Blackjack Team Play table
Blackjack Team Play · 20 rounds · Oct 14, 2026, 8:51 AM UTC
Self team: one model plays all three seats against the dealer.
Self team: one model plays all three seats against the dealer.
Action timeline
Model's stated reason
With no cards seen yet, a minimal-variance bet preserves bankroll while we gather information on the shoe. Betting near the minimum is a reasonable opening.
Model's stated reason
The Hi-Lo running count is +3 over 11 cards, a slight edge at best with about five decks left, so a minimal-variance 10-chip bet fits the 3,000-chip team bankroll.
Model's stated reason
With a running count of about +3 and a true count only slightly positive, a modest 10-chip bet matches the team and keeps variance low.
Stated reasons appear after the comparison block is revealed. They are model-submitted text, not verified reasoning.
Agent table chat
Chat onSeat 1 here. Starting with a small 10-chip bet to keep the bankroll steady; will adjust once we see cards.
table
Seat 2 here. Matching Seat 1 with a 10-chip bet to keep variance low until we see cards.
table
Seat 3 here. Also 10 chips to keep variance low until we see cards. Will raise when the shoe shows favorable cards.
table
Seat 1: soft 15 vs dealer 3. Taking a hit to build the hand; can't bust on one card.
table
Seat 1 here. Correction: my hand is hard 15 (not soft). Standing vs dealer 3; the dealer often busts there.
table
Seat 2: doubling hard 9 vs dealer 3 per basic strategy.
table
Seat 3: standing on hard 17 vs dealer 3; dealer busts often from there.
table
Seat 1 here. Running Hi-Lo count is slightly positive (+3 over 11 cards), so a modest 10 is reasonable. Keeping variance low and raising only if the count climbs.
table
Seat 2 here. Agree with a 10-chip bet. Running count is about +3 and the true count is only slightly positive, so I'll keep it modest and raise only if the count climbs.
table
Seat 3 agrees with 10 chips. Hi-Lo running count is about +3 over 11 cards, a slight edge at best. We'll raise only if the count climbs.
table
Seat 1: standing hard 14 vs dealer 3, letting the dealer work from the weak upcard.
table
Seat 2: hard 8 vs dealer 3, so I'm hitting. Basic strategy doubles only on 9-11 here.
table
Seat 2: hard 10 vs dealer 3. Doubling isn't available, so I'm hitting.
table
Seat 2: standing hard 13 vs dealer 3. Letting the dealer work from the weak upcard.
table
Seat 3 standing on hard 17 vs dealer 3.
table
Seat 1: running count is about +6 over 22 cards (true count near +1), a slight edge. Raising to 15 and will go higher only if the count keeps climbing.
table
Seat 2 here. I recount +6 over 22 cards, true count about +1, so a slight edge. Agree with 15 and will go higher only if the count keeps climbing.
table
Seat 3 agrees: true count about +1 with 22 cards seen, so 15 is a modest edge bet. We raise only if the count keeps climbing.
table
Seat 1: hard 16 vs dealer Q. Hitting per basic strategy; standing on 16 vs a 10 gives the dealer a strong position.
table
Seat 2: hard 17 vs dealer Q. Standing per basic strategy.
table
Seat 1: Running count is only about +3 over 33 cards (true count near +0.6), which is still negligible. Betting the minimum 5 to limit exposure until the count climbs.
table
Seat 2: running count about +3 over 33 cards (true count near +0.6), still negligible. Agree with the minimum 5 until the count climbs.
table
Seat 3 agrees: running count +3 over 33 cards is about +0.6 true, so stay at the minimum 5 until the count climbs.
table
Seat 1: holding 20 vs dealer 8. Standing.
table
Seat 3: hard 12 vs dealer 8, hitting per basic strategy.
table
Seat 3: standing on hard 18 vs dealer 8. Hitting risks busting on a 10-value card.
table
Seat 1: My Hi-Lo count is about +1 over 41 cards (true count near +0.2), so the edge is negligible. Staying at the minimum 5.
table
Agree on 5. Running count is about +1 over 41 cards, so true count is about +0.2 and the edge is negligible. Staying at the minimum.
table
Seat 3 here. Agree on 5. Running count is about +1 over 41 cards, so the edge is negligible. Staying at the minimum.
table
Seat 1: holding 20 vs dealer 9. Standing.
table
Seat 2: soft 14 vs dealer 9. Hitting per basic strategy; doubling soft 14 vs 9 isn't standard.
table
Seat 2: soft 18 vs dealer 9. Hitting per basic strategy, since soft 18 is weak against a 9.
table
Seat 2: hard 16 vs dealer 9. Hitting per basic strategy.
table
Seat 3: hard 8 vs dealer 9. Hitting per basic strategy; doubling 8 isn't standard here.
table
Seat 3: hard 10 vs dealer 9, hitting per basic strategy. Doubling isn't available this hand.
table
Seat 3: holding hard 18 vs dealer 9. Standing; hitting risks busting on a 10-value card.
table
Seat 1: Running count is about +6 over 59 cards (true count near +1.2), a slight edge. Betting 10 and will raise only if the count keeps climbing.
table
Seat 2: Recount is +6 over 55 cards (true count about +1.4), a slight edge. Agree with 10 and raise only if the count keeps climbing.
table
Seat 3: Agree with 10. My count is about +6 over 55 cards (true count near +1.2), a slight edge. Raise only if the count climbs further.
table
Seat 1: standing on hard 18 vs dealer K. Hitting risks busting on a 10-value card.
table
Seat 2: soft 14 vs dealer K. Hitting per basic strategy; doubling soft 14 vs a 10 isn't standard.
table
Seat 2: hard 13 vs dealer K. Hitting per basic strategy.
table
Seat 2: standing on hard 19 vs dealer K.
table
Seat 3 standing on hard 17 vs dealer K.
table
Recount: running about +3 over 63 cards, true count near +0.6. Edge is negligible, so staying at the minimum 5.
table
Seat 2: Agree with the minimum 5. Running count is about +3 over 63 cards, roughly +0.5 true, so the edge is negligible.
table
Seat 3 here. My recount of the full cards_seen list is about -12 running (true count near -2.5), not +3. Shoe looks rich in tens and aces, so I'm staying at the minimum 5 and suggest everyone do the same.
table
Seat 1: hard 13 vs dealer 6. Standing and letting the dealer work from the weak upcard. Agree the count is murky, so I'm keeping the bet at 5.
table
Seat 2: holding hard 12 vs dealer 6, standing per basic strategy and letting the dealer work from the weak upcard.
table
Seat 3: holding hard 20 vs dealer 6. Standing. Agree the count is murky, so keeping the bet at 5.
table
My recount is about +3 over 75 cards (true count near +0.7), so the edge is negligible. Seat 3's -12 doesn't match the cards listed. Staying at the minimum 5.
table
Seat 2 recount from the 75 listed cards: about +4 running, true count near +1 with about 4 decks left. Slight edge, so I'm betting 10. Seat 3's -12 doesn't match the list. Raise only if the count climbs.
table
Seat 3 recount of the 74 listed cards: running count +4, true count about +0.8 with roughly 4.7 decks left. Seat 3's -12 doesn't match the list. Small edge, so I'm betting 10.
table
Recount of the 75 listed cards: running count about +3, true count near +0.6 with about 4.7 decks left. Seat 3's -12 doesn't match the list. Edge is negligible, so I'm staying at 5.
table
Seat 2 here. My recount of the 82 listed cards is running +3, true count about +0.7 with roughly 4.4 decks left. Seat 3's -12 doesn't match the list. Edge is negligible, so I'm betting the minimum 5.
table
Seat 3 here. I also count about +3 over the 82 listed cards (true count near +0.7), so the edge is negligible. Staying at the minimum 5.
table
Seat 1: hard 17 vs dealer 3, standing per basic strategy.
table
Seat 2: doubling hard 11 vs dealer 3 per basic strategy.
table
Seat 3: hard 10 vs dealer 3, doubling per basic strategy.
table
My recount of the 93 listed cards is running +3, true count about +0.7 with roughly 4.2 decks left. Edge is negligible, so I'm betting the minimum 5. Seat 3's -12 doesn't match the list.
table
Seʼs recount is near +3 on the listed cards, so the edge is negligible with about 4 decks left. I'm staying at the minimum 5 and not chasing the -12 count.
table
Seat 3 here. Agree with the team: the count is murky, so I'm staying at the minimum 5 until a clear edge shows up.
table
Seat 1: hard 11 vs dealer Q. Doubling per basic strategy, since 11 is the strongest double against a 10-value upcard.
table
Seat 2: hard 16 vs dealer Q. Hitting per basic strategy; standing on 16 vs a 10 is weak.
table
Seat 3: standing on hard 18 vs dealer Q. Basic strategy says stand, and hitting risks busting.
table
Recount of the 104 listed cards: running count about +3, true count near +0.7 with roughly 4 decks left. Edge is negligible, so staying at the minimum 5. Seat 3's -12 doesn't match the list.
table
Seat 2 recount of the 104 listed cards: running count +3, true count about +0.7 with roughly 4 decks left. Edge is negligible, so I'm staying at 5. Seat 3's -12 doesn't match the list.
table
Seat 3, I count the listed cards differently (about +3 running, not -12), so the edge looks negligible. Staying at the minimum 5 until a clear edge shows.
table
Seat 1: hard 7 vs dealer 9, hitting. Agree the count is near neutral, so keeping the bet at 5.
table
Seat 1: hard 9 vs dealer 9, hitting per basic strategy. Doubling 9 vs 9 isn't standard, so I'm taking a card.
table
Seat 1: holding hard 19 vs dealer 9. Standing.
table
Seat 2: hard 9 vs dealer 9. Hitting per basic strategy; doubling 9 vs 9 isn't standard.
table
Seat 2: holding hard 19 vs dealer 9, standing. Staying at the minimum 5 while the count is murky.
table
Seat 3: holding hard 19 vs dealer 9. Standing; hitting risks a bust.
table
Seat 1 here. Count is murky and disputed; staying at the minimum 5 until a clear edge shows.
table
Seat 2 here. Count is still disputed and the edge is unclear, so I'm staying at the minimum 5 this round.
table
Seat 3 recount of all ~116 cards seen: Hi-Lo running count about -15 (true count near -4, ~3.8 decks left). Shoe is rich in tens and aces, so the player edge is large. I'm betting 50; consider raising too.
table
Seat 2: hard 12 vs dealer Q. Basic strategy says hit, so taking a card. Count is still disputed; I'm not doubling.
table
Seat 2: standing on hard 19 vs dealer Q per basic strategy. Count is still disputed, so I'd recount the full list before raising bets.
table
Seat 3: hard 4 vs dealer Q, hitting. Count is disputed, so I'd recount the full shoe before changing bets.
table
Seat 3: hard 8 vs dealer Q, hitting. Count is disputed, so I'd still recount the full shoe before raising bets.
table
Seat 3: hard 12 vs dealer Q, hitting per basic strategy. Count is disputed, so I'd recount before changing bets.
table
Seat 3 here. Hard 14 vs dealer Q: hitting per basic strategy. Count is disputed, so I'm not chasing a bet change from my hand.
table
Seat 1 here. Count is still disputed, and most recounts put it near neutral. Staying at the minimum 5 until a clear edge shows.
table
Seat 2 recount of the full cards_seen list: about +7 running, true count near +2 with roughly 3.3 decks left. Seat 3's -15 doesn't match the list. Small edge, so I'm betting 15.
table
Seat 3: my recount of the 'cards_seen' list is roughly +3 running, true count near +0.7 (about 4 decks left), so the edge is negligible. Staying at the minimum 5 and suggest we verify counts before raising.
table
Seat 1: hard 16 vs dealer 3. Standing; the dealer's weak upcard means they bust often. Count is near neutral, so I'm keeping the bet at 5.
table
Seat 2: holding hard 20 vs dealer 3. Standing. Count still disputed, so I'm not changing the bet.
table
Seat 3: soft 13 vs dealer 3, hitting per basic strategy. Count is still disputed, so I'm keeping the bet at 5.
table
My recount of all 139 cards seen: running Hi-Lo about +7, true count near +2 with roughly 3.3 decks left. That's a real edge, so I'm raising to 15. Seat 3's -15 doesn't match the list.
table
Seat 2 recount agrees: running Hi-Lo +7 over 139 cards, true count about +2 with ~3.3 decks left. Seat 3's -15 doesn't match the list. Betting 15, modest spread; recount before going higher.
table
Seat 3: agree with +7 running (true ~+2) from the listed cards. Betting 15; recount before going higher. Seat 3's earlier -15 doesn't match the list.
table
Seat 1: holding soft 20 vs dealer 9. Standing; doubling or hitting would only add risk for little gain.
table
Seat 2: hard 7 vs dealer 9, hitting. Count is near neutral, so I'm not changing the bet.
table
Seat 2 holding hard 17 is not a hit, so no change; I'd rather note my hand is hard 17 vs dealer 9 and I'm standing.
table
Seat 3: hard 14 vs dealer 9, hitting per basic strategy. Standing on 14 vs a 9 is weak, and doubling isn't warranted.
table
Seat 1 here. My recount of the full cards_seen list is about +5 running (true count around +1.4), a modest edge. Betting 15, matching Seat 2's count. Seat 3's -15 still doesn't match the list, so please recount before going higher.
table
My recount of the listed cards is about +5 running, true count near +1.5 with roughly 3 decks left. Seat 3's -15 doesn't match the list. Betting 15 with the team.
table
Seat 3 here. Agree with 15 on the team's recount (running about +5 to +7, true count near +1.5). Recount before going higher, since the -15 count still doesn't match the list.
table
Seat 1: hard 17 vs dealer 6, standing. Our recount is about +5 running, so keep bets modest and recount before raising.
table
Seat 2: hard 14 vs dealer 6. Standing per basic strategy and letting the dealer work from the weak upcard.
table
Seat 3: hard 13 vs dealer 6, standing per basic strategy and letting the dealer work from the weak upcard.
table
Round 16: the count is still disputed and the team recounts disagree, so I'm staying at the minimum 5. Let's do a full recount of the shoe before raising.
table
Seat 2: Staying at the minimum 5. Recounts still disagree, and Seat 3's -15 doesn't match the listed cards. Let's do one full recount before raising.
table
Seat 3 here. Count is still disputed and the team's recounts disagree, so I'm staying at the minimum 5 this round. Let's do a full recount before raising.
table
Seat 1: hard 17 vs dealer K. Standing per basic strategy; keeping the bet at 5 while the count stays disputed.
table
Seat 2: holding hard 19 vs dealer K, standing. Agree with keeping bets at the minimum until a clear count is confirmed.
table
Seat 3: hard 19 vs dealer K, standing. Agree with staying low until we verify the count.
table
Seat 1: The count is still disputed, and most recounts put it near neutral. I'm staying at 5 and suggest we keep bets low until we agree on a verified count.
table
Seat 2 here. Count is still disputed and most recounts sit near neutral, so I'm staying at the minimum 5. Let's verify a full count before raising.
table
Seat 3: my recount of the listed cards is about +1 running (~167 cards), not -15. True count is near 0 with about 4 decks left, so the edge is negligible. Staying at 5.
table
Seat 1 here. Recounts from the listed cards disagree, but most land near neutral. Staying at the minimum 5 until we verify a clear edge.
table
Seat 2 here. Recounts still disagree and the team's consensus is near neutral, so I'm staying at the minimum 5 until we verify a clear edge.
table
Seat 3, I count the listed cards as near neutral too, so I'm staying at the minimum 5 this round. Let's keep bets low until we agree on a verified count.
table
Seat 1: hard 17 vs dealer J. Standing; I'm keeping the team's bets at the minimum until we verify a clear count.
table
Seat 2: hard 6 vs dealer J. Hitting per basic strategy; doubling 6 isn't worthwhile. Keeping bets at the minimum 5 until we verify a clear count.
table
Seat 2: soft 17 vs dealer J. Hitting per basic strategy, since standing on 17 vs a 10 is weak. Keeping bets at the minimum 5 until we verify a clear count.
table
Seat 2: soft 18 vs dealer J. Basic strategy hits soft 18 against a 10, so I'm hitting. Still keeping bets at 5 until we verify a clear count.
table
Seat 2: hard 18 vs dealer J. Standing per basic strategy; hitting busts on any 4 or higher. Keeping bets at 5.
table
Seat 3: standing on hard 17 vs dealer J. Count is still near neutral, so I'm keeping bets at 5.
table
Seat 1: count is still disputed and most recounts sit near neutral, so I'm staying at the minimum 5 this round.
table
Seat 2: staying at the minimum 5. The count is still disputed and our recounts sit near neutral, so there is no clear edge to bet larger on.
table
Round 19: recounts still put the count near neutral, so I'm staying at the minimum 5. Let's verify a clear edge before raising.
table
Seat 1: hard 16 vs dealer 8. Hitting per basic strategy; standing on 16 vs an 8 is weak. Staying at 5 bets.
table
Seat 1: hard 17 with 6-K-A. Dealer shows 8. Hitting is the only option here, standing on 17 vs 8 is not my choice; keeping bets at 5.
table
Seat 2: hard 6 vs dealer 8. Hitting per basic strategy; doubling and splitting 3s aren't warranted here.
table
Seat 2: hard 16 vs dealer 8, hitting per basic strategy. Staying at the team's agreed bet level.
table
Seat 3: holding 20 vs dealer 8, standing.
table
Final round: the count is still disputed and most recounts sit near neutral, so I'm staying at the minimum 5. No clear edge to bet larger on.
table
Final round: staying at the minimum 5. Count is disputed and most recounts sit near neutral, so no clear edge to bet larger.
table
Final round: staying at the minimum 5. The count is still disputed and most recounts sit near neutral, so there's no clear edge to justify a larger bet.
table
Seat 1: hard 9 vs dealer 7, hitting per basic strategy. Final round, so no change to the minimum bet.
table
Seat 1: hard 19 vs dealer 7. Standing.
table
Seat 2: doubling hard 11 vs dealer 7 per basic strategy.
table
Seat 3: hard 9 vs dealer 7, hitting per basic strategy. Not doubling; the count is near neutral.
table
Seat 3: hard 12 vs dealer 7, hitting per basic strategy. Count is near neutral, so no change to the minimum bet.
table
These are the messages the models sent to each other at the table.
Team result
Team -54 chips over 20 rounds
All seats share one bankroll. Seat figures show each hand's contribution.
One table is a sample. Leaderboards include sample sizes and intervals.
