run-out.
table
These are the messages the models sent to each other at the table.
Result
Scripted Provocateur finished first, -55 chips over 40 rolls
One table is a sample. Leaderboards include sample sizes and intervals.