A couple weeks before the WSOP Main Event final table, I got an email from Amir Lehavot (founder of PokerWit.com and 3rd in this year’s WSOP ME), asking if I’d be interested in helping him prepare for the final table. After discussing what I might be able to do to help him, and making sure we had similar expectations, I agreed. My role was to analyze existing footage of his opponents for possible behavioral patterns, and to be present for a final table simulation that Lehavot was doing, and look for possible patterns he might have. During the final table, I was also watching the 15-minute-delay broadcast in a hotel room, weighing in occasionally with my thoughts through Skype.
I plan on writing a more detailed article on this experience. I’ll just give a few thoughts on it for now.
Going in, Amir and I both agreed that it was a long shot that there was much in the way of poker tells I’d find on his opponents (for a couple reasons which I’ll address in a second). Amir appreciated that I had the same opinion he did on this, and I think respected the fact that I wasn’t super-enthusiastic about the idea or super-confident that I would be able to help him. I expressed concern that he was already such a good player that focusing on physical tells so close to the final table might be a hindrance to his performance. I also wasn’t sure I was the best person for the job. (I had a horrible vision of me giving him a read that he acted on that was horribly wrong and busted him.)
But we did agree that it could have theoretical value if I did find some good tendencies, and it was obvious to me that Amir was a smart enough guy to not let my opinions sway him without thinking about the ideas himself first. Also, regarding me not being the best person for the job, it occurred to me that even though there are a lot of very experienced players who’d be better at analyzing poker tells than I, those players might not have been as enthusiastic about the job as I was, or perhaps would be much more expensive than I was, or perhaps wouldn’t have been free at that moment to work on such a thing. All of these things made me feel better that maybe I was the right guy for the job after all.
The main problem with trying to find reliable reads on his opponents prior to the final table was twofold: 1) most of them were quite experienced players (Jay Farber was the only non-poker-professional) and 2) there wasn’t much existing footage to make good comparisons or reach much in the way of good conclusions beforehand. For example, even the most amateur player, Farber, only was shown a few times in pre-final table footage, and pretty much all of his significant spots were ones where he had strong hands. There were only a few hands shown for most of the players; not enough to establish good reads. (One exception was Ryan Riess, whose 2012 WSOPC Hammond final table, was online; all 9 hours of it. But his style and demeanor had changed a lot since then, so it was of questionable value. For the other players who had past footage of them you could find, like Newhouse, Tran, and Benefield, not only was it quite old footage, but they were very experienced live players, which made finding good reads on them unlikely.) Anyway, long story short on this topic, with a small sample size of hands, and most of the players being smart and skilled, it was a tough assignment.
I will say that it was very educational to try to think of the best way to approach this challenge. Trying to think of the best way to structure and organize the information I had, and to try to communicate it in clear ways to Amir without just rambling on about possibilities. Luckily I’d already had some experience in logging televised poker footage for my own notes and database that I keep, so it wasn’t like I was starting from scratch. But it definitely required a new way to do some things, and it was a learning process. (Amir and I have a conversation planned in future to go over ways that it could have been done better, and what things he liked about it.)
During the final table, I took notes as I watched the broadcast. While I started out pretty rough, by the end I felt like I had pretty good reads on both Farber and Loosli, with a few very minor reads on the other players. (I will go into more detail on these reads in future.) As it turned out, due to a couple bad spots, Amir was short stacked most of the game, so my reads were basically useless to him. I believe it could have been different; for example, if he had gotten a bigger stack and gotten in more post-flop situations, my opinions could have come into play. (This would have especially been the case if he’d gotten short-handed and fairly deep with Farber or Loosli.)
As for Amir’s own poker tells, I assumed going in that he was probably giving very little away. I was present at a final table simulation he did a few days before the real final table, with several good and well-known players there, including Carlos Mortensen, Matt Stout, Steve Gross, Ryan Welch, and James Carroll. Unfortunately the simulation was a fairly informal practice session, with nothing really at stake, so it did not replicate real conditions or real high-stakes behavior. Still, I had a couple small opinions of stuff that could theoretically have been imbalanced in Amir’s actions.
All in all, it was a learning experience in a lot of ways, and I feel not only am I better equipped for such an assignment in the future, I also, in focusing so much on these specific players, got better at understanding the range of behavior that can be present and at picking out what is potentially useful information more quickly.