// push/fold charts · 6-max

6-max push/fold charts

The shove-or-fold equilibrium for a 6-handed table with a big-blind ante, from 2 to 25 big blinds, for every seat. At 10 bb it opens 23.3% from UTG and 73.9% from the small blind. Our solver produced every number here; the two assumptions it needed are printed at the bottom of the page.

Open-shove range size by seat and depth

Share of all 1,326 starting hands each seat shoves when it is folded to, weighted by real combos and by the frequency of mixed hands. Open a depth for every chart, every calling range and the EV of all 169 hands.

StackUTGHJCOBTNSBCharts
2 bb100.0%100.0%100.0%98.2%100.0%2 bb →
3 bb90.0%90.0%90.0%89.4%100.0%3 bb →
4 bb65.0%65.0%66.8%71.9%94.6%4 bb →
5 bb37.6%43.3%48.4%56.6%85.8%5 bb →
6 bb32.3%35.1%43.3%52.0%82.2%6 bb →
7 bb29.6%32.7%39.1%49.5%81.0%7 bb →
8 bb27.0%31.1%37.0%47.3%77.3%8 bb →
9 bb25.2%30.5%34.2%45.2%76.1%9 bb →
10 bb23.3%28.5%33.5%43.6%73.9%10 bb →
11 bb21.7%27.0%32.1%42.1%71.0%11 bb →
12 bb20.3%26.0%31.8%39.6%69.9%12 bb →
13 bb19.7%24.3%31.1%37.8%66.8%13 bb →
14 bb18.1%23.1%29.7%36.3%64.7%14 bb →
15 bb17.1%21.7%29.4%36.1%64.2%15 bb →
16 bb16.6%20.8%28.2%35.1%61.6%16 bb →
17 bb15.6%20.1%26.2%34.0%60.6%17 bb →
18 bb15.0%18.8%25.5%34.1%59.7%18 bb →
19 bb14.0%17.9%23.8%33.3%56.9%19 bb →
20 bb13.7%17.6%22.9%33.3%55.6%20 bb →
21 bb13.2%16.6%22.6%32.4%53.5%21 bb →
22 bb12.9%16.1%21.4%31.1%52.8%22 bb →
23 bb12.4%15.1%21.3%31.0%51.2%23 bb →
24 bb11.7%14.3%19.5%29.8%49.3%24 bb →
25 bb11.2%14.0%19.0%29.6%48.3%25 bb →

One chart per seat, every depth

Each cell is the deepest stack, in big blinds, from which that hand still open-shoves from that seat, checking every depth from 2 bb up. At or below the number: shove. Above: fold. 25+ means it shoves at every depth we solved; a dash means it folds even at 2 bb. An asterisk marks a hand that stops and starts again deeper; hover or tap it for the depths.

UTG: shove up to

HJ: shove up to

CO: shove up to

BTN: shove up to

SB: shove up to

How wide the big blind defends

The big blind's calling range against a shove from each seat, as a share of all hands. A later shover's range is wider, so the call is too. Every other seat's calling range is on the depth pages.

Stackvs UTGvs HJvs COvs BTNvs SB
2 bb100.0%100.0%100.0%100.0%100.0%
3 bb100.0%100.0%100.0%100.0%100.0%
4 bb100.0%100.0%100.0%100.0%100.0%
5 bb88.3%89.1%92.8%94.6%95.5%
6 bb62.4%64.1%68.0%75.9%82.2%
7 bb42.4%45.4%51.7%58.1%75.0%
8 bb32.4%36.3%40.9%48.8%66.9%
9 bb26.4%31.2%33.9%42.4%61.3%
10 bb22.6%26.8%30.8%37.3%56.6%
11 bb18.5%23.4%27.9%34.9%53.2%
12 bb16.0%19.4%25.8%32.1%49.8%
13 bb14.5%18.8%24.3%29.4%47.3%
14 bb13.2%16.9%22.2%28.6%45.3%
15 bb11.5%14.2%19.2%26.4%42.3%
16 bb10.8%13.4%18.4%24.7%40.5%
17 bb9.8%12.8%17.2%23.3%38.2%
18 bb8.3%11.5%15.9%22.3%36.4%
19 bb8.3%10.9%14.5%21.6%35.4%
20 bb7.8%10.4%13.3%19.8%33.9%
21 bb7.8%9.6%12.8%18.4%32.7%
22 bb7.7%8.3%12.3%18.1%31.5%
23 bb6.7%8.3%11.5%17.8%30.4%
24 bb5.9%8.3%11.4%16.7%29.7%
25 bb5.9%7.8%10.6%15.5%28.5%

When these charts apply

Short-stacked in a tournament, folded to you, with no reason to fear the payout structure: this is the arithmetic of the shove. Two things move the real answer away from it. Near the money or a pay jump, chips you lose are worth more than chips you win, and chip-EV ranges are too wide; that is ICM, and these pages do not model it. And as stacks get deeper, a real player also has a min-raise; that option is worth something this model was not allowed to use, so treat the deep end of the table as the price of the shove rather than the best play.

For the two-player version of the same game, with no ante, see the heads-up charts.

The trainer deals real hands at the frequency they really arrive, in the seat you pick, and prices every wrong shove or call in big blinds.

Drill 6-max push/fold →
// what was solved, and what was assumed
  • 6-max, blinds 0.5 / 1, 1 bb big-blind ante (dead money, posted by the big blind). Every player starts the hand with the same stack, counted before posting, so the big blind has 1 bb less to play with.
  • The first player in may only shove or fold. Facing a shove, each player behind may only call or fold. No limps, no raises short of all-in.
  • Assumption: once one player calls, everyone behind folds. Three-way all-ins are left out of the model, because pricing them needs a 169×169×169 equity table. In real play a big blind sometimes overcalls with aces; here it cannot, so shoves into several players are priced slightly too well.
  • Assumption: players who folded held random cards. Card removal is applied between the shover and each possible caller, not for the hands that folded in front. One consequence: a seat's ranges depend only on the players left behind it, which is why every 6-max chart matches the 9-max chart for the seat with the same players behind it (6-max UTG is 9-max LJ; HJ through BB are the same seats).
  • Solved by fictitious play (linearly averaged) over a 169×169 all-in equity matrix from our engine, 200,000 sampled runouts per matchup, card removal corrected, 3,000 iterations per seat. Measured, not assumed: no seat, in any spot on these pages, could gain more than 0.046 mbb (thousandths of a big blind) per hand by deviating.
  • Checked against our heads-up table: the same solver run with two players and no ante reproduces the heads-up charts, with no disagreement bigger than 0.1 percentage points of equity.
  • EV is the aggressive action minus folding, in big blinds, against everyone else's equilibrium ranges. Chip EV only: no ICM, so near a bubble or a pay jump these ranges are too wide.
  • Raw data: 6max.json (summary) and one file per depth, e.g. 6max/10.json.