Positional Strategy Simulator
Poker’s most reliable edge isn’t a secret hand — it’s where you sit. This simulates thousands of hands to show bb/100 (big blinds won per 100 hands) by seat.
Why the blinds lose: the small and big blind post money before seeing their cards, and act first on every postflop street — the worst combination of forced investment and least information.
Why the button wins: the button (and seats near it) act last preflop and postflop, see more information before committing chips, and can enter pots with a wider range of hands profitably.
Flat vs. positional range: the “flat” model uses the same hand-strength threshold to enter a pot regardless of seat. The “positional” model tightens that threshold in early position and loosens it on the button — closer to how strategy is actually taught.
The postflop edge toggle adds a simplified “in-position steal” effect: in heads-up pots, the player who acts last on later streets wins some uncontested pots purely from position, independent of hand strength. Toggling it on isolates how much of the button’s edge comes from acting last, not just entering more pots.
Position is worth less than you think, until it isn’t
“Play tight early, loose on the button” is the most repeated advice in poker, and this simulator exists to find out where that edge actually comes from. The answer is not where most people assume.
Turn both layers off — a flat entry range for every seat, no postflop adjustment — and the button stops being special. At 6-max, the best non-blind seat under that model is UTG at 18.12 bb/100, with the button on 17.78. If everyone plays the same cards and nothing is credited for acting last, the seat you sit in barely matters.
Now switch the postflop edge on. The button jumps by +17.01 bb/100 while the first seat to act moves by -5.59. That single layer is what makes position pay — not the cards you choose to enter with, but what it is worth to see everyone else act before you decide.
This is the whole reason the model is built in layers rather than as one number. A simulator that only ever showed the finished figure would tell you the button earns most, which is true and useless. Separating the two effects says why, and the why is the part that transfers to a table.
The blinds are the other half of the story
Every variant on this page has the blinds losing, and losing heavily — BB is the worst seat at -44.75 bb/100 under the settings above. That is not a flaw in the model, it is what a blind is: money in the pot before you have seen anything, posted by someone who then has to act first for the rest of the hand.
It is also the arithmetic check that the simulation is sound. Poker is zero-sum, so every seat’s winnings have to come from another seat, and the figures add to roughly zero across the table by construction. When you see the button winning, you are looking at money that came from the blinds.
What this model assumes, and where it stops
A single raise size, entry decided by a Chen-score threshold per seat, and real five-card showdowns for the pots that get there. What it does not have is betting on later streets, bluffing, or any opponent who adjusts to you — the three things that make real poker hard. The postflop edge is applied as a modelled layer precisely because simulating actual postflop play properly is a far larger problem than this page is trying to solve.
So read the shape rather than the absolute numbers: which seats gain, which lose, and what happens when you turn a layer on or off. The bb/100 figures are internally consistent and not a prediction of your own results. Full assumptions on the methodology page, and the dataset and its checks are listed on the data page.