Tuesday, June 01, 2010

UCLA Poker Panel and Another Poker Riddle

Last week, as I mentioned, I went to see Chris Ferguson and Bill Chen at UCLA. The event was entertaining, as the panelists recounted anecdotes about Chris and joked about how he well-known he is. I had been hoping it would be more informative, perhaps with some online poker-data analysis or even a mathematical proof or two. However, nothing even approached that level, and it seemed like nobody there had ever even tried looking at online data. There was some discussion of the state of poker game theory and whether poker computer programs would ever be able to beat the best human players. Ferguson and Chen agreed that heads-up limit hold'em might be solved within ten years, and that computer programs might already be as good as people (only in heads-up limit hold'em), but it would be very hard to judge who is "best" unless many hundreds of thousands of hands were played. Chris's dad added that poker in general is unsolvable, since with more than two players, you can expect more than one equilibrium point (a well-known, general result of game theory). The event seemed to be at least partly for promotional purposes, as it was put online and Bill Chen made frequent book references, while the backdrop was initially a Full Tilt advertisement.

At the reception afterward, I didn't see Chris Ferguson, but I spoke a bit with Chris's dad and a bit more with Bill Chen. I asked Bill a little about his book, and jokingly asked when the sequel would be coming out. He said that they might be writing an example book where they look at 100 hands and analyze them with game theory. Sounds interesting. In order to slip away from my group of 3-4 people, Chen left us with the following riddle: if you hold 9333 in Omaha, what are 3 ways you can make the nuts? One way is to have 999 on the board (without a bigger quads or straight flush possible, of course). The other two are harder to see... as a group it took three of us a few minutes to figure them out, but none of us are Omaha players.

Here is the video of the event. I don't recommend viewing it, but I think it's about time for this blog to become a multimedia experience. I didn't watch most of it, but it seems to begin in the middle of Professor Tom Ferguson's tribute to his son.


Tilt said...

Tricky, I had to google the answer. I got the 22223 on my own, but the royal flush on the board was very difficult to figure it out.

Craig Berger said...

I got them, but needed two big hints. The first was "the 9 can be the same suit as the 3," which enabled me to get one pretty quickly, and the second was "a similar idea where it looks like the board should play." What's nice about the problem is that the answer he gives you, 999xy, is a completely different idea, so its a nice way to throw potential guessers off the track.