Sklansky-Chubukov Rankings
The full SC rankings table: each starting hand's worst-case profitable push EV at any stack depth. The original push-fold framework, with stack-size lookups for every hand.
Sklansky-Chubukov Rankings
The full SC rankings table: each starting hand's worst-case profitable push EV at any stack depth. The original push-fold framework, with stack-size lookups for every hand.
Sklansky-Chubukov number: the maximum effective stack (in big blinds) at which shoving from the small blind is still profitable, even if the big blind plays perfectly with your cards face-up. Shove if your stack is at or below the number. All 169 starting hands.
| Rank | Hand | Max push BB | Push at 10bb? |
|---|
The mathematical floor for all-in decisions
The Sklansky-Chubukov rankings answer a specific hypothetical: if your hole cards were face-up and your opponent played perfectly against them, what is the largest stack from which you could still shove all-in profitably from the small blind? The answer is a ranked number for each of the 169 starting hand categories, with larger numbers meaning the hand can shove from deeper stacks. These are not strategy recommendations; they are mathematical proofs about the minimum profitable all-in threshold.
The derivation
The S-C ranking for a hand H is the stack size S (in big blinds) at which the EV of shoving with H face-up is exactly zero against a perfect caller. The perfect caller's calling range is every hand that has equity greater than their pot-odds threshold against H specifically. The equilibrium shove size S satisfies:
EV(shove) = FE * (SB + BB) + (1-FE) * [equity_H * (2S) - (1-equity_H) * S] = 0
where FE is the fraction of hands that fold to the shove (hands below the equity threshold), and equity_H is the average equity of H against the calling range. Solving for S gives the S-C ranking.
For the strongest hand, AA: the opponent folds nothing (every hand has positive equity calling against a known AA), so FE = 0. The ranking for AA is effectively infinite, meaning you can shove AA from any depth profitably (though the calculation technically produces a finite value because of split-pot considerations with specific hands). For weak off-suit hands like 72o, the opponent calls extremely thinly and the S-C number is very small, around 1-2bb.
Worked example
KJo has an S-C ranking of approximately 12-13bb. This means: even if your opponent knows you hold KJo, shoving from up to 12bb is profitable against a perfect caller. Below that threshold, any shove with KJo from the SB versus BB is mathematically sound regardless of what the opponent knows. This is the baseline. In reality, since cards are hidden, KJo becomes profitable to shove from even deeper stacks because the opponent's calling range is calibrated against your whole range, not the specific hand.
Compare to T9s, which ranks around 14bb. The suitedness adds approximately 2bb of shove-depth range over the offsuit version. Flush draws and backdoor equity against known bluff-catchers give suited hands meaningful additional value even when you're exposed.
What each output means
S-C ranking per hand is the maximum stack depth (in bb) from which that hand can shove profitably against a perfect opponent in the SB vs. BB matchup. Sorted table ranks all 169 hand categories from highest (AA, then KK, etc.) to lowest (72o, 32o). Button approximation halves the SB values as a rough guide to button shove thresholds, accounting for the additional SB fold equity. Interactive derivation shows, for any hand you click, the specific calling range that produces the threshold and the equity calculation at that threshold.
S-C vs. Nash push ranges
The Sklansky-Chubukov number and the Nash equilibrium push range answer different questions. S-C asks: "What's the deepest stack from which this specific hand profits against a perfect caller?" Nash asks: "What set of hands should I push with, given that my opponent is also playing their optimal strategy against my unknown range?" The Nash range is usually wider than S-C thresholds because range-mixing provides additional equity that face-up calculations miss.
A hand with an S-C ranking of 10bb might be a Nash shove from 12-13bb when the opponent can't isolate it and must construct a range-based calling strategy. This means S-C is the conservative floor, not the ceiling. You can often profitably shove deeper than S-C suggests, once you account for the information advantage of holding cards face-down.
Where the model breaks
S-C was developed without antes. Modern tournament structures, particularly those with big-blind antes (which can equal or exceed the big blind itself), substantially increase the pot that a shove is stealing. This makes profitable shove depths considerably deeper than S-C rankings indicate. A hand with an S-C ranking of 12bb may be profitable to shove from 18-20bb in a heavily anted structure. The ante toggle in the calculator adjusts the rankings accordingly.
ICM invalidates S-C near the bubble and final table, for the same reason it invalidates any chip-EV calculation: losing chips costs more than winning chips earns when pay jumps are nearby. A hand that is S-C-profitable at 10bb from a pure chip-EV perspective might be correctly folded near the money if the ICM cost of busting outweighs the chip-EV gain from the shove. For a head-to-head comparison of chip-EV and ICM thresholds, the pushbot trainer's ICM mode shows both side-by-side. The ICM deal calculator handles the full multi-player prize equity calculation.