PKO Strategy: Why Covering the Table Changes Your Preflop Ranges

Work through one progressive knockout final table and you can watch bounties rewrite a preflop range in real time. What follows is a single solved example scenario, not a universal chart. Every number below belongs to this one spot, and the point is the reasoning it exposes rather than the frequencies themselves, because those move with the structure in front of you.

The example comes out of GTO Ranges+: one solved PKO final table, one decision node, read three different ways and then compared against the same decision with no bounties in play. In this particular spot the bounty on the opener's head is worth more than four times the next pay jump, and that ratio is what pulls the button's range apart from the chart you would use in a regular tournament.

The example scenario

Eight players at a final table, blinds at 100k/50k with a 10k ante, stacks running from 9.9bb up to 49.9bb. In this solve's currency, first place pays 957 and eighth pays 142.5, and the eight bounties on the table run from 108 up to 540. Prizes and bounties are quoted in the same units here, which is what lets us compare them directly. Your own tournament will have a different spread, so treat the table below as the shape of the problem rather than a set of values to memorise.

This is a standard progressive structure: half of each bounty you win is paid immediately, half is added to your own head. In this scenario the pay jumps are 43 from eighth to seventh and 55 from seventh to sixth, so the average bounty here pays roughly four pay jumps in cash for winning a single pot. That ratio is the number to work out at your own table, because it is what decides how much the rest of this applies.

The decision: UTG opens, folds to you

In this scenario UTG opens to 2.1bb and the action folds around. Three seats face the identical decision with different stacks and, crucially, different players left behind them. Here is what the solve does in each seat.

The button plays almost twice as many hands as the hijack, and the way it adds them is the interesting part. The hijack 3-bets slightly more often than it calls. The cutoff calls a little more than it 3-bets. The button calls twice as often as it 3-bets. The extra volume arrives almost entirely as flats, which runs against the instinct that flatting an open in position is a leak and you should be 3-betting or folding.

The mechanism is coverage, not aggression

Count who is left behind each seat with a bigger stack:

• Hijack (39.9bb): covered by both the cutoff and the button. Two players behind can end its tournament.
• Cutoff (44.9bb): covered by the button only. One.
• Button (49.9bb): covered by nobody. It is the biggest stack at the table.

In this solve the VPIP gradient tracks that count exactly, which is the clue to what is driving it. This is the same asymmetry covered in covering vs covered, sharpened by bounties. The button cannot be eliminated this hand, so its downside is chips rather than tournament life. Everyone else's chips carry the usual ICM tax on stacking off.

Flatting also does something a 3-bet cannot. Calling leaves both blinds live behind it at 19.9bb and 24.9bb, exactly the depths where a squeeze is often a jam. When they jam, the button plays for their bounty with a stack that covers them. A 3-bet folds those hands out before the money can go in.

The two ranges, side by side

Here is the button's full range in the PKO solve next to the same decision at a chip-EV table with equal 40bb stacks and no bounties. Both grids are decoded from the solve files, so all 169 hand classes carry real frequencies rather than a sample.

Source: GTO Ranges+. Left: PKO final table, 8-handed, button 49.9bb facing a 2.1bb UTG open. Right: chip EV, 8-handed, equal 40bb stacks, button facing a 2.2bb UTG open. Hover any cell for its exact action mix.

One caveat worth stating plainly. The PKO button is 49.9bb and the chip-EV reference is 40bb, and deeper stacks flat a little more in general. That difference pushes toward more calling, so it can explain part of the calling gap. It cannot explain the 3-betting gap, which moves the other way: the PKO button 3-bets 8.1% against the chip-EV button's 5.8%.

What actually changes, hand by hand

Three patterns come out of the comparison.

The button plays hands it would otherwise fold, and 3-bets them. K4s and K3s are pure folds in the chip-EV solve. In the PKO solve they enter the pot 86% and 72% of the time, and when they do they are 3-bets far more often than calls. Q8s goes from 17% played to 75%. K5s from 42% to 100%. These are hands with a blocker and enough playability to continue, and they are being used to build a pot the button is happy to get all-in with against a stack it covers. If you want the chip-EV version of that logic, the same blocker reasoning drives our 3-bet and 4-bet ranges article.

The middle of the range stops mixing and becomes a pure call. In the chip-EV solve, 99 3-bets 12% of the time, 88 17%, AQo 22%, and the suited connectors from 54s up to 87s all carry a 15% to 25% 3-betting frequency. In the PKO solve every one of those hands calls 100% of the time. Nothing in that group raises at all.

The top of the range stops mixing too, in the other direction. AKo 3-bets 100% in the PKO solve against 64% at chip EV, where it calls the other 36%. QQ goes from 82% to a pure 3-bet.

Put together, the PKO button runs a cleaner split than the chip-EV button. Premiums and blocker hands raise, everything else that continues just calls, and most of the mixing that fills the middle of a chip-EV strategy disappears. The bounty gives the button two clear plans instead of a blend: build a pot it wants to play for stacks, or keep it small and keep the covered blinds involved.

How to use this at the table

• Work out the bounty-to-pay-jump ratio before the level starts. If busting the average stack pays several pay jumps, the bounty dominates and your ranges widen. On a real money bubble with small bounties it does not, and normal ICM caution applies. The ICM Calculator gives you the pay jump side in seconds.
• Check who covers you before you widen. The button's licence comes from being uncoverable. Take the same range to the hijack with two bigger stacks behind and it becomes the loosest player at the table with the most to lose.
• Flat more in position when you cover the blinds. Keeping short covered stacks in the pot behind you is the point, not an accident. This is the opposite of the standard advice to deny equity by raising.
• Do not widen your calls against a shove from a stack that covers you. The bounty you cannot win does nothing for you, and the elimination risk is real. ICM still governs that side of the decision.

The takeaway

One spot cannot give you a chart, but it can give you a method. In this example the same open drew a 13.4% response from the hijack and a 24.3% response from the button, with the difference arriving almost entirely as flat calls that kept two covered blinds in the pot behind. What transfers to your own table is not those two numbers: it is the order of operations. Price the bounties against the pay jumps, work out who covers whom, and let the ranges follow from that instead of from a chart built for a tournament without bounties.

The PKO, near-bubble and final-table range libraries this article is built on live in GTO Ranges+, alongside the chip-EV charts to compare them against. If you want the postflop half of the same tournament, Postflop+ puts you in the spot and grades the decision.