In this post, I’m going to describe my analysis of Sylvain Loosli that I did prior to and during the 2013 WSOP Main Event final table. I had been hired by Amir Lehavot to analyze his opponents and himself for possible behavioral leaks. The point of this post is not just to talk about how I went about trying to find patterns in Loosli and what those patterns may have been, it’s also to talk about the difficulty in trying to do this kind of analysis.
There are two main problems with trying to do this kind of analysis.
The first main problem is the small sample size of available hands to study. Pre-final-table, I perused all the existing WSOP footage on the players and took notes on all the significant spots. Then I compared and contrasted these hands, looking for possible patterns that could be meaningful. But there are very few hands shown for any of the November 9ers pre-final-table. The hands they do show can take a long time, and they have to spread out their coverage over many players.
The other problem is editing and the limitations of television coverage. For instance, there is editing of time, making any analysis of bet-timing tells of questionable value. There is also the fact that what the coverage shows is at the whim of the TV director. Many times, I’d like to see what a person is doing immediately after betting (where they look, how they move, etc.) but the camera cuts to the person whose decision it is.
These are the two main problems even before the final table begins. The fact that the final table is played months after the majority of the tournament is another difficulty. No matter how sure I might have been of my reads prior to the final table (which wasn’t sure at all for the reasons discussed above), I still had to go into the final table with mostly a blank slate, knowing that these players had been preparing for the final table, and may have altered their strategy and behavior drastically. For instance, even though Loosli was very new to live play when he played the bulk of the WSOP and displayed a lot of behavioral variety in the footage I’d seen, he had had months to prepare and months to formulate a better behavioral strategy (which I believe he did.)
Finally, there are so few hands played even during the final table that it is hard to develop much of a read over that time, even after the fact, when I’ve had more time to look at the footage. For instance, in Loosli’s case, over the course of the 7 hours of televised final table play, there were only a couple of 3-bets w/ very weak hands, and only one river spot where he was betting a significant amount while quite weak (the hand where he was betting 33, which wasn’t even obviously a bluff but more of a range-merge where he maybe could have gotten a call from Ace high). As is typical for most tournament poker, there just weren’t that many significant big bets post-flop; most spots were pre-flop maneuvers where hand strength isn’t well-defined or were small flop or turn bets or calls, and so it’s hard to establish much in the way of good comparison points.
All of these difficulties were known by Lehavot and me going into this venture; it’s not like either he or I were super-enthusiastic about the idea or thought that we’d find some great information that would alter the course of the game. It was more like we went into it thinking that finding a few decent patterns was possible and, if found, could theoretically play a role. It was kind of an experiment.
So now I’ll go into more detail about how I analyzed Sylvain Loosli, but you should keep in mind that I am far from certain on all of these reads. I was more certain of my reads on Farber (part of this was due to him being more of a recreational player and part of it had to do with seeing more of him due to him playing a long heads-up battle), and I might write a similar breakdown of those, too, but I wanted to write about Loosli because he was more of a challenge and I thought it’d be more educational for me to do a post-game analysis of him.
Pre-final table footage
Here’s a video where I compiled all the pre-final-table hands I thought were significant from Loosli. I’ll reference this video below, so you can see what I’m talking about. (In the notes of the YouTube video, you can see the specific hands that are included in the compilation.)
First, my general impression after watching all of this footage was that Loosli showed a lot of behavioral variability. Much more so than any of the other November 9ers. Some examples of the kind of behavioral variation I’m talking about:
- Taking a long time to count out chips and make a bet, and making a lot of movements while he did that
- Putting his hand on his head and face at various times while in a hand
- Shuffling chips and sometimes flipping chips (turning chips end over end on the table) before/after making a bet
- Announcing bet-sizes and raises verbally
- Lot of body movement in general during/after betting
Knowing that Loosli was new to live poker, it makes sense that you’d see him doing more things that you wouldn’t see other, more stoic players doing. Before I even started really looking at the video, my general theory going in is that if I see a lot of behavioral variation, there’s going to be more stoicness/stillness in those spots where Loosli has a weak hand, and more movement/looseness in those spots where he’s got stronger hands.
While knowing that I obviously had a small sample size, I think this general theory held up pretty well in the pre-final-table footage. Here are a few things I saw that I noted as possible patterns that were worthy of further analysis.
Hand on head/face seemed correlated to good hands:
- At 2:00 in the video, when he has AA and gets in a raising war with Morgenstern, he has his hand under his chin for much of this.
- At 2:48 in the video, when he has the full house versus Newhouse and bets the river, he has his hand on his mouth.
- At 5:40, after betting his set of Kings and getting called by Benefield, he puts his hand up on his head for a few moments.
- At 6:30, where Loosli is calling with A6 (2 pair) on the flop, he has his hand on the side of his face.
- At 8:38, with AJ on a JT6 flop vs Brummelhuis, he has his hand on the side of his head.
If you look at the weaker hands, you’ll see that he doesn’t put his hand on his head in any of those. (I’ll discuss the final table after this, but it’s worth noting I only saw him put his hand on his head once during the final table, and it was when he had limped in w/ AA.)
Extended stacking of chips after announcing bet/raise possibly correlated with strength
- 3:25 Full house vs Newhouse, when he 3-bets river, he’s cutting out chips in multiple stacks for quite a long time, as opposed to just stating the amount and being still.
- 5:50 With the set of Kings on the river, after he bets, he’s putting the bet together in stacks for a long time up until Benefield calls.
This is actually a fairly common tell for a lot of people, when they announce a bet or all-in and then voluntarily start to stack up the chips for an extended period of time. You can compare this to Loosli’s weaker bets and bluffs, where he doesn’t do this. (Again small sample size, but it was something to keep in mind for later.)
Verbalizing bets/raises possibly correlated to strength
A very small sample size, but there are several hands where Loosli verbalizes his bet or raise with a strong hand, whereas this is less common with his weaker hands. Also, when he does verbalize, it’s usually with a questioning, upward-inflected tone, like, “2.5?”. I wondered if the questioning tone could be more correlated toward strength. I didn’t feel that these were necessarily important; more of just something to keep in mind for later.
Extended length of time preparing a bet or raise
This was fairly unique to Loosli amongst the November 9; the amount of time he would take to stack up chips and make a bet or raise. My general assumption was that he’d tend to take longer stacking chips up when strong and be more quick and more direct in putting chips in when making significant bets that were weaker or bluffing, so that was something I’d want to look at more at the final table.
Also, small pauses were worth looking for. An example would be when he’s betting his set of Ks vs Benefield on the river at 5:50. He puts his hands down on the table, which is basically announcing he’s going to bet, but then waits several seconds before announcing a bet.
Another example in the same hand, at 4:20, when he gets ready to 3-bet; he drops a chip on his cards and then seems to wait a few seconds before gathering chips (hard to tell with possible editing).
If you look at the weaker hands in this compilation, even though he might take a while to gather chips, he puts them in more straightforwardly, without any non-productive pauses.
Small extraneous movements seemed correlated to good hands
- 2:00 Holding AA, adjusting glasses after being raised by Morgenstern, before re-raising.
- 4:45 After 3-betting pre-flop w/ KK, Loosli makes a good amount of movement and facial movement here after Benefield calls
- 5:22 After betting the turn with his set of Ks, his right hand slides back over the felt and makes some loose movements in a way I wondered we’d see if he was weak
- 7:30 After betting his rivered trip 7s, he has that same long slide of his hand over the felt after betting
- 8:38 As he says “all in” when 3-betting with his AJ on the JT6 flop vs Brummelhuis, his hand brushes in front of his mouth.
While Loosli has other small, extraneous movements in other, weaker hands, my general theory was that he would be less likely to have these movements after making significant bets with hands that were weak or a bluff.
Chip flipping possibly associated with good hands
As a general rule, “chip flipping,” when someone flips a chip or chips end over end while in the act of betting or right after it, is associated with high hand strength. This can vary a lot, as most things do, person to person, but I start out with that assumption. It was just something I’d keep an eye out for later.
- 1:08 Betting his flopped straight, Loosli does a little chip flip before reaching for his chips
- 3:40 3-betting his full house on the river vs Newhouse, Loosli does some chip flipping as he announces his bet.
- 8:38 Before 3-betting on the JT6 flop w/ his AJ vs Brummelhuis, he does some chip flipping.
- 13:00 It’s possible he’s flipping his chips a little here after bluff-raising the flop with 9T on an AK9 flop, although it’s hard to tell if it’s riffling or flipping because it’s partially covered.
It’s worth studying the two big bluff bets from Loosli in this compilation; the hand where he puts his opponent all in on the turn w/ QJ on a 9J2K 3-club board, and the hand where he bluff-raises the flop w/ 9T on an AK9 board in a multiway pot. In both of these hands, Loosli is mostly still during after betting. There is a little bit of movement, like after shoving that turn, when he places a chip on his cards, or when he looks over at Ortiz for a moment as Ortiz thinks. I watched these hands a good amount, just internalizing what Loosli might look like when he’s betting while vulnerable.
Final table footage
Here’s a compilation of final table footage that I’ll reference below. I chose only to include the strong and weak 3-bets pre-flop because I felt like the standard raises were too normal and standard to expect to learn much from.
Knowing that Loosli had several months to prepare, is a smart guy, and has a lot of well-known poker friends, I was assuming he’d have a much different demeanor on the final table. So while I had a list of things to look for, I was prepared to basically start from scratch and know that he might be much more stoic than before.
As the game started, it was obvious he was doing some things differently.
For one thing, he no longer riffled his chips during a hand. He used to riffle his chips pretty constantly. So maybe someone told him to stop that.
For another, he was wearing a scarf, so it’s possible someone told him he was giving away information with swallowing or his heartbeat.
Also, he stopped verbalizing his bets or raises at all, that I saw. This was probably another thing he figured out wasn’t helping him. (On the other hand, this could be a contributing factor to his bet-amount mistakes in this match, if he is used to verbalizing his bet amounts.)
He also seemed more cognizant of being stoic, and seemed to have a standard pose he’d return to, with one arm on the rail and the other sticking out forward onto the table. Only once did I see him put his hand up on his head, and that was the hand where he limped with AA. And overall he had much less extraneous hand and arm movement, or movement over all.
He was, however, taking a very long time to gather chips and put them in the pot; even longer than I’d seen him take pre-final-table. He also did a lot of holding the chip stacks sideways so he could look at them. It was kind of a weird and unique way to gather chips; most people do not take nearly as long as he does, nor with as much movements. My reads on him were kind of thrown off because I at first thought that the long time gathering chips was related to strength, but he was capable of taking a long time to gather his chips even when weak (although in the small sample size I have it does seem like his stronger 3-bets do take a little longer).
Returning quickly to standard pose after betting
What I started keying into more was how quickly after making a big bet did he return to his standard stoic pose. For example, it seemed like if he was betting a strong hand, there was a tendency to allow himself some extra movement before returning to his pose. He’d gather chips, or shift around, for a few more seconds. Whereas if he was making a significant bet with a weaker hand, he’d be more likely to go to his standard pose quite quickly.
As examples of this, watch the first three hands in the compilation, where he 3-bets with QQ, KK, and KK. Watch after he puts in the bet, how much movement he has. Then compare this to the couple weak 3-bets where he returns a little more quickly to his pose.
In the post-flop situations where he was betting with a weak hand, he also seems to return more quickly to his standard pose. It’s pretty hard to reach much of a conclusion on this, though, because there just really aren’t many post-flop hands. Again, all of this is a pretty small sample size.
He does have a few occasions where he does some chip flipping with strong hands. He does it right before 3-betting with QQ. He also does it right before 3-betting with KK (the first one). He also does it once when c-betting the flop with AQ high.
Pauses in chip gathering seem heavily correlated to strength
After having more time to study the footage, I do think Loosli was more likely to have pauses and stalls in his chip gathering when he actually had a hand, and take longer in gathering chips in those cases. A few examples of this:
- 00:40 With QQ, after he puts a chip on his cards and it’s evident he’s going to play, he just pauses for a couple seconds, flipping chips, before he starts to put chips together.
- 1:35 With KK, after he puts a chip on his cards and it’s evident he’s going to play, he again just pauses for a couple seconds, flipping chips.
- 2:30 Again with KK here, he puts a chip on his cards, so we know he’s going to play, then he just sits there for almost ten seconds not doing anything.
- 5:30 After limping with AA and being checked to on the flop, he puts his hand down from his face and looks down at his chips, so it’s evident he’s going to bet. Then he just shuffles chips for a while before placing the bet.
If you compare these spots with the spots where he is vulnerable or bluffing, you’ll see that while he may take a while to gather and place bets in those spots, he is much more straightforward and is engaged in “productive” actions the whole time, without these “unproductive” pauses. Of course, there are times when he straightforwardly makes a bet with strong hands, too, but my point is that I believe any sort of stalling or hesitation in the process of making a bet is heavily weighted towards value hands. This can also include a lot of the very extraneous movements Loosli makes when he is value-betting, such as the long, drawn-out process of adding chips one at a time to his bet amount before putting them in.
The above theory makes sense if we assume that players with strong hands want to give a sense of having to make a hard decision to bet or raise(even if this is an unconscious desire they have to give this impression).
Considering the sample size, none of this is very convincing. I would have loved to have seen more hands with Loosli, just to see how my reads progressed and to test their accuracy. I do feel pretty confident about the stalls/hesitation in his chip-gathering process meaning strength, but unfortunately there weren’t a lot of hands to compare.
How would this information have best been used? If I trusted my read of how Loosli was putting his 3-bets together pre-flop, I could imagine 4-betting him lightly if I felt like his chip-gathering was more straightforward, without pauses, whereas if I saw a lot of pauses or hesitation I would be more inclined to fold or call pre-flop. Same with post-flop situations, if I saw evidence of hesitation/stalling in the process of him betting, I’d be more likely to give him credit for a hand.
Same thing with how quickly he returned to his standard pose after betting or raising. If I saw him make a bet or raise and then saw him quite quickly go to his standard, stoic pose, I would be less likely to give him credit for a hand. This could make the difference in deciding to 3-bet or 4-bet pre-flop, or make a raise or call post-flop.
I’m reminded of talking to Lehavot at the beginning of starting to work for him on this. He seemed to think that it was possible I didn’t think highly enough of the power of poker tells. He said something to the effect that when it comes to very good players who are well-balanced and tough to play, a tiny amount of information could go a long way in making your mind up on borderline decisions. I did agree with him, and said that some very good players had told me that that was how they found the most use of poker tells; by using slight bits of information to sway decisions that were fairly close one way or the other. I think of that now and think that these slight reads I had on Loosli, although probably not huge game changers (unless they turned out to be more accurate with more of a sample size) could make the difference in the long-term by giving a little bit more information with which to make close decisions.
I also feel fairly confident some of my old reads on Loosli were accurate. I wish I could see more footage of him from back then. I think if there wasn’t such a big gap between most of the tournament and the final table I think my old reads would have come more into play. It’s obvious that Loosli dramatically improved his unreadability during that time, but I still think he had a few behavioral leaks. Whether or not they were the ones I talked about in this post, I’m far from certain.