Position and Playing OOP

What position costs, in chips

Position is not a feeling either, and unlike almost everything else in this camp it is not something the cards can tell you — our rule layer refuses to run without being told it, rather than guessing. Here is the part of it that is pure arithmetic.

You hold J♠T♥7♦6♣ on 9♥8♦2♣, 20 outs and 14 nut outs, and you want to see both remaining cards — that is the 69.7% you learned to compute. Pot is 100 and your opponent bets the pot on every street.

• Flop: he bets 100, you call. Pot 300.
• Turn: he bets 300, you call. Pot 900.
• You paid 400 — 4 times the pot you started with — to see two cards.

In position, against the same opponent checking, that same pair of cards costs 0.

Rule of the game: no bet may exceed the pot, which is why this ceiling exists at all — 400 is the most this sequence can ever cost, and each individual call was priced at 33.3%.

Every amount above came out of the legal-bet calculator js/core/plo-pot.js, the same one behind the pot-odds and SPR lessons, whose formula was checked against three independent public rule pages (docs/M2-SOURCES.md, section T20).

Out of position, treat "seeing the next card" as a purchase with a price, and price the whole sequence before the first call — not one street at a time.

When your opponent is betting every street. It is an upper bound on the cost, not a prediction that he will.

The cost is not a property of the hand. Same cards, same board, same opponent, same equity: 400 out of position and 0 in position. Anything you conclude about "is this draw worth continuing with" that does not mention which seat you are in has skipped the largest term.

Checking is a line, not a surrender

The previous lesson measured what sizes do. Read the same table backwards and it tells you when the answer is not a size at all.

On A♠K♦Q♥ (nut shift 18.4%) the fraction of hands that can profitably call goes 9.2% at the smallest size, 8.3% at half pot, 8.3% at a full pot. You have three sizes and one outcome. On T♦9♦5♣ (nut shift 75.5%) the same three choices give 46.9% and 19.1% at the ends — there the size is the decision.

When the callers barely thin out as you size up, the lever you are pulling is the wrong one. Choose which hands to bet instead of how much to bet, and let the rest check.

Static boards, heads up, and especially out of position, where a check keeps a hand playable on later streets instead of committing you now.

Do not carry it to dynamic boards. On T♦9♦5♣ sizing up removes more than half the callers (46.9% to 19.1%), so checking there gives away something real. And K♠7♦2♣ shows the third case: the largest size still leaves 0% of hands able to call, meaning betting big costs you nothing and gains you nothing — the reason to check there is about which of your hands wants a bigger pot, not about protection.

Acting last is not the same as being protected

Position is usually taught as "you act last, so you have more information". True, and incomplete — here is a spot where acting last is the expensive seat.

3 players, pot 100. The first player bets 100. The second raises to 400. It is on you, and you have put nothing in on this street: you owe 400 into a pot that will be 1,000, so you need 40.0%. The original bettor, facing exactly the same raise, needs 33.3% — the pot-limit ceiling — because his earlier 100 is already working for him.

Out of position, decide how far this hand is going before you decide what to do on this street. In position you are allowed to do it the other way round.

Heads up, and it holds because pot-limit caps every street, which is what makes the whole sequence priceable in advance.

Being last to speak protects nobody once a third player is involved: the seat with the best information in the hand above is the one being asked for 40.0% instead of 33.3%. Information and price are different currencies.

One more claim you will hear, and how we are handing it to you: that position is worth more in Omaha than in hold'em. This course gives you that as reasoning, not as a measurement. We have never measured the hold'em side of that comparison in this project, so we are not going to quote a number for it, and you should be suspicious of anyone who does without saying how they measured. What we did measure is on the first page of this lesson: the same draw costs 400 or 0 depending on where you sit.

What to take away

• Position has a price tag. Two pot-sized calls to see both cards is 400 into a pot of a quarter that size — 4 pots — against 0 in position.
• Price the sequence, not the street. Out of position you commit to the plan before you get the information.
• Checking is a line. When sizing up barely thins the field (9.2% to 8.3%), the decision was never about the size.
• Last to act is not protected. In a three-handed pot the last seat was the one asked for 40.0% instead of 33.3%.

All of it is a rule of this course, not an equilibrium solution — and one claim on the previous page is explicitly reasoning rather than measurement, which we said so out loud rather than dressing it up in a percentage.

One limit belongs on this page rather than in a document: the drills for this course cannot test position. Our rule layer requires you to state it and refuses to assume one — but no threshold in this book actually varies with it. Flip the seat on any spot and every verdict comes back identical, because separating "you are out of position" from "your hand is simply worse here" needs postflop range data and this project has none. The requirement exists so that the day a position-aware rule does arrive, no caller is already quietly passing a wrong default. Until that day the questions below measure price and commitment arithmetic, and the seat printed in each scene is scene-setting, not part of the answer. There is no date on it: it needs the same missing data as range advantage. What this page can give you is what position costs in chips, which is measurable — rather than a threshold dressed up as one.

Next: what changes when there are more than two of you.