Party Bad Beat Jackpot EV

[Gotmilk]

I asked this a while back, and we couldn't figure out an answer, but I found a stat in Pokerstat that might allow as a group to come up with some kind of reasonable guess. If you go into Misc. Stats it gives you showdown % for different types of hands. For example, I have 27,000 hands in my database. I have shown down quads or straight flush 11 times. My showdown percentage is 100%. Looking at the hands, in none of the instances would my victory have triggered the bad beat, so to our group stats I add 27,000 losing hands. If we get a bunch of us to post those stats, we can guesstimate something at least.

[Happy Hour]

I can tell you without looking that in 90,000+ hands I have never had my quads beat (any) quads. I have never had a straight flush beat quads or another straight flush. I have never lost with either.

Edited to add: Also, I have never seen, at any table I've been at, a showdown that would qualify for the bad beat.

[zooey]

I've thought about this too, and I think its nearly impossible to calculate how often a bad beat "should" occur in any game where all 10 players aren't blindly going to showdown.

So here's how I would go about it: if someone would keep a record of how big the bad beat jackpots are when won, eventually one could get a reasonable guess at how often they hit (relative to # hands played) and a ev calc could be made.

The standard dev of jackpot size is large though, so it would take a largish # to converge. If someone could post the size of the jackpot on its last 20 hits, we could get a reasonable guess at the SD and thus how many more we need to watch to be able to do a reasonable EV calc.

best,

zooey

[radek2166]

I have ssen it once in ~60k hands. Not sure what the four of a kind was but it was beat by a st flush.

[Baulucky]

Calculate the odds of quads in holdem and multiply by the odds of str8 flush in holdem. I did it once and promptly went out to buy a lotto ticket because it was better value.

edit: This is NOT exact, but gives you the order of magnitude.

IMO, it'd be easy for Party to rig this up, especially after a huge jackpot builds. I know I would do it, if I were them...Who is going to stop them?.

[Baulucky]

Actually the odds of quads squared is a better estimate...

Still too big a number.

[waffle]

[ QUOTE ]
I asked this a while back, and we couldn't figure out an answer, but I found a stat in Pokerstat that might allow as a group to come up with some kind of reasonable guess. If you go into Misc. Stats it gives you showdown % for different types of hands. For example, I have 27,000 hands in my database. I have shown down quads or straight flush 11 times. My showdown percentage is 100%. Looking at the hands, in none of the instances would my victory have triggered the bad beat, so to our group stats I add 27,000 losing hands. If we get a bunch of us to post those stats, we can guesstimate something at least.

[/ QUOTE ]

I wrote a computer simulation. 10 players in a ring game. Players will play all pocket pairs, any two cards ten or above (like AK, KJs or QT), any suited ace, and all suited connectors except 23s 34s. Puts them at about 20-23% VPIP. Everyone always went to the river. The odds for the BBJ hitting in these conditions are 1 in 200,000 games.

Assume that the BBJ happens 1 in 400,000 games (because people are sometimes folding before the river). Assume that you win 5% of hands played - so you pay 2.5 extra cents per hand of BBJ rake. Finally, given party's payout structure, if you're at a table that hits the BBJ, your equity in the jackpot is 7%.

SO: Jackpot * (.07) >= .025 * 400000

This means given my assumptions, Jackpot needs to be >= about 143,000 to make playing at the tables +EV.

[BlueBear]

An interesting conclusion, does this consider the extra 10% party charges in adminstrative fees when the jackpot is won?

Your conclusion also suggests that playing for the jackpot has positive EV most of the time, because the jackpot is higher than your states figure most of the time.

[Chief911]

You can add my 50k hands where I've never had a jackpot qualifying hand.

Nick

[Awesemo]

It wouldnt be that hard to actually calculate. There are only 3 cases.

Case 1: Quad beats quad. Board must have 2 pair of both 8s or higher, and two people out of 10 must have two exact cards.

Case 2: SF beats quad. Board must have a pair and a 3 straight flush. Depending on the board, one, two, or three sets of cards could make the sf, and one could make quads.

Case 3: Sf beats sf. Board must have 3 consecutive straight flush cards. Two players must have two exact cards.

Through combanitorics this problem is not that difficult. First you figure out the chance that the board could qualify for a bad beat jackpot, and then you figure out the chance that the qualifying hands are held.