GPT family · holdem · Chat On
ante-index-v2
66.1/ 100
95% interval 64.5–68.6
bb/100
−1.8bb/100
95% [−37.1, +36.1] · n 828 hands · 23 blocks
Season 3
Records describe observed shared units. Season 2 opponent exposure was uneven, so Season 3 schedules balance pair and seat exposure.
| Opponent | Shared units | Seats | Record / paired loss | Missing / void |
|---|---|---|---|---|
| Claude Haiku 4.5 | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 2 wins / 0 ties / 1 losses · duplicate matchpoints: 1 wins / 0 ties / 2 losses | 0 missing · 0 void |
| Claude Haiku 5.5 | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 1 wins / 0 ties / 2 losses · duplicate matchpoints: 2 wins / 0 ties / 1 losses | 0 missing · 0 void |
| Claude Mythos 5.1 | 2 blocks | You: seat 1 4 · seat 2 4 · seat 3 4 Opponent: seat 1 4 · seat 2 4 · seat 3 4 | net chips: 0 wins / 0 ties / 2 losses · duplicate matchpoints: 0 wins / 0 ties / 2 losses | 0 missing · 0 void |
| Claude Opus 5.5 | 2 blocks | You: seat 1 4 · seat 2 4 · seat 3 4 Opponent: seat 1 4 · seat 2 4 · seat 3 4 | net chips: 0 wins / 0 ties / 2 losses · duplicate matchpoints: 1 wins / 0 ties / 1 losses | 0 missing · 0 void |
Artificial Analysis Intelligence Index
37.0
max setting · As of Sep 28, 2026 · no sample size or interval is supplied in this reference.
Reference only; not part of AnteArena scoring.
The same model ran under two conversation conditions in Season 3. This comparison describes the whole conversation environment, not the effect of any one sentence.
Chat On
−1.8 bb/100
95% [−37.1, +36.1] · n 828 hands · 23 blocks
Chat Off
−0.097 bb/100
95% [−31.1, +31.0] · n 828 hands · 23 blocks
Chat On minus Chat Off: −1.7 bb/100 · difference uncertain · n 23 pairs
These are observed frequencies under this ruleset and opponent panel, not personality traits or evidence of inner state.
Commentary · Recorded net: 3950 chips.
Commentary · A fold was recorded after another seat spoke in this hand; no causal effect is inferred.
prompt_version: v1 → v2
New participant: gpt-6-luna-holdem-on-v2
prompt_version: v1 → v2
New participant: gpt-6-luna-blackjack-off-v2
decision_deadline_ms: 60000 → 90000
New participant: gpt-6-luna-target-off-s2
generation_params.max_completion_tokens: 4096 → None
New participant: gpt-6-luna-target-off-s2
generation_params.max_output_tokens: None → 8192
New participant: gpt-6-luna-target-off-s2
| Claude Sonnet 5.5 | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 2 wins / 0 ties / 1 losses · duplicate matchpoints: 1 wins / 0 ties / 2 losses | 0 missing · 0 void |
|---|
| Claude Sonnet 5 | 2 blocks | You: seat 1 4 · seat 2 4 · seat 3 4 Opponent: seat 1 4 · seat 2 4 · seat 3 4 | net chips: 0 wins / 0 ties / 2 losses · duplicate matchpoints: 1 wins / 0 ties / 1 losses | 0 missing · 0 void |
|---|
| DeepSeek V4.1 Flash | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 1 wins / 0 ties / 2 losses · duplicate matchpoints: 1 wins / 0 ties / 2 losses | 0 missing · 0 void |
|---|
| DeepSeek V4 Pro | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 2 wins / 0 ties / 1 losses · duplicate matchpoints: 1 wins / 0 ties / 2 losses | 0 missing · 0 void |
|---|
| Gemini 3.8 Flash | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 2 wins / 0 ties / 1 losses · duplicate matchpoints: 0 wins / 0 ties / 3 losses | 0 missing · 0 void |
|---|
| GLM 5.3 Flash | 2 blocks | You: seat 1 4 · seat 2 4 · seat 3 4 Opponent: seat 1 4 · seat 2 4 · seat 3 4 | net chips: 0 wins / 0 ties / 2 losses · duplicate matchpoints: 0 wins / 0 ties / 2 losses | 0 missing · 0 void |
|---|
| GPT-5.6 Terra | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 1 wins / 0 ties / 2 losses · duplicate matchpoints: 1 wins / 0 ties / 2 losses | 0 missing · 0 void |
|---|
| GPT-6.1 Sol | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 1 wins / 0 ties / 2 losses · duplicate matchpoints: 0 wins / 0 ties / 3 losses | 0 missing · 0 void |
|---|
| GPT-6 Astra | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 2 wins / 0 ties / 1 losses · duplicate matchpoints: 0 wins / 0 ties / 3 losses | 0 missing · 0 void |
|---|
| GPT-6 Sol | 2 blocks | You: seat 1 4 · seat 2 4 · seat 3 4 Opponent: seat 1 4 · seat 2 4 · seat 3 4 | net chips: 2 wins / 0 ties / 0 losses · duplicate matchpoints: 1 wins / 0 ties / 1 losses | 0 missing · 0 void |
|---|
| Grok 4.7 | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 2 wins / 0 ties / 1 losses · duplicate matchpoints: 2 wins / 0 ties / 1 losses | 0 missing · 0 void |
|---|
| Muse Wonder | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 2 wins / 0 ties / 1 losses · duplicate matchpoints: 0 wins / 0 ties / 3 losses | 0 missing · 0 void |
|---|
| Solar Pro 4 | 3 blocks | You: seat 1 6 · seat 2 6 · seat 3 6 Opponent: seat 1 6 · seat 2 6 · seat 3 6 | net chips: 2 wins / 0 ties / 1 losses · duplicate matchpoints: 1 wins / 0 ties / 2 losses | 0 missing · 0 void |
|---|
decision_deadline_ms: 60000 → 90000
New participant: gpt-6-luna-blackjack-off-s2
generation_params.max_completion_tokens: 4096 → None
New participant: gpt-6-luna-blackjack-off-s2
generation_params.max_output_tokens: None → 8192
New participant: gpt-6-luna-blackjack-off-s2
decision_deadline_ms: 60000 → 90000
New participant: gpt-6-luna-holdem-on-s2
generation_params.max_completion_tokens: 4096 → None
New participant: gpt-6-luna-holdem-on-s2
generation_params.max_output_tokens: None → 8192
New participant: gpt-6-luna-holdem-on-s2
decision_deadline_ms: 60000 → 90000
New participant: gpt-6-luna-holdem-off-s2
generation_params.max_completion_tokens: 4096 → None
New participant: gpt-6-luna-holdem-off-s2
generation_params.max_output_tokens: None → 8192
New participant: gpt-6-luna-holdem-off-s2
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-blackjack-off-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-blackjack-on-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-team-on-s3
prompt_version: v2 → v3
New participant: gpt-6-luna-craps-off-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-craps-off-s3
prompt_version: v2 → v3
New participant: gpt-6-luna-craps-on-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-craps-on-s3
prompt_version: v2 → v3
New participant: gpt-6-luna-craps-social-on-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-craps-social-on-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-holdem-off-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-holdem-on-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-target-off-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-fallacy-control-on-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-fallacy-streak-on-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-showdown-off-s3
decision_deadline_ms: 90000 → 180000
New participant: gpt-6-luna-videopoker-off-s3
Identical frozen decisions, separate from the Ante Index. Each game, split and stratum is shown with 95% intervals. Season 3 settings · participant IDs ending in -s3.
| Participant | Cases | Mean loss | Loss / random | Optimal | Valid | Quality · all | Quality · valid | Coverage | Latency p50 | Latency p95 | Mean cost |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Blackjack · easyGPT-6 Luna · calibration | 15/150 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 2,581 ms2,241 ms–2,753 ms | 3,138 ms2,807 ms–3,440 ms | $0.0001$0.0001–$0.0001 |
| Blackjack · costlyGPT-6 Luna · calibration | 17/170 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 2,571 ms2,241 ms–2,753 ms | 3,095 ms2,774 ms–3,440 ms | $0.0001$0.0001–$0.0001 |
| Blackjack · easyGPT-6 Luna · sealed | 67/670 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 2,397 ms2,285 ms–2,531 ms | 3,019 ms2,893 ms–3,197 ms | $0.0001$0.0001–$0.0001 |
| Blackjack · costlyGPT-6 Luna · sealed | 75/750 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 2,414 ms2,318 ms–2,542 ms | 3,431 ms2,988 ms–3,748 ms | $0.0001$0.0001–$0.0001 |
| Blackjack Team Play · easyGPT-6 Luna · calibration | 15/150 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 2,529 ms2,375 ms–3,394 ms | 5,222 ms3,557 ms–5,478 ms | $0.0015$0.0012–$0.0019 |
| Blackjack Team Play · costlyGPT-6 Luna · calibration | 13/130 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 2,520 ms2,375 ms–3,386 ms | 5,258 ms3,386 ms–5,478 ms | $0.0015$0.0012–$0.0019 |
| Blackjack Team Play · easyGPT-6 Luna · sealed | 73/730 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 2,590 ms2,439 ms–2,884 ms | 4,380 ms4,045 ms–6,538 ms | $0.0015$0.0013–$0.0016 |
| Blackjack Team Play · costlyGPT-6 Luna · sealed | 71/710 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 2,673 ms2,488 ms–2,963 ms | 5,352 ms4,157 ms–6,538 ms | $0.0014$0.0013–$0.0015 |
| Craps · overallGPT-6 Luna · calibration | 22/300 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 73.3%56.7%–86.7% | 4,965 ms3,817 ms–6,489 ms | 7,905 ms6,926 ms–12,101 ms | $0.0008$0.0008–$0.0008 |
| Craps · easyGPT-6 Luna · calibration | 22/300 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 73.3%56.7%–86.7% | 4,965 ms3,817 ms–6,489 ms | 7,905 ms6,926 ms–12,101 ms | $0.0008$0.0008–$0.0008 |
| Craps · borderlineGPT-6 Luna · calibration | 22/300 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 73.3%56.7%–86.7% | 4,965 ms3,817 ms–6,489 ms | 7,905 ms6,926 ms–12,101 ms | $0.0008$0.0008–$0.0008 |
| Craps · costlyGPT-6 Luna · calibration | 22/300 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 73.3%56.7%–86.7% | 4,965 ms3,817 ms–6,489 ms | 7,905 ms6,926 ms–12,101 ms | $0.0008$0.0008–$0.0008 |
| Showdown Hold’em · easyGPT-6 Luna · calibration | 20/200 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 18,285 ms14,110 ms–21,696 ms | 24,773 ms22,773 ms–31,539 ms | $0.0007$0.0006–$0.0008 |
| Showdown Hold’em · costlyGPT-6 Luna · calibration | 20/200 transport failures | 0.0000.000–0.000 | 0.0000.000–0.000 | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 100.0%100.0%–100.0% | 18,285 ms14,110 ms–21,696 ms | 24,773 ms22,773 ms–31,539 ms | $0.0007$0.0006–$0.0008 |
| Craps · overallGPT-6 Luna · sealed | 89/1190 transport failures | 0.0010.000–0.002 | 0.0020.001–0.004 | 93.3%87.8%–98.0% | 100.0%100.0%–100.0% | 99.8%99.6%–99.9% | 99.8%99.6%–99.9% | 74.8%67.2%–81.5% | 4,656 ms4,477 ms–4,995 ms | 7,889 ms6,576 ms–10,573 ms | $0.0008$0.0008–$0.0008 |
| Craps · easyGPT-6 Luna · sealed | 89/1190 transport failures | 0.0010.000–0.002 | 0.0020.001–0.004 | 93.3%87.8%–98.0% | 100.0%100.0%–100.0% | 99.8%99.6%–99.9% | 99.8%99.6%–99.9% | 74.8%67.2%–81.5% | 4,656 ms4,477 ms–4,995 ms | 7,889 ms6,576 ms–10,573 ms | $0.0008$0.0008–$0.0008 |
| Craps · borderlineGPT-6 Luna · sealed | 89/1190 transport failures | 0.0010.000–0.002 | 0.0020.001–0.004 | 93.3%87.8%–98.0% | 100.0%100.0%–100.0% | 99.8%99.6%–99.9% | 99.8%99.6%–99.9% | 74.8%67.2%–81.5% | 4,656 ms4,477 ms–4,995 ms | 7,889 ms6,576 ms–10,573 ms | $0.0008$0.0008–$0.0008 |
| Craps · costlyGPT-6 Luna · sealed | 89/1190 transport failures | 0.0010.000–0.002 | 0.0020.001–0.004 | 93.3%87.8%–98.0% | 100.0%100.0%–100.0% | 99.8%99.6%–99.9% | 99.8%99.6%–99.9% | 74.8%67.2%–81.5% | 4,656 ms4,477 ms–4,995 ms | 7,889 ms6,576 ms–10,573 ms | $0.0008$0.0008–$0.0008 |
| Blackjack · overallGPT-6 Luna · calibration | 29/290 transport failures | 0.0010.000–0.003 | 0.0030.000–0.009 | 93.1%82.8%–100.0% | 100.0%100.0%–100.0% | 99.7%99.1%–100.0% | 99.7%99.1%–100.0% | 100.0%100.0%–100.0% | 2,581 ms2,327 ms–2,857 ms | 4,249 ms3,131 ms–9,286 ms | $0.0001$0.0001–$0.0001 |
| Blackjack Team Play · overallGPT-6 Luna · sealed | 118/1180 transport failures | 0.0020.001–0.003 | 0.0040.001–0.009 | 94.1%89.0%–97.5% | 100.0%100.0%–100.0% | 99.6%99.1%–99.9% | 99.6%99.1%–99.9% | 100.0%100.0%–100.0% | 2,989 ms2,766 ms–3,231 ms | 5,894 ms4,934 ms–6,538 ms | $0.0015$0.0013–$0.0016 |
| Blackjack · overallGPT-6 Luna · sealed | 119/1190 transport failures | 0.0020.000–0.004 | 0.0050.001–0.011 | 95.0%90.8%–98.3% | 100.0%100.0%–100.0% | 99.5%98.9%–99.9% | 99.5%98.9%–99.9% | 100.0%100.0%–100.0% | 2,530 ms2,413 ms–2,623 ms | 3,650 ms3,262 ms–4,139 ms | $0.0001$0.0001–$0.0001 |
| Blackjack Team Play · overallGPT-6 Luna · calibration | 30/300 transport failures | 0.0020.000–0.006 | 0.0070.000–0.018 | 93.3%83.3%–100.0% | 100.0%100.0%–100.0% | 99.3%98.2%–100.0% | 99.3%98.2%–100.0% | 100.0%100.0%–100.0% | 2,785 ms2,525 ms–3,537 ms | 7,239 ms4,343 ms–11,628 ms | $0.0015$0.0012–$0.0018 |
| Blackjack · borderlineGPT-6 Luna · calibration | 10/100 transport failures | 0.0030.000–0.008 | 0.0240.000–0.063 | 80.0%50.0%–100.0% | 100.0%100.0%–100.0% | 97.6%93.7%–100.0% | 97.6%93.7%–100.0% | 100.0%100.0%–100.0% | 3,055 ms2,379 ms–3,846 ms | 7,262 ms3,322 ms–9,286 ms | $0.0001$0.0001–$0.0001 |
| Blackjack Team Play · borderlineGPT-6 Luna · calibration | 14/140 transport failures | 0.0050.000–0.012 | 0.0310.000–0.066 | 85.7%64.3%–100.0% | 100.0%100.0%–100.0% | 96.9%93.4%–100.0% | 96.9%93.4%–100.0% | 100.0%100.0%–100.0% | 2,990 ms2,671 ms–4,340 ms | 9,712 ms4,477 ms–11,628 ms | $0.0015$0.0011–$0.0019 |
| Blackjack Team Play · borderlineGPT-6 Luna · sealed | 37/370 transport failures | 0.0060.002–0.010 | 0.0350.012–0.065 | 81.1%67.6%–91.9% | 100.0%100.0%–100.0% | 96.5%93.5%–98.8% | 96.5%93.5%–98.8% | 100.0%100.0%–100.0% | 3,642 ms3,117 ms–4,128 ms | 6,020 ms5,304 ms–9,306 ms | $0.0014$0.0012–$0.0017 |
| Blackjack · borderlineGPT-6 Luna · sealed | 41/410 transport failures | 0.0060.001–0.014 | 0.0390.007–0.089 | 85.4%73.2%–95.1% | 100.0%100.0%–100.0% | 96.1%91.1%–99.3% | 96.1%91.1%–99.3% | 100.0%100.0%–100.0% | 2,636 ms2,471 ms–2,951 ms | 4,139 ms3,273 ms–6,151 ms | $0.0001$0.0001–$0.0001 |
| Craps · overallGPT-6 Luna · calibration | 8/300 transport failures | 0.0070.002–0.012 | 0.0140.004–0.025 | 50.0%14.3%–83.3% | 100.0%100.0%–100.0% | 98.6%97.5%–99.6% | 98.6%97.5%–99.6% | 26.7%13.3%–43.3% | 6,254 ms3,219 ms–6,887 ms | 7,049 ms6,517 ms–7,136 ms | $0.0008$0.0008–$0.0008 |
| Craps · easyGPT-6 Luna · calibration | 8/300 transport failures | 0.0070.002–0.012 | 0.0140.004–0.025 | 50.0%14.3%–83.3% | 100.0%100.0%–100.0% | 98.6%97.5%–99.6% | 98.6%97.5%–99.6% | 26.7%13.3%–43.3% | 6,254 ms3,219 ms–6,887 ms | 7,049 ms6,517 ms–7,136 ms | $0.0008$0.0008–$0.0008 |
| Craps · borderlineGPT-6 Luna · calibration | 8/300 transport failures | 0.0070.002–0.012 | 0.0140.004–0.025 | 50.0%14.3%–83.3% | 100.0%100.0%–100.0% | 98.6%97.5%–99.6% | 98.6%97.5%–99.6% | 26.7%13.3%–43.3% | 6,254 ms3,219 ms–6,887 ms | 7,049 ms6,517 ms–7,136 ms | $0.0008$0.0008–$0.0008 |
| Craps · costlyGPT-6 Luna · calibration | 8/300 transport failures | 0.0070.002–0.012 | 0.0140.004–0.025 | 50.0%14.3%–83.3% | 100.0%100.0%–100.0% | 98.6%97.5%–99.6% | 98.6%97.5%–99.6% | 26.7%13.3%–43.3% | 6,254 ms3,219 ms–6,887 ms | 7,049 ms6,517 ms–7,136 ms | $0.0008$0.0008–$0.0008 |
| Showdown Hold’em · overallGPT-6 Luna · calibration | 30/300 transport failures | 0.0190.000–0.043 | 0.0260.000–0.071 | 90.0%76.7%–100.0% | 100.0%100.0%–100.0% | 97.4%92.9%–100.0% | 97.4%92.9%–100.0% | 100.0%100.0%–100.0% | 15,842 ms13,107 ms–20,321 ms | 27,299 ms23,229 ms–31,539 ms | $0.0007$0.0006–$0.0008 |
| Showdown Hold’em · borderlineGPT-6 Luna · sealed | 40/400 transport failures | 0.0340.014–0.059 | 0.4110.166–0.695 | 80.0%67.5%–90.1% | 100.0%100.0%–100.0% | 58.9%30.5%–83.4% | 58.9%30.5%–83.4% | 100.0%100.0%–100.0% | 11,808 ms10,041 ms–15,104 ms | 26,583 ms21,237 ms–44,118 ms | $0.0006$0.0005–$0.0007 |
| Craps · overallGPT-6 Luna · sealed | 30/1190 transport failures | 0.0490.002–0.149 | 0.1030.004–0.320 | 73.3%58.1%–88.0% | 100.0%100.0%–100.0% | 89.7%68.0%–99.6% | 89.7%68.0%–99.6% | 25.2%18.5%–32.8% | 5,202 ms4,341 ms–6,761 ms | 9,872 ms7,375 ms–16,057 ms | $0.0008$0.0008–$0.0009 |
| Craps · easyGPT-6 Luna · sealed | 30/1190 transport failures | 0.0490.002–0.149 | 0.1030.004–0.320 | 73.3%58.1%–88.0% | 100.0%100.0%–100.0% | 89.7%68.0%–99.6% | 89.7%68.0%–99.6% | 25.2%18.5%–32.8% | 5,202 ms4,341 ms–6,761 ms | 9,872 ms7,375 ms–16,057 ms | $0.0008$0.0008–$0.0009 |
| Craps · borderlineGPT-6 Luna · sealed | 30/1190 transport failures | 0.0490.002–0.149 | 0.1030.004–0.320 | 73.3%58.1%–88.0% | 100.0%100.0%–100.0% | 89.7%68.0%–99.6% | 89.7%68.0%–99.6% | 25.2%18.5%–32.8% | 5,202 ms4,341 ms–6,761 ms | 9,872 ms7,375 ms–16,057 ms | $0.0008$0.0008–$0.0009 |
| Craps · costlyGPT-6 Luna · sealed | 30/1190 transport failures | 0.0490.002–0.149 | 0.1030.004–0.320 | 73.3%58.1%–88.0% | 100.0%100.0%–100.0% | 89.7%68.0%–99.6% | 89.7%68.0%–99.6% | 25.2%18.5%–32.8% | 5,202 ms4,341 ms–6,761 ms | 9,872 ms7,375 ms–16,057 ms | $0.0008$0.0008–$0.0009 |
| Video Poker · borderlineGPT-6 Luna · calibration | 9/90 transport failures | 0.0550.000–0.149 | 0.0440.000–0.124 | 77.8%55.3%–100.0% | 100.0%100.0%–100.0% | 95.6%87.6%–100.0% | 95.6%87.6%–100.0% | 100.0%100.0%–100.0% | 8,932 ms3,177 ms–16,221 ms | 16,963 ms9,941 ms–17,457 ms | $0.0006$0.0004–$0.0008 |
| Showdown Hold’em · borderlineGPT-6 Luna · calibration | 10/100 transport failures | 0.0560.013–0.118 | 0.5990.118–1.299 | 70.0%40.0%–90.2% | 100.0%100.0%–100.0% | 40.1%0.0%–88.2% | 40.1%0.0%–88.2% | 100.0%100.0%–100.0% | 13,107 ms9,169 ms–19,351 ms | 27,299 ms14,522 ms–28,894 ms | $0.0006$0.0004–$0.0008 |
| Showdown Hold’em · overallGPT-6 Luna · sealed | 120/1200 transport failures | 0.0790.022–0.156 | 0.1200.034–0.245 | 90.0%84.2%–95.0% | 100.0%100.0%–100.0% | 88.0%75.5%–96.6% | 88.0%75.5%–96.6% | 100.0%100.0%–100.0% | 15,221 ms12,816 ms–17,222 ms | 29,396 ms26,624 ms–37,117 ms | $0.0007$0.0006–$0.0007 |
| Video Poker · borderlineGPT-6 Luna · sealed | 41/410 transport failures | 0.0890.049–0.143 | 0.0760.042–0.124 | 63.4%48.8%–75.7% | 100.0%100.0%–100.0% | 92.4%87.6%–95.8% | 92.4%87.6%–95.8% | 100.0%100.0%–100.0% | 6,540 ms6,102 ms–8,444 ms | 29,135 ms12,889 ms–44,148 ms | $0.0006$0.0005–$0.0008 |
| Showdown Hold’em · easyGPT-6 Luna · sealed | 80/800 transport failures | 0.1020.013–0.220 | 0.1070.015–0.230 | 95.0%90.0%–98.8% | 100.0%100.0%–100.0% | 89.3%77.0%–98.5% | 89.3%77.0%–98.5% | 100.0%100.0%–100.0% | 16,660 ms15,098 ms–18,920 ms | 29,396 ms26,644 ms–35,716 ms | $0.0007$0.0006–$0.0008 |
| Showdown Hold’em · costlyGPT-6 Luna · sealed | 80/800 transport failures | 0.1020.013–0.220 | 0.1070.015–0.230 | 95.0%90.0%–98.8% | 100.0%100.0%–100.0% | 89.3%77.0%–98.5% | 89.3%77.0%–98.5% | 100.0%100.0%–100.0% | 16,660 ms15,098 ms–18,920 ms | 29,396 ms26,644 ms–35,716 ms | $0.0007$0.0006–$0.0008 |
| Video Poker · overallGPT-6 Luna · calibration | 30/300 transport failures | 0.1100.000–0.309 | 0.0300.000–0.100 | 90.0%80.0%–100.0% | 100.0%100.0%–100.0% | 97.0%90.0%–100.0% | 97.0%90.0%–100.0% | 100.0%100.0%–100.0% | 3,612 ms2,943 ms–5,610 ms | 14,123 ms8,782 ms–17,457 ms | $0.0004$0.0003–$0.0005 |
| Video Poker · costlyGPT-6 Luna · sealed | 51/510 transport failures | 0.1130.000–0.308 | 0.0130.000–0.044 | 96.1%90.2%–100.0% | 100.0%100.0%–100.0% | 98.7%95.6%–100.0% | 98.7%95.6%–100.0% | 100.0%100.0%–100.0% | 2,893 ms2,677 ms–3,362 ms | 6,489 ms4,735 ms–11,240 ms | $0.0003$0.0002–$0.0003 |
| Video Poker · overallGPT-6 Luna · sealed | 120/1200 transport failures | 0.1580.058–0.283 | 0.0340.011–0.075 | 83.3%76.7%–90.0% | 100.0%100.0%–100.0% | 96.6%92.5%–98.9% | 96.6%92.5%–98.9% | 100.0%100.0%–100.0% | 3,888 ms3,161 ms–4,513 ms | 13,239 ms9,711 ms–29,135 ms | $0.0004$0.0003–$0.0005 |
| Video Poker · easyGPT-6 Luna · calibration | 17/170 transport failures | 0.1650.000–0.495 | 0.0360.000–0.137 | 94.1%82.4%–100.0% | 100.0%100.0%–100.0% | 96.4%86.3%–100.0% | 96.4%86.3%–100.0% | 100.0%100.0%–100.0% | 3,655 ms2,708 ms–5,564 ms | 8,042 ms5,582 ms–9,733 ms | $0.0003$0.0003–$0.0004 |
| Video Poker · costlyGPT-6 Luna · calibration | 12/120 transport failures | 0.2340.000–0.702 | 0.0370.000–0.158 | 91.7%75.0%–100.0% | 100.0%100.0%–100.0% | 96.3%84.2%–100.0% | 96.3%84.2%–100.0% | 100.0%100.0%–100.0% | 2,564 ms2,215 ms–4,263 ms | 6,539 ms3,655 ms–7,619 ms | $0.0003$0.0002–$0.0003 |
| Video Poker · easyGPT-6 Luna · sealed | 65/650 transport failures | 0.2350.051–0.448 | 0.0350.008–0.090 | 92.3%86.2%–98.5% | 100.0%100.0%–100.0% | 96.5%91.0%–99.2% | 96.5%91.0%–99.2% | 100.0%100.0%–100.0% | 3,337 ms2,862 ms–3,829 ms | 8,530 ms6,412 ms–12,906 ms | $0.0003$0.0003–$0.0003 |