Brag/Beat Watch - Will my poker study get funded?

[DING-DONG YO]

Are all you guys 18? I don't want to show up at someone's parents house and run into Chris Hansen.

"Look, I just wanted to play poker with him. that's it, I SWEAR!!"

[T50_Omaha8]

Which professor you looking at to mentor this...Keenan?

And by upper level micro-econ you mean 4000 level? Cause 4010 was a joke... Who knows maybe the 8xxx grad courses are harder.

BTW I came very close to doing poker for my game theory project (w/ Keenan), backwards inducing from the last betting round to demontrate fixed expected values each player has before a hand is dealt. The girl I was working with didn't seem enthralled with it though.

[T50_Omaha8]

[ QUOTE ]
[ QUOTE ]
again, i call bullsht on the poker bots being amazing at LHE, names of "pros" please.

[/ QUOTE ]

It was a vague recollection, I'll try to go look one of the studies up, the bots are all LHE but i thought i remembered one where they did really well.

I'd be willing to host the game. BEAT: It'd beat at my parents house. Still, I could probably host. I'd definitely want to set it up.

[/ QUOTE ]
In the zoo there's a massive thread about the $5/10 and $10/20 bots on Party that ran near 1bb/100 I think. That's damned impressive for a bot at those stakes. I think it was in October or so.

[SkeetyMcdoogle]

econ sucks and if you think you are smarter than me come study biomedical engineering you graph making GDP finding mother Fer

[DING-DONG YO]

lol professionalstudentsgetarlfackingjob

[cheer]

[ QUOTE ]
[ QUOTE ]
maybe you would like it less if you actually bought the books and read them

[/ QUOTE ]

I lollerskated. If you think you like econ, take an upper level micro-econ class. If your brain still works afterward, then we can talk. I alternately love econ so much and hate it so much.

[/ QUOTE ]

MicroEcon and Behaviors is my bitch. I am 2nd in my class of 40. (3001 Micro @ the UofM)

[DannyOcean_]

i had a mildly insane micro-econ teacher... I did pretty well in the class but half the people dropped. I'm doing it with keenan T50, correcto.

DDY- We're all 18.

SkeetyMcDoogle- I do think I'm smarter than you, but it has nothing to do with my major. lol at arrogant science majors.

[DannyOcean_]

No email today, and the time for leaving the office has passed. I'm guessing they lied about their deadline. Typical of the honors program here to be off deadline. I wouldnt be pissed except that I'm about to go on spring break so what if I don't know for another week and a half?

grrrrrrrrrrrrr.

oh well

[venom007]

Whats all this talk about people being from atlanta, i live in decatur.

[MooBot3000]

[ QUOTE ]
again, i call bullsht on the poker bots being amazing at LHE, names of "pros" please.

[/ QUOTE ]

That's why I said "pretty good." From my understanding (which comes from reading their journal papers a few years ago) they only had bots that could beat novices. Any mid-stakes online LHE winner could beat up their bot.

Making a good poker bot is inherently harder than other types of games because it's in incomplete information game. For example, chess has all the information available to the computer, namely the current board state, and then the computer can crunch away at different scenerios (the more powerful the computer the more scenarios it can search over) it then scores them all according to some evaluation criteria and picks the move that has the best score. Poker you can't do this because the complete game information isn't known (namely what cards your opponents have). You can mold the poker problem into a similar framework as the chess problem by searching over all possible card holdings of your opponents, but as we know each holding isn't equally as likely. So the crux of poker bot making is in the opponent modeling (i.e. how to distribute weights on each poker hand so that the more likely holdings have higher weight...). In other words, they have to learn how to do conditional EV calculations given the opponents. This problem as we know, most humans even struggle with (especially in NLH) so why should we think it would be easy for us to write a bot that could do it for us? The poker problem isn't a "brute force helps us" kind of problem. This is why I don't forsee pokerbots beating the best "pros" anytime in the near future.