Push or Fold Poker Charts From 15 Big Blinds Down

Below roughly 15 big blinds in an unopened pot, the raise-fold line stops existing: your raise is too large a slice of your stack to abandon. Your real options collapse to all-in or fold, and push/fold charts tell you which hands go in from each seat at each stack depth. They come from Nash equilibrium solutions to that two-action game, and they are close enough to unexploitable that memorising them is the highest-return hour in tournament poker.

Why raise-folding stops working

Blinds are 500/1000 with a 1,000 big blind ante and you have 12,000 behind — 12bb. There is 2.5bb sitting in the middle before anyone acts.

Open to 2.2bb and you have 9.8bb left. The big blind shoves. You are now risking 9.8bb to win a pot of roughly 15bb, needing under 40 per cent equity, and you hold the hand you raised with. Folding is a disaster and calling was never your plan. You have built a raise with no fold in it.

Shoving all 12bb instead does three things at once. It removes that trap. It denies the equity of every hand that would have called a small raise and outflopped you. And it wins 2.5bb uncontested when everyone folds, which is a 20 per cent stack increase for zero variance.

That last number is the one people underrate. At 12bb you do not need to win pots — you need to win blinds, repeatedly, without ever seeing a flop.

Where the threshold actually sits

Fifteen big blinds is a guideline, not a wall. The real trigger is how much of your stack the raise commits and how many players are left behind you.

Shove when your open would commit more than about a third of your stack and you could not fold to a re-shove behind it.

Position moves the threshold. From under the gun at a nine-handed table you have eight players to get through, so raise-folding retains some value down to about 12bb. From the small blind, with only the big blind left to act, shoving is correct past 20bb — and heads-up, past 25bb. More players behind means more chances to be raised, which is exactly what makes a foldable raise valuable again.

Antes push the threshold up by two big blinds or so, because the pot you are attacking is bigger relative to the risk. Above the threshold you are back in normal raising territory, where the solver-built preflop opening ranges for each seat take over.

Shoving ranges by position

Approximate Nash shoving ranges in an unopened pot, no antes, nine-handed:

Effective stack	UTG	CO	BTN	SB
15bb	~10%	~15%	~22%	~32%
12bb	~11%	~18%	~26%	~38%
10bb	~13%	~22%	~32%	~45%
7bb	~17%	~28%	~42%	~55%
5bb	~22%	~35%	~50%	~65%

Two patterns do most of the work. Roughly halving your stack roughly doubles your range, and each seat closer to the button widens you by about half again.

The hand ordering inside those ranges is not the ordering you use at 60bb. Short-stacked equity is about how you fare all-in against a calling range, not about playability after the flop.

• Small pairs rise sharply. 44 is never behind more than a coin flip against an unpaired hand, so it shoves from almost anywhere at 10bb.
• Suited aces beat offsuit broadways. A5s carries a blocker to the ace-x calls and a flush to fall back on; KJo is dominated by most of what continues.
• Suited connectors fall. 87s needs flops to realise its value, and there are no flops when the stack goes in preflop.

Calling a shove is a different chart

Never call with your shoving range. When you shove, folds win you the pot; when you call, nothing wins but the showdown.

Put a number on it. A button with 10bb shoves, the small blind folds, and you are in the big blind. The pot holds 11.5bb and you must put in 9bb more, so you need about 44 per cent equity. Add a big blind ante and that price drops to roughly 42 per cent.

Against a button shoving a third of all hands, 44 per cent is a real bar. AKo and 88 clear it comfortably. A9o and KTs sit around break-even. Everything below is a fold, however painful it looks when you are the one being run over.

The practical shape: calling ranges run around half the width of the range being shoved at you. If you never let a short stack fold through you, you are the one paying for their aggression.

Where ICM makes the chart wrong

Nash charts assume every chip is worth the same amount. In a tournament with a payout ladder, they are not. The chips you can lose are worth more than the chips you can win, because busting ends your equity in every prize above you.

That tax lands hardest on calls. On a bubble with a meaningful pay jump, a big blind call that is break-even in chips can lose real money, and the correct calling range tightens noticeably — sometimes to pairs and ace-king only.

Shoving ranges move too, but in both directions:

• Covered by the table, near a pay jump: tighten. Your stack is worth more surviving into the money than gambling for a marginal edge.
• Covering everyone at the table: widen. Every player behind you is calling tighter than the chart says, so your fold equity goes up and you can attack with hands Nash would fold.
• Two short stacks behind you: tighten, and let them bust first.

Run your stack, the remaining stacks and the payout structure through an ICM calculator and the size of that tax becomes a number rather than a feeling. Do it after sessions on the spots that felt close, and the pattern in which bubble calls to abandon starts to stick. The wider strategic frame around short-stack play — when to accumulate, when to wait — is covered in short stack MTT strategy.

Antes change every number

Antes are dead money, and dead money widens shoves. Without antes at 0.5/1 you risk your stack to win 1.5bb. With a full big blind ante the prize becomes 2.5bb — a 67 per cent increase in the reward for a steal that costs the same.

The effect is large. A typical shoving range at 10bb widens by roughly a third once antes are in, and the earlier positions feel it most, since they were closest to folding in the first place. The detail on how antes widen shoving ranges is worth reading before you trust a no-ante chart at an ante table.

One structural note: the big blind ante format posts a single ante from one player rather than collecting from everyone. It puts the same money in the middle but changes who pays and when, and the big blind ante affects the short stack in the big blind most of all, since posting it can leave them with almost nothing behind.

Making the charts automatic

You will never recall a 169-hand grid at the table. You will recall boundaries.

1. Learn small blind and button first. They are where most of your shoves happen and where the ranges are widest, so errors there cost the most.
2. Memorise thresholds, not tables. At 10bb on the button: any pair, any ace, any two broadway, suited kings. That approximation costs a fraction of a big blind against the exact chart.
3. Then learn big blind calls. The second-most frequent short-stack decision, and the one where ICM does the most damage.
4. Drill under time pressure. Recognition at three seconds is a different skill from recall at thirty.

Sharkling's preflop trainer deals these spots at the stack depths you actually play and grades each decision against the solver, so the feedback arrives while the hand is still in your head. Twenty minutes a day for a week is enough to make the button and small blind automatic.

Push/fold is the most solved region of poker and the cheapest to get right. Every big blind you save by shoving instead of raise-folding is one more orbit of survival — and survival, at 12bb, is the whole strategy.