Our next installment of viewer mail:

Q: How soon before poker bots get sophisticated enough to spoil the games?

A: The last I heard, they had a LIMIT hold em program that could beat humans: http://www.stoxpoker.com/man_vs_machine.html. I'm pretty sure I could come up with a winning SHORT STACK algorithm for no limit holdem - which means it's probably out there. STTs and MTTs are somewhat extensions of the short stacking game. There is analysis software out there now for STTs and MTTs that give you ev analysis for every hand played. I don't see why something similar couldn't be incorporated into a playing bot. Limit holdem, short stacking and tournaments are particularly attractive to bots because the downside to making a wrong play is limited.

NO LIMIT holdem cash games are a little more challenging because of the larger initial stack sizes. Even still, I think we're pretty close (if not there already) to having a machine play winning no limit hold em.  I don't worry about it for a few reasons:
* The poker sites are allegedly vigilant about looking for cheaters, and it should be in their best long term interests to do so.
* I don't think a bot would crush the games. Even if it was technologically possible, I think they would prefer to stay under the radar.  So I don't know that having a bot at the table would be that much worse than having a decent winning reg at the table.
* I actually think it's possible to beat a bot playing "against" your stats (i.e. playing differently from what your stats would indicate).  But I haven't completely thought this out.

FWIW, I have a buddy who used to be part of a chess club in Manhattan.  Last year, I met a few of his chess club buddies a few times and they seemed to be preoccupied with programming online poker bots.  There's no money in chess - plenty of money in poker.  They seemed to be cognizant of possibly doing the data processing and actual playing from separate machines, so that detection may be more difficult.  I would also be surprised if in every advanced computer science class in the country, there wasn't someone (if not the whole class) working on poker algorithms or bots for a project.  For this and many other reasons, the game gets harder every year.

Q:  Will the NLHE games die out from over-analysis before this ever happens?

A:  I DON'T think the NLHE games will die out for a two reasons:

* There will always be fish (assuming no adverse regulation).  How do I know?  The same reason casinos are constantly popping up and state lotteries are booming.  People love to gamble.  It's just part of human nature - it's exciting and fun (when you win).  If there are hordes of people who like throwing $100 in a slot machine when their ev is -8%, or people who pay $10 for lottery tickets when their ev is -50%, I don't see how playing a virtually zero sum game against other human competitors won't be more appealing.

* Poker is a zero sum game with costs.  The "game dying out" scenario requires that everyone starts becoming equally skilled, so the only one making money is the casino.  However, I have no reason to think this would be true.  While it may be that it would take a greater amount of skill to make the same money five years from now, it's not clear to me that the top players 5 years from now won't be making more money than the top players are now.  It's like how in sports people get better every year, but the top players still clearly make more than the rest of the pack.  I suppose it is possible that everyone's play will approach what is considered game-theory optimal, it's not clear to me that game theory optimal can't be beaten by creative players - again, I need to think about it some more.

I'd be interested to hear what other players think.