My name’s Zachary Elwood, and I’ve written some well-known books on poker tells: Reading Poker Tells, Verbal Poker Tells, and Exploiting Poker Tells. I also have a poker tells video series.
I was watching the Cate Hall versus Mike Dentale “grudge match” the other day, and Doug Polk and Shaun Deeb were commentating. Doug and Shaun got into a conversation (57:09 into the stream) about the value of poker tells that went like this (edited slightly for clarity, scroll down a bit to watch for yourself):
Doug: I don’t make big generalizations [about opponent tendencies], because that’s the only way I can really lose to people, is by making big adjustments and it being wrong.
Shaun: Have you ever thought about, maybe, hiring someone for live reads to teach you? Because you’re so good in theory, and you’re good at conversing, you’re such a great personality, I think that just to add that 10% edge to your game.
Doug: I think everyone in live just thinks they have good live reads. And so you’ve just got like a bunch of bullshit. Honestly, I just think that so many live players are just total garbage at poker, that I just don’t want some random garbagio person to be like: “Yo, he like, talked on the turn, and that means this thing.” Because it does mean that thing against some people, it doesn’t against other people.
Shaun: But most live reads are personal.
Doug: I actually think it makes you a worse player to go down that road. I think you become a much worse player because you lose fundamentals and you lose the ability to create a good strategy by focusing on things that are not important in the grand scheme of playing good player.
Shaun: I definitely think you convince yourself to make marginal calls and play some spots and play some bigger pots than are necessary with weaker hands because you trust your instincts based on, you know, timing and physical tells and recognize when they’re middle or bottom of their range.
Doug: The thing is, I see so much when people just guess wrong. That happens so much. I analyze poker hands on YouTube. That’s what I do everyday. And I see just in a plethora of spots, where people just go with their reads. And afterwards we always have the same conversation. Me: ‘This was a bad play for these reasons.’ Them: ‘Oh no, but I had a read that…’ And it’s like, am I supposed to take you seriously? Because you played the hand like an idiot. But it’s like, so you played your hand badly, but because you quote “had a read”, we have to pretend we’re on the same footing.
In case you want to hear this convo for yourself:
A lot of people assume because I’ve done a lot of work on poker tells, that I think poker tells are a very important part of live poker. But this is far from the case: I believe that reading opponent tells is the “icing on the cake,” something to think about long after you’re a solid, winning player.
In other words: I have made a lot of content about poker tells because I find the subject interesting, not because I believe it’s the most important part of poker.
So I can understand what Doug is saying and what his frustrations are. I think the following things are true:
- Most predominantly-live poker players, even professional ones, are not nearly as good at poker as players who have honed their games online.
- Most live reads acted on are probably acted on for bad reasons. In other words, most live reads, even many made by experienced players, are probably not well-justified.
- For the large majority of the poker-playing population, especially the less skilled, time spent learning about poker tells would better be spent on learning fundamental strategy.
Having said all of that, let me clarify where I disagree with Doug and think learning about poker tells can be a good idea.
To start with, we should recognize that Doug’s objections to using physical tells (i.e., they’re sometimes wrong) can just as easily be applied to any exploitative, non-game-theory-optimal approach to poker. I’ll include an excerpt from my last book, Exploiting Poker Tells, here, that deals with that idea:
It’s important to understand: When you decide to incorporate poker tells into your game, your reads will sometimes be wrong. This is completely expected and inevitable. For all exploitative strategies (i.e., strategies aimed at exploiting opponent flaws), your decision will sometimes be wrong in a specific hand.
To give a strategic, non-tell-related comparison: When an opponent has been three-betting pre-flop every single hand, and you make an exploitative decision to four-bet him light, he will sometimes have pocket aces. This doesn’t mean your exploitative decision was wrong. You made your decision because you thought it would be profitable in the long run. You hopefully didn’t expect that it would be correct every time you did it.
In the same way, acting on opponent behavior is a numbers game. If you have a read that is slightly reliable, or even very reliable, it will sometimes be “wrong” in the context of a single hand. This doesn’t necessarily mean that it was a bad read.
I wanted to point this out because I think some people, when they make a read and it turns out to be wrong, get frustrated with trying to use poker tells. This is understandable, because it is frustrating to make a move or a fold based on a read, only to discover you were completely off-base. You’re using a more speculative source of data, so it is understandable that it stings more when you’re wrong. “If I’d only played the hand the normal way,” you might think, “I wouldn’t have messed up.”
But you should remember that, if you’re reading opponents well in general, you will be right more often than you are wrong and that will improve your results. It’s just usually much easier to remember the times you were wrong than the times you were right.
Doug could just as easily have said, “I see so much when people just guess wrong” about non-behavior-related adjustments to an opponent’s frequencies. But I don’t think Doug would object to using the more obvious/demonstrable changes in an opponent’s strategy as basis for decisions (although I think he would probably say that people also base decisions on those things too frequently, too).
For example, I don’t think Doug would object to adjusting a river-bluff range based on an opponent who has been shown in many hands to fold too much on the river. And aside from player-specific history, I also think there are probably certain general lines from recreational-seeming players that Doug would feel okay making exploitative adjustments to (for example, maybe raising or calling unusually small bets lighter than normally). Both of these have analogies for behavior-based decisions: noticing player-specific behavioral patterns, or noticing behaviors that, even without any player history, are usually reliable for recreational players.
As the excerpt above talks about, I think for most players there can be negative emotion around using what is perceived as a more-speculative/less-known source of information. And I would imagine this is especially true when you have become a very good player based on a pure GTO approach to the game.
If we look at physical/verbal tells as a similar thing to perceived changes in opponent frequencies, then I think we have a more realistic view of how tells can be on a spectrum from very obvious and likely to be reliable to less obviously reliable and more ambiguous.
Here’s where I think learning about tells can be very valuable for players:
- Defense. The readers of my books and videos, many of whom are not experienced players, will mostly not make good use of that information in regards to reading opponents. This is simply because many are not experienced enough to use that info and will misuse it or misapply it in many ways. But where those players do see a lot of value in my stuff is in learning how to eliminate their own tells. I’ve had many customers send me messages saying, basically, “I was doing such-and-such a lot until I read about it in your book.” Fixing such leaks probably improve win rates (or decrease loss rates) more than using that info to read others.
- Very obvious tells exist. Most tells are on the ambiguous “This probably means that” side, but obvious tells do exist. My blog posts, books, and videos talk about many of the common somewhat-reliable, more ambiguous patterns, but also a good amount of the super-reliable variety. (Speaking of very reliable tells, I once tweeted “If Jay is bluffing here, I’ll eat my hat” during the no-hole-cards live airing of the 2013 WSOP Main Event, based solely on Farber’s behavior. Ryan Riess made a daring call with Queen high and Farber had the flush.) Even if a tells-skeptical, GTO-focused player like Doug only deviated from his usual strategy when he saw tells that a majority of skilled live players would consider very reliable, that would represent an improvement to his win rate.
- Even less-reliable tells can be used for tough/borderline spots. Assuming you have reason to believe a tell is only slightly reliable, it is still reliable. If you’re in a spot where you could go either way, that information will help you in the long run, even if you’re wrong in specific spots. This may especially be the case when playing strong opponents who play in a balanced way and who often will put you in spots where your decision between calling or folding is close to neutral EV. And this is how many skilled live players have told me tells play a role in their game; when they’re in a spot they think is fundamentally break-even, they may use some small piece of behavioral information (a player’s loose back-and-forth glances at them, an opponent’s breathing, whether an opponent pushed in chips or not when saying “all in”) to make up their mind. Assuming they have good reason to think that behavioral info is reliable (which, I agree with Doug, is something to be skeptical about for many players), then those decisions will be made better in the long run.
So, to summarize: Tells are a very small part of playing strong live poker. And I agree with Doug Polk that many live poker players are wasting a good amount of time and money in 1) thinking too much about poker tells at the expense of thinking more about strategy, and 2) making reads too often based on behavioral information they don’t have good reason to believe will be accurate.
Having said that, I also think that, as long as you have a realistic viewpoint that tells will be used infrequently (especially versus strong players), then I think learning about poker tells will 1) protect you from leaking info, and 2) lead to many +EV spots, even for players who only want to base decisions on the most reliable patterns.
Zachary Elwood is the author of Reading Poker Tells, Verbal Poker Tells, and Exploiting Poker Tells. He has consulted for two WSOP Main Event final tablists and done video analysis for several high-stakes cash and tournament players. His video series is at www.readingpokertells.video.
(Note: Join the 151,389 people that have upgraded their poker skills with Upswing’s preflop charts – for FREE. Click here to get the charts now!)