Poker-AI.org

Hi guys,

My bots were running on party poker for 2,5 years. In the last few months my 11 bots got banned. First all 8, then last month I created a 2 new bots and got banned after 12 days, and 2 weeks ago tried another one and got banned after 11 days. I run party poker client on VM, bot is on OS. I do image scraping, human like mouse movements, random position and timing mouse clicks on the buttons. My bots were having on NL4 and NL10 winrate of 1-2bb/100 hands in the first 2 years on party poker, but when I made a big strategic improvement, they started winning 5-8bb / 100 hands on NL4 and 4-6 on NL10 (never had time to try NL25 and higher because of bann problems). My bots were like 20% -30% of the population, so It might be that party poker decided to bann my bots because they were crashing the NL4 and NL10, making it too tough for the other players, and when they were winning just a little bit, they were no problem (just one assumption). I can hear from friends botters that many people are able to run bots with similar configuration as I have for years without any problem. So in order to run bots again without getting banned I would have to make big changes in mouse movements, clicks, table size, table look (I was replacing table images so I had unique look), in VM configuration in order to just maybe not get banned. That is the reason why I am looking for someone who has bots on party poker, and is willing to take my Decision making dll to run bots on NL4-Nl25 cash games (fast tables and slow tables) for profit share, so that I can focus on strategy.
In the attachment you can find Graph NL4 after 10 days which my last bot played before getting banned.
I am also having bots on ipoker and 888, chinese casinos, if you want to run bots on some other network it is ok for me.

I am working alone on the strategy part, father was helping with ocr and DB. I have own DB for player stats and population stats. I can query from it e.g how often regulars (or fish, tight) are bluffing e.g flop when they bet e.g. 33% pot size (for different bet sizes, board type, number of opponents, position etc). It is also able to adjust if opponent changes strategy, e.g if we bluff overbet turn a lot, and see that opponent called light in last few samples, it is sign for us that we should adjust. I have a simulator where I can define hand scenario (either by copying hand history from pt4, or by using real print screen images, or by defining everything myself) and see the bot ranges for that spot which is very useful for testing purpose (similar like snowie).