what do you mean exactly by doing an ICM calculation. In other words. How do you derive the bold numbers.

Disregard. Got it

still got another question about the number of future knockouts. Say stack x will collect 4.0 future knockouts. How is this converted to a $ value. Just 4 times the average $ bounty values of the remaining players?

so if player 1 has a bounty of: 1$
players 2: 2$
players 3: 3$

The avg bounty is 2$. So if a stack will collect 4 future bounties, this corresponds to 4 x 2 = 8$?

and if the remaining regular prize pool was 100$. Then a value of 8/100 is added to the regular ICM eq to get the equity including bounties?

We currently only support regular knockouts with equal bounties for all players. Support for progressive KO is planned and should be ready within a month or so.

The first part is correct. I'm not entirely sure if I understand the last bit correctly. The value added by future bounties is calculated in the way you quoted above, simply multiply the # of expected knockouts with the $ per bounty. So in this case the total regular prizes would be 100$, and the total bounties 8$.

The old equity mode would display equities in relation to the total regular prize pool, so the equity on that table would sum up to 108% when including bounty equity. The new "table equity" mode shows the total remaining prizes of 108$ as 100%. (That setting is available at Window: Preferences: Equity Display, "table equity" is the default in the current beta.)

Thanks!

What i meant by the last part. Say, we are facing a shove of a player and the ICM value of our stack is we call is 0.20. And if the total prizepool (without bounties) is 100$, this
stack is worth 20$. Now assume that according to your model we get another 8$ worth of bounties. So infact our stack is actually worth 28$.

So now the icm value when we call becomes: 0.20 + 8/100 = 0.28.

which corresponds with a stack worth of 28$ ( 0.28 * original prize pool = 28$).

Does this make sense? (in other words i leave the prizepool the same, and just give players more % of this cause of knockouts)

Yes, those figures would be in line with the "old" equity display, which is displayed relative to the total regular prize pool.

Note that the total equity for folding is also affected by future knockouts, so you would need to adjust that value accordingly for the EV calculation. You can't simply use the 0.2 figure, that only accounts for the regular prize pool and would considerably underestimate the actual equity.

(I'll probably post further details and maybe some examples on the site when the feature is ready for the stable version.)

Yeah makes sense. Thanks! Really looking forward for the PSKO support

It's tough to calc but that's a really good thing for SnG ecosystem just imo. Recs seem to prefer KO to non KO and regs make more mistakes if they can't just jam/fold the ranges they learned from practicing

im curious what your model for PSKO will be. Will you still be calculating a number of future knockouts for each player, and then give each of the remaining bounty a weight to indicate how much this bounty will contribute ( cause the bounty on the short stack is more likely to earned then the bounty of the chip leader)

or will you calculate a probability for each players bounty to be earned for each finishing position?