Every number worth knowing before you post a blind: exact dealing odds, what the table is likely holding against you, and all 154 playable starting-hand shapes ranked by how often they actually win a showdown.
Exact odds — every one of the 2,598,960 possible five-card hands enumerated.
| Holding off the deal | Odds | Holding off the deal | Odds |
|---|---|---|---|
| deuce + 1 more wheel card (3 4 5 7) | 27.36% | pat T-low or better | 4.75% |
| deuce + 2 more wheel cards | 11.14% | pat 9-low or better | 2.04% |
| deuce + 3 more wheel cards | 1.54% | pat 8-low or better | 0.71% |
| any 2 distinct wheel ranks | 58.47% | pat 7-low | 0.16% |
| any 3 distinct wheel ranks | 17.91% | 3 distinct ranks 2–8 | 45.11% |
| any 4 distinct wheel ranks | 1.91% | 4 distinct ranks 2–8 | 11.17% |
| all 5 wheel ranks dealt | 0.04% | dealt straight or flush | 0.55% |
200,000 simulated deals per table size, card removal honored. A “premium start” here is three or more wheel cards.
| Opponents | ≥1 pat T-low | ≥1 premium start | ≥1 strong one-card draw* | How many premium starts out there |
|---|---|---|---|---|
| 1 | 4.7% | 17.8% | 11.1% | none 82% · one 18% |
| 2 | 9.2% | 33.2% | 21.3% | none 67% · one 31% · both 2.6% |
| 3 | 13.9% | 46.5% | 30.8% | none 53% · one 39% · two 7% · three 0.3% |
| 4 | 17.8% | 57.1% | 38.9% | none 43% · one 44% · two 12% · 3+ 1.1% |
| 5 | 21.9% | 66.2% | 46.6% | none 34% · one 46% · two 18% · 3+ 2.6% |
*four distinct ranks 8 or lower — a one-card draw at an 8-low shell.
Second draw is done: you’re in position and everyone drawing missed. If you stand pat, these are the exact odds you get outdrawn on the final draw. Heads Up: one opponent on the listed draw, outs over 43 unseen cards. Vs. 2 Opponents: both opponents hold the listed draw type and each catches one card — the odds at least one gets there.
| You stand pat with | vs 7-4-3-2 (smooth 7, deuce) | vs 8-5-4-2 (8 draw) | vs 2-3-4-5 (wheel cards) |
|---|---|---|---|
| 9-7-4-3-2 | 27.9%12 outs: 5·6·8 | 23.3%10 outs: 3·6·7 | 23.3%10 outs: 7·8·9 |
| 9-6-4-3-2 | 25.6%11 outs: 5·6·8 | 23.3%10 outs: 3·6·7 | 25.6%11 outs: 7·8·9 |
| 9-5-4-3-2 | 25.6%11 outs: 5·6·8 | 25.6%11 outs: 3·6·7 | 18.6%8 outs: 7·8 |
| 8-7-4-3-2 | 18.6%8 outs: 5·6 | 16.3%7 outs: 3·6 | 14.0%6 outs: 7·8 |
| 8-6-4-3-2 | 16.3%7 outs: 5·6 | 7.0%3 outs: 3 | 16.3%7 outs: 7·8 |
| 8-5-4-3-2 | 16.3%7 outs: 5·6 | 0 outschop only | 9.3%4 outs: 7 |
| 7-6-4-3-2 | 9.3%4 outs: 5 | drawing dead | 7.0%3 outs: 7 |
Winning catches are ranks; counts already account for every card you and the drawing hands block. Chops (identical hands) are not counted as outdraws. Exact enumeration — no simulation noise.
All 154 two-, three-, and four-card classes (X = ten or higher), 10,000 showdowns each: you draw to the shape, everyone finishes all three draws against random hands. Toggle the table size from heads-up to 6-handed — at each size a random hand’s baseline is 100% divided by the players. Sorted by win rate; red rows are hands that can fill their own straight.
Method: hero’s first draw discards exactly the X cards; later draws follow the frozen rollout heuristic on both sides, no betting, ties split. 10k samples puts ±0.5pt of noise on each figure — adjacent ranks are interchangeable, tiers are not. Generated by the same engine that trained the app’s blueprints (verified bit-exact vs its Swift port).