PKO and Mystery Bounty Math: Pricing the Bounty Into Every All-In

In bounty formats, nominal equity is only half the equation. Learn to price a bounty into required equity so you stop folding your way out of the most profitable spots.

Bounty tournaments are where solid players quietly leak the most edge — not because the postflop is harder, but because they keep evaluating all-ins with the wrong number. They look at a shove, run the chip-EV in their head, see they're a touch short of the price, and fold. In a freezeout that fold is correct. In a PKO where you cover the shover, that same fold is often a sizable mistake, because there's cash sitting on top of the pot that never entered the math.

A bounty is extra cash equity you collect only when you eliminate a covered opponent. That single sentence contains the entire strategic adjustment. The job is to turn that cash into chip-equivalent equity and add it to the pot you're contesting — so your required equity to call drops. Get the conversion right and you'll call shoves that look marginal and print. Get it wrong (or ignore it) and you'll fold the most profitable spots in the format.

This piece works through the math: how PKO money is split, how to convert a bounty into equity points, a numeric example showing how much a bounty actually lowers your calling threshold, why short stacks with bounties become magnets, and the genuine tension between bounty-EV and ICM survival near pay jumps. Then mystery bounty, which is the same engine with a probabilistic payout.

How the money is split in a PKO

In a standard Progressive Knockout, your buy-in splits in half. One half feeds the regular prize pool (the payout ladder you climb by surviving and finishing high). The other half becomes your bounty — the cash sitting on your head that anyone who eliminates you collects.

The "progressive" part is the wrinkle that makes the format. When you knock someone out, you don't pocket their entire bounty. Typically half of the bounty you win is paid to you immediately as cash, and the other half is added to your own head, growing the bounty other players are now chasing. So eliminations compound: a player who busts three opponents early can be carrying a bounty several times the starting size, and every future knockout they make is bigger.

That progressive mechanic matters for two reasons:

Immediate cash is locked in. The half you collect on a knockout is real money in your account, independent of where you finish. It doesn't care about ICM, pay jumps, or whether you bust the next hand.
The half added to your head is future equity. It raises your own bounty, which is partly an asset (you'll carry it if you win more) and partly a target painted on your back.

Contrast this with non-progressive / Bovada-style flat bounties, where the entire bounty you win is paid as cash and your own head value stays fixed at the starting amount. Flat bounties are simpler — there's no compounding, no growing target — but the core call/shove adjustment is identical: a bounty is cash you win only when you cover and eliminate.

The core adjustment: bounty as equity points

Here's the mechanism that should run in your head on every all-in where a bounty is live.

Normally, to call an all-in profitably by chip-EV, you need enough equity to justify the chips you're risking versus the pot you're winning. Standard pot-odds: if you call X to win a pot of P (including your call), you need roughly X / P equity.

A bounty changes the prize. When you cover the shover, winning the hand wins you the pot plus the bounty cash. The bounty is extra reward that arrives only on the win-and-eliminate branch. So it functions like adding chips to the pot — but only chips you collect when you knock the player out.

To make the bounty commensurable with chips, convert it using the chip-EV-to-prize ratio: roughly how much one tournament chip is worth in cash at this stage. If the average stack is, say, 40,000 chips and the average cash equity of a stack is some dollar figure, you can back out "what is X dollars of bounty worth in chips?" In practice you don't need to be precise — you need the order of magnitude. The cleaner way to think about it operationally:

Express the bounty as a fraction of the pot. If the pot you're playing for is worth, in chips, some cash value C, and the bounty is worth B in cash, then the bounty is "B/C extra pot" — and that fraction is what lowers your required equity.

Why does it lower required equity? Because the reward on your winning branch went up while the cost of calling stayed the same. More reward for the same risk means you can win less often and still break even.

A useful approximation for the adjusted threshold:

Required equity ≈ (chips risked) / (chips in pot + bounty-in-chips), with the bounty only counted on the branch where you win AND eliminate.

Strictly, the bounty attaches only to the "you win and the covered short stack is eliminated" outcome — in a clean heads-up all-in where you cover, that's the same as "you win the hand," so you can fold it directly into the pot. In multiway spots it's messier (more on that below), but heads-up against a covered shover, the bounty just inflates the pot.

A worked example

Let's price one out.

Setup. Mid-stage PKO. A short stack open-shoves for 10 big blinds from the small blind. You're in the big blind, you cover comfortably, and the action is heads-up. Antes and blinds make the math, but to keep it clean: you're calling 9 bb more to win a pot that — after their 10 bb shove, your 1 bb already posted, plus blinds/antes — totals roughly 22 bb.

Chip-EV only (no bounty). You're risking 9 to win 22.

Required equity ≈ 9 / 22 ≈ 41%.

So in a freezeout, you need about 41% equity to call. Against a 10 bb SB shoving range, plenty of hands clear that — but a hand like K9o, Q9s, small offsuit aces, weak suited kings sit right around the threshold, and many players fold them.

Now add the bounty. Suppose the short stack's bounty, in cash, converts to roughly the value of 8 bb of chips at this stage. (You can get this conversion from a bounty/ICM tool — the point is the order of magnitude: a meaningful but not enormous bounty.) Because you cover and it's heads-up, the bounty attaches cleanly to your winning branch, so it inflates the effective pot:

Effective pot ≈ 22 (chips) + 8 (bounty-in-chips) ≈ 30.

You're still risking 9 chips. Your required equity becomes:

Required equity ≈ 9 / 30 ≈ 30%.

The bounty just moved your calling threshold from ~41% to ~30% — about 11 equity points.

That is enormous. Eleven points is the difference between folding the bottom third of your calling range and calling it profitably. Hands you'd muck in a freezeout — K7s, Q8s, J9o, A2o, small pairs you were worried were flips at best — become clear calls, because against a 10 bb shoving range almost everything has 30%+ equity. In fact, with a bounty that size, you're often calling any two cards mathematically, since even 32o has roughly 30%+ against a wide shove.

The size of the swing scales with the bounty relative to the pot. A bounty worth 8 bb against a 22 bb pot is huge. The same bounty against a 120 bb pot (deeper stacks, bigger all-in) barely moves the needle — maybe a couple of equity points. The shallower the all-in relative to the bounty, the more the bounty dominates. That's the single most important intuition in the format.

Why short stacks with bounties are targets

The example explains the dynamic directly. The cheaper it is to put a covered bounty at risk, the more the bounty distorts your price. So:

Short stacks carry disproportionate bounty value. A 10 bb stack's bounty might be worth nearly as much as a 60 bb stack's bounty (in a flat format, exactly as much; in progressive, it depends on knockouts), but it costs you far fewer chips to contest. Bounty-per-chip-risked is highest against short stacks.
You want to isolate them. If a short stack with a juicy bounty shoves and you can over-call or re-shove to get heads-up, you deny everyone else the chance to win that bounty. Isolating to play the bounty heads-up is worth real equity, because the bounty is a winner-take-all prize and you want to be the only one with a shot at it. This is why, in PKOs, you'll see players over-shove to isolate covered short stacks with hands they'd never get involved with in a freezeout.
Covering is everything. None of this applies if you don't cover. If the short stack has you covered, you can't win their bounty by eliminating them — there's nothing to collect, because you'd be the one eliminated if you lose, and if you win you've merely doubled, not knocked out. A bounty only lowers your required equity when you cover the opponent. Misapplying the discount to spots where you're the shorter stack is a classic and expensive error. When you're covered, play the spot as chip-EV (and ICM) — the bounty is irrelevant to your call.

The tension: bounty-EV vs ICM survival

Everything above maximizes bounty-EV = chip-EV + immediate cash from knockouts. Early and mid-tournament, when survival is cheap and the payout ladder is far away, bounty-EV is close to the right objective and you should hunt aggressively.

But bounties don't repeal ICM. As you approach a pay jump — the money bubble, a final-table ladder, a satellite-style flat structure — busting costs you real prize-pool equity, and that cost can exceed the bounty you're chasing.

The honest framing: near a pay jump, the bounty partially offsets ICM risk but does not fully override it.

Far from pay jumps: weight bounty-EV heavily. Call and isolate liberally against covered short stacks. The ~11-point discount from the example is close to fully realizable.
Near a meaningful pay jump: the chips you risk are worth more than their raw count (the ICM tax), which raises your required equity, while the bounty lowers it. These pull in opposite directions. The net adjustment might be small — the bounty buys back some of the ICM premium, but not all of it, so your real threshold lands between the freezeout number and the full bounty-discounted number.
On a hard bubble with a big bounty: you can be in a spot where the bounty fully justifies a call by chip-EV-plus-cash, yet ICM still says fold because busting before the bubble forfeits a near-certain min-cash. The bounty is real cash, but it can't bring back the pay jump you torched.

The practical rule: the bounty discount is largest exactly when ICM pressure is lowest, and shrinks as ICM pressure grows. Don't apply the full mid-game discount on the bubble, and don't apply bubble-tight ICM ranges in level six when survival is nearly free.

This is precisely where running the numbers beats feel. shadepoker's ICM Calculator let you price the bounty into required equity and weigh it against ICM survival in the same spot, so you can see whether the cash on the head actually buys back the ICM tax or just dents it.

Multiway and the progressive head

Two complications worth flagging.

Multiway pots. When more than one player can contest a covered bounty, the bounty no longer attaches cleanly to "you win the hand." You only collect it on the branch where you are the one who eliminates the short stack — if a third player wins the pot and knocks them out, you get nothing. So in multiway spots, count the bounty at a discount proportional to how often you're the eliminator, not the full value. This is another reason isolating is valuable: going heads-up converts a discounted multiway bounty into a full-value heads-up bounty.

The progressive head. In a true PKO, half of every bounty you win lands on your own head. That has a subtle effect: it makes you a bigger target, which slightly reduces the fold equity of your future shoves (people call wider to win your inflated bounty) and means others get the same discount against you that you get against them. It's a second-order effect — don't let it stop you collecting bounties — but it explains why deep-stacked bounty leaders get shoved on relentlessly. Carrying a big head is mostly an asset, but it's an asset that invites action.

Mystery bounty: same engine, probabilistic payout

Mystery bounty reformats the bounty as a lottery. You knock someone out, and instead of a fixed amount you draw a sealed bounty from a pool — most are small, a few are large, and one or two are life-changing jackpots. Crucially, in most mystery formats there's a bounty phase: bounties "activate" at a certain point (often the money bubble or a set level), and only knockouts after activation pay out a draw.

The math adjustment is clean: the EV of a knockout is the average bounty value across the envelope distribution. If the pool is, say, $1,000,000 spread over 1,000 envelopes, each knockout is worth $1,000 in expectation regardless of the jackpot tail. You price that expected value into your required equity exactly as in the PKO example — convert the average bounty to chip-equivalent, add it to the pot on the win-and-eliminate branch, recompute the threshold.

Two practical notes specific to mystery:

Use the mean, not the dream. The jackpot tail is psychologically loud but the expected draw is what's mathematically real. Players overcall before activation and underrate the steady EV of average draws after it. Price the average.
Activation timing reshapes ranges. Just before bounties activate, there's often less reason to gamble for knockouts (the bounty isn't live yet) and more ICM pressure if activation coincides with the money bubble. Right after activation, every covered short stack becomes a target carrying the full average bounty — the discount from the example snaps on at once, and calling and isolating ranges should widen accordingly.

The same progressive-vs-flat distinction applies. PokerStars-style mystery formats may layer mystery draws on top of a progressive structure; Bovada-style and many live mystery events use flat draws with no head growth. The conversion is identical — average bounty in, chip-equivalent out — only the compounding differs.

The takeaway

In bounty formats, nominal equity is only half the equation. The other half is the cash you collect when you cover an opponent and eliminate them, converted into equity points and added to the pot. Skip that conversion and you'll fold hands that are clearly profitable; the shallower the all-in relative to the bounty, the bigger the mistake.

The discipline is mechanical:

Do I cover? If not, the bounty is irrelevant — play chip-EV and ICM.
Convert the bounty to chip-equivalent (average bounty in mystery formats) and add it to the pot on the win-and-eliminate branch.
Recompute required equity — it drops, often by many points against shallow shoves.
Check ICM. Near a pay jump, the bounty partially offsets the ICM tax but doesn't override it. Far from pay jumps, take the full discount.
Isolate covered short stacks to convert a multiway, discounted bounty into a full-value heads-up one.

Misprice the bounty in either direction — ignoring it when you cover, or chasing it through an ICM wall on the bubble — and you bleed EV. Price it correctly and the bounty stops being a vague "bonus for knockouts" and becomes what it actually is: a precise, quantifiable reduction in the equity you need to stack chips and cash.