Push Bot Charts and Calling Ranges for Donkaments

[jcm4ccc]

I searched through hundreds of $12 turbo tournaments (PS) and calculated the following calling range when you are in push-fold mode. This is the calling range that the average player will use when you are pushing 10M or less in the typical low buy-in tournament:

<ul type="square">[*]100%: JJ+,AKs,AKo[*]77%: TT-55,AQs-A9s,KQs,AQo-AJo[*]31%: 44-22,A8s,KJs-KTs,JTs,ATo-A8o,KJo+[*]16%: A7s,QTs+,T9s,98s,A7o,KTo,QJo[*]6%: A6s-A2s,K9s-K7s,Q9s,J9s-J8s,T8s,A6o-A2o,QTo,J9o+,T8o+,98o[/list]
What this means is that, if you push and your opponent has, for instance, cards within the following range (TT-55,AQs-A9s,KQs,AQo-AJo), there is a 77% chance that they will call you.

Below I describe how I derived all of this, and how you can use it to create push-bot charts.

<u> Method </u>

What I did was to search hand histories for times when a player pushed for 3 – 10M in an unopened pot, and was called by someone who was not short stacked. I noted the hands that the player called (not what the player pushed with, but what they called with). Here is the data for JJ – AA, AKo, AKs:

<ul type="square">[*]AA: 19[*]KK: 14[*]QQ: 22[*]JJ: 21[*]AKs: 7[*]AKo: 30[/list]
My assumption was that nobody ever folded these hands. These hands, on average, occur 40 times every 1326 hands. I counted 113 instances of these hands, so I think it’s a fair assumption that, in the histories that I looked at, a total of 3746 opportunities occurred (113*1326/40) in which someone might have called the preflop pusher. Using that number, it’s easy to calculate the number of times someone actually called with a particular hand and what that represents in percentages.

For example, each pair should occur 17 times for every 3746 hands. Here’s what I actually counted, along with the percentage of the expected total number of times someone was holding this hand and could have called:

<ul type="square">[*]AA: 19 (112%)[*]KK: 14 (83%)[*]QQ: 22 (130%)[*]JJ: 21 (124%)[*]TT: 16 (94%)[*]99: 15 (88%)[*]88: 11 (65%)[*]77: 15 (89%)[*]66: 12 (71%)[*]55: 13 (77%)[*]44: 8 (47%)[*]33: 4 (24%)[*]22: 4 (24%)[/list]
Then I lumped similar percentages and came up with the calling range shown above

<u> push-bot charts </u>

This data can obviously be used (along with pokerstove data) to create push-bot charts for low –limit tournaments where you have no knowledge of your opponents and their calling tendencies. I have created these charts, an example of which is shown below:



K8s<ul type="square">[*]SB: 10[*]Button:10[*]CO: 9.2[*]4th from the BB: 6.5[*] 5th from the BB: 4.9[*] 6th from the BB: 3.8[*] 7th from the BB: 3.1[*] 8th from the BB: 2.6[/list]
This indicates that, for example, if you are holding K8s in the 4th position and have less than 6.5M, you can profitably push.

You can create your own push-bot charts using the data above and pokerstove. I’ll share the charts that I created (and the Excel worksheet used to create the chart) for anyone willing to help me gather more data to make these charts even more precise. You can cut out 25 hand histories from low level tournaments in which someone pushed for 3-10M and someone who was not short stacked called. Just cut out the first 25 that you run across in your hand histories of Poker Stars tournaments and send them to me at jmcclintock@nc.rr.com. Or, if you’d rather not bother with cutting out hand histories but still want the charts and data, you can transfer $12 (one donkament buy-in) to my PS account (jcm4ccc) and send me an email, and I’ll send the charts and data along. I’ll also send updates as I create them.