A7o ace-seven offsuit
Ace-seven offsuit is 12 of the 1,326 two-card starting hands (0.90%). All-in before the flop it wins 58.8% against a random hand, 36th of 169. Below: its equity against every other hand and against the four opponent ranges, and the solved push/fold answer at every stack from 2 to 25 bb, heads-up, 6-max and 9-max. Every number is computed, and the working is at the bottom.
A7o against every hand
All-in preflop, five cards to come. A7o is ahead (better than 50%) against 87.5% of the 1,225 hands an opponent can still hold, and behind against 11.9%. Its own two cards remove 101 combos from the deck, which is why the live counts below do not always match the usual 4, 6 and 12. Tap a cell to open that hand.
Best matchups
Worst matchups
What A7o makes, street by street
The best hand A7o holds on every flop, turn and river, with every board dealt (19,600 flops, 2,118,760 rivers). The last column is any two cards, for comparison.
| Best hand | Flop | Turn | River | River, any two |
|---|---|---|---|---|
| Royal flush | 0% | 0.00043% | 0.0023% | 0.0032% |
| Straight flush | 0% | 0.0026% | 0.014% | 0.028% |
| Four of a kind | 0.010% | 0.046% | 0.13% | 0.17% |
| Full house | 0.092% | 0.63% | 2.22% | 2.60% |
| Flush | 0% | 0.43% | 1.96% | 3.03% |
| Straight | 0% | 0.77% | 3.05% | 4.62% |
| Three of a kind | 1.57% | 3.06% | 4.45% | 4.83% |
| Two pair | 4.04% | 11.4% | 22.8% | 23.5% |
| One pair | 40.4% | 48.0% | 45.9% | 43.8% |
| High card | 53.9% | 35.6% | 19.5% | 17.4% |
A7o after the deal
How often a hand like this flops something, and what its draws are worth, every card counted:
A7o against AA in the equity calculator → Add more hands, a flop or dead cards, and every board is dealt again.
A7o against the four opponents
The same matrix, weighted by the ranges the trainer's bots play. These ranges are assumptions, not solver output, and are printed in full on /opponents so you can argue with them.
| Opponent | A7o equity | Live combos in range |
|---|---|---|
| The Nit | 26.0% | 42 |
| The Calling Station | 41.8% | 216 |
| The Maniac | 45.2% | 301 |
| The Reg | 39.2% | 149 |
Push/fold with A7o
Heads-up, the small blind shoves at every depth from 2 to 25 bb when folded to; the big blind calls at every depth from 2 to 25 bb against the equilibrium shove. At 6-max, A7o open-shoves at every depth up to 10 bb from UTG and open-shoves at every depth from 2 to 25 bb from the button; at 9-max from UTG it open-shoves at every depth up to 6 bb. Each cell is the solved frequency at that stack, read from the same tables as the push/fold charts; hover or tap for the EV against folding.
Heads-up · no ante
6-max · 1 bb big-blind ante
| bb | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Open-shove, folded to | ||||||||||||||||||||||||
| UTG | ● | ● | ● | ● | ● | ● | ● | ● | ● | 38 | ||||||||||||||
| HJ | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | |||||||||||
| CO | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | 51 | ||||||
| BTN | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● |
| SB | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● |
| Big blind calls a shove from | ||||||||||||||||||||||||
| vs UTG | ● | ● | ● | ● | ● | ● | ● | ● | ● | |||||||||||||||
| vs HJ | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ||||||||||||
| vs CO | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | |||||||||
| vs BTN | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | 40 | |
| vs SB | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● |
9-max · 1 bb big-blind ante
| bb | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Open-shove, folded to | ||||||||||||||||||||||||
| UTG | ● | ● | ● | ● | ● | |||||||||||||||||||
| UTG+1 | ● | ● | ● | ● | ● | ● | ||||||||||||||||||
| UTG+2 | ● | ● | ● | ● | ● | ● | ● | |||||||||||||||||
| LJ | ● | ● | ● | ● | ● | ● | ● | ● | ● | 38 | ||||||||||||||
| HJ | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | |||||||||||
| CO | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | 51 | ||||||
| BTN | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● |
| SB | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● |
| Big blind calls a shove from | ||||||||||||||||||||||||
| vs UTG | ● | ● | ● | ● | ● | ● | ● | |||||||||||||||||
| vs UTG+1 | ● | ● | ● | ● | ● | ● | ● | 25 | ||||||||||||||||
| vs UTG+2 | ● | ● | ● | ● | ● | ● | ● | ● | ||||||||||||||||
| vs LJ | ● | ● | ● | ● | ● | ● | ● | ● | ● | |||||||||||||||
| vs HJ | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ||||||||||||
| vs CO | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | |||||||||
| vs BTN | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | 40 | |
| vs SB | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● | ● |
Knowing where A7o sits on the chart is half of it. The trainer deals every hand at the depth and seat you pick and prices each wrong shove or call in big blinds.
- Hand vs hand: our Rust engine (the same WebAssembly file the trainer runs) dealt 200,000 random five-card boards for every one of the 14,196 class-vs-class matchups, with card-conflicting combos excluded and ties counted as half. Sampling error is about ±0.1 percentage points per matchup.
- Hand vs a range: each villain class is weighted by how many of its combos are still live once A7o's two cards are out, so against a random hand the equity is Σ (live combos × equity) ÷ 1,225 = 58.8%. Against an opponent, the sum runs over that opponent's range only.
- Equity against a random hand is not how good a hand is to play. It ignores position, stack depth and what your opponent actually continues with. The push/fold rows are the playable answer, and only for the shove-or-fold endgame.
- Push/fold rows are copied from the solved tables, which state their own assumptions on the heads-up, 6-max and 9-max hubs (6-max and 9-max ignore three-way all-ins). Frequencies between 20% and 80% count as mixed.
- Raw data: a7o.json and the summary hands.json, built by
tools/build-hands.mjs.