Online poker’s meritocracy has been a wonderful thing: how the hardest-working and most skilled players, over a long enough time period to reduce short-term variance, almost inevitably rise to the top. But as the industry matures, this close correspondence between talent and results is in danger. Many players point the finger of blame at the wealth of videos, books, and other instructional material on the market today. In this article, I’ll argue that other factors, such as legislation, unfair rake schemes, and the sites’ increasing market power, are at fault.

Although there are now more ways than ever to learn about poker, this trend is nothing new. Herbert Yardley’s The Education of a Poker Player is one of the earliest strategy books, dating back to the 1950s. There are now books and videos which cover nearly every form of poker commonly played and skill levels have risen dramatically.

Though, it’s not clear whether the books that are available always represent the state of the art strategies that the best players use. Where once a novice could read only one book and expect the strategies inside to provide them with a sizeable edge over weak opposition, now such a player might have to spend a great deal of time learning a much wider range of skills. Just playing tight might not cut it alone. In fact, you could argue that selling these educational materials is an important way for the best players to capitalize on their skills. These players might not be able to always get sufficient action at their desired stakes, so becoming a coach or an author is just another way to profit from their knowledge.

Furthermore, this probably misses the most important aspect of learning: imitation. Players have always learned by observing and adopting the strategies of those who seem to win the most. It happens in every game, and the cumulative effects are probably more significant than all other learning put together. This is also something that we’re powerless to stop.

Online-specific aspects of poker, such as database programs and heads-up displays (HUDs) are also often blamed for the worsening state of online poker. Although these tools help winning players beat the fish faster, and perhaps help discourage some fish from re-depositing, this misses some important points. The cumulative aspect of the rake, taking money out of the online poker economy with every hand played, shouldn’t be underestimated. Banning HUDs would help the fish lose slower, but it would also mean that more of their losses end up in the hands of the sites. It might be the case that removing these tools from online poker would help none of the professional players if any increased deposits end up in the rake. These tools are in fact part of the skill element of online poker, and skill is the only thing that keeps some players consistently winning. At the extreme, removing all skill aspects of online poker would reduce it to other forms of gambling, such as roulette, where nobody wins in the long run. Aspects of the game such as learning cutting-edge strategy and using these programs more efficiently than others are what keeps poker a profitable career for the best players. We wouldn’t want to change that.

Legislation has been a major factor in recent years, with some countries segregating their player pools and restricting them to regulated sites. Other countries enacted full-scale bans. Not only does this prevent many losing players from accessing the sites, but it also disrupts the player pool on other sites as skilled players are much more likely to leave their country of origin to continue playing.

When a country segregates its player pool, this endangers poker’s meritocracy in two ways. A professional might start winning more not because they’re one of the best at their form of poker, but because they’re lucky enough to live in a country with below-average players. This disrupts the close correspondence between skill and winnings which makes poker fair. Segregation also almost inevitably leads to higher rakes, either in the form of new government taxes, or through regulated sites realizing that they no longer need to keep rakes down to stay competitive.

At low levels, I’d argue that less generous rake schemes are the major problem that online poker faces. Actual rakes haven’t changed much, but effective rakes have greatly increased as bonus, rakeback, and incentive schemes have been scaled back. I remember being able to receive roughly 100% rakeback on one site playing $1/$2 limit hold ‘em some years back. When the rake is so high compared to the betting limits, this makes a real difference, compared to the roughly 50% which only a very high volume grinder might be able to receive in those games today. Also, the effective burden of the rake is much higher now that skill levels are generally higher and before-rake winrates subsequently lower. A combination of high rake and the need to fund living costs is surely preventing a number of skilled players from moving up and achieving their full potential.

Rake is a much less significant problem at high stakes, but there are still problems at these levels of the game. These games will always surely always get harder due to the high rewards on offer and the weak worldwide economy, but even so the status quo seems problematic. These games are facing a collective action problem, where nobody wants to play unless a known fish is in the game. This means that many fish who’d be happy to jump into a running game are disincentivized from joining. Furthermore, when these games do run they often end as soon as the fish leaves or sits out, making their predatory nature even more obvious. Unfortunately, this problem is very difficult for the professionals at these limits to tackle directly, as any costs of starting or maintaining a game are localized while the benefits are spread among many. It’s so easy to free ride on the good work of others that any attempt to tackle the problem is bound to unravel. It’s a classic tragedy of the commons. As I’ve argued elsewhere, I believe that this problem could be most easily addressed if the sites gave players the proper incentives to provide the common good of starting and maintaining games.

At the end of the day, the sites have to realize that online poker is fundamentally changing, and that they should recognize these changes to return online poker to its early days. Although there’s nothing they can do to directly address legislative issues, there are many ways that they can make the game fairer, restoring the association between skill and earnings. This might not seem to be in their short-term interest, as it would surely lead to lower rake collections, especially at small stakes. But it may help their long-term profits, as having professionals starting games and playing at all levels could lead to more rake being collected over time. As I’ve tried to argue here, pointing the finger of blame at books and HUDs is missing more fundamental issues. The danger is that, as there’s now less competition between sites due to legislation and shrinking player pools (increasing the market power of the few biggest sites), they might ignore this opportunity to improve the common lot.

Philip Newall is the author of The Intelligent Poker Player, and you can follow him on twitter.

A lot of people, even experienced players, struggle with how to think about poker tells and how to incorporate them into their game. A lot of this struggle comes from not understanding some basic poker tell concepts. Many times, it’s this confusion that leads to doubt or ambivalence about the usefulness of tells. Even if you’re a player who already uses tells successfully, misunderstanding these basic concepts can prevent you from using tells optimally.

Many of the concepts and terms discussed below are ones I’ve had to invent while working on blog posts and my book, Reading Poker Tells. There simply wasn’t a resource that discussed poker tells from a rigorous scientific or statistical standpoint like I wanted to do. (Some of my concepts, though, may be similar to those used in behavioral science.) While I’ve had to make some simplifications in studying poker tells (hand strength is one of the most necessary simplifications), thinking about tells in this rigorous way has led to a more logical understanding of how tells work, which has in turn improved my reads.

Situation is Important

There are many situations in poker, and some of them are very different from others. Just as a check from a first-to-act pre-flop raiser on the flop means something completely different from a check from a player last-to-act on the river, tells are best understood by trying to put them into similar categories so that they may be better understood.

When writing my book, the most important situational categories I came up with were “waiting-for-action,” “during-action,” and “post-bet.” A player’s “waiting-for-action” behavior is what he exhibits when waiting for an opponent or opponents to act. A player’s “during-action” behavior is what he exhibits when it’s his turn before he acts. A player’s “post-bet” behavior is what he exhibits after making a bet. 

A player can have a lot of variation between his “waiting-for-action” behavior and his “post-bet” behavior. To make a long story short, this is primarily due to the increased emotional tension that making a bet and being studied can create for a player, compared to the more passive, more defensive emotions sometimes present when a player is waiting for an opponent or opponents to act.

For example, let’s say there’s a player with a strong hand who is “waiting-for-action,” and who doesn’t look at his opponent at all. After he makes a bet with his strong hand, however, he engages in a lot of eye contact with his opponent. If you didn’t realize this pattern of his and how the situation impacts his behavior, you could draw the wrong conclusions about how to use this information.

There are many players who have tells that are much more reliable in either the waiting-for-action situation or the post-bet situation. Two of the main ways people can vary their behavior in these situations is in how much eye contact they make with their opponent and in how loose or still their body movements are.

Tells Are Most Obvious When Accompanying Significant Actions

Many people, when first trying to study tells, make the mistake of trying to look for tells everywhere, regardless of the situation or the size of the bet. They make the mistake of comparing how a player acts when he makes a standard pre-flop raise to how that player acts when he’s making a large river bet.

When I study most emotion-based poker tells (to separate them from the more passive, lazy tells), I’m trying to compare behavior that accompanies significant actions. This means I don’t usually spend much time studying how a player makes a standard pre-flop raise or a standard continuation bet. I will only start to analyze a player’s behavior when the action gets significant, like when someone is putting in a large turn or river bet, or moving all-in. That’s when your mental database should start “recording”, so to speak.

If you are observing people too much in non-important spots, you will just tire yourself out and collect weak, insignificant data for the most part. Plus, you’re just distracting yourself from gathering the much more important information about your opponents’ fundamental playing styles.

Reliability and Frequency

To illustrate the concepts of reliability and frequency, I’ll use a famous tell that many people are familiar with: the Teddy KGB Oreo tell from the movie Rounders. (I know this is a ridiculous, unrealistic tell, but it’s a useful one for illustrative purposes.) This was the tell where KGB cracked open an Oreo and ate it when he had a strong hand. This tell was exhibited as a waiting-for-action tell, meaning KGB did it before betting or raising.

The term reliability has been used in a general way to indicate the usefulness of a poker tell. But I’ve never seen an exact description of how one might determine that, and I wanted to define it more accurately. The equation I came up with is:

Reliability of a poker tell = # of times behavior is displayed with specific hand strength in specific situation  /  Total # of times behavior is displayed in that situation

Let’s say we wanted to determine the reliability of KGB’s Oreo tell as a waiting-for-action tell of high hand strength, we would have:

Reliability of KGB’s Oreo-eating tell = Number of times Oreo-eating behavior was exhibited with a strong hand while KGB was waiting for opponent to act  /  Total number of times Oreo-eating behavior was exhibited while KGB was waiting for opponent to act

If we observed KGB performing the Oreo behavior before betting with a strong hand eight times, and we never saw him do it any other time, his tell would be 100% reliable (as far as we know based on observation).

If we observed there were eight times that KGB did his Oreo thing while he was waiting for action with a strong hand, and we observed two times that he did the Oreo thing while he was waiting for action with a weak or middle-strength hand, then the reliability would be 8/10, or 80%.

In the real world, a tell will usually be displayed to some extent when a player has a range of different hand strengths, so a tell will seldom have a reliability of 100%. But even a reliability of 51%, while far from what we’d consider very practical, would still be theoretically enough information to sway borderline decisions (in other words, those decisions where, from a fundamental strategy point of view, two opposite actions have the same expected value.)

Reliability, as I’ve defined it, is a much different thing than the concept of frequency. Frequency refers to how often a behavior is exhibited in a certain situation.

For example, if KGB exhibited his waiting-for-action Oreo tell every single time he had a strong hand, the frequency of this tell would be 100%. If this tell had a frequency of 20%, it would mean KGB was only exhibiting the tell 2 out of 10 times he has a strong hand.

It’s important to differentiate the ideas of reliability and frequency. A tell could have a very low frequency and still be highly reliable. For example, KGB’s waiting-for-action high-hand-strength Oreo tell could have a frequency of 20% and still be 100% reliable, and very useful. Basically this would mean that every time you saw that tell, no matter how infrequently he displayed it, you would know exactly what it meant; that he had a good hand.

Just Because a Tell is Reliable Doesn’t Mean Its Absence Is

Leading directly from the previous concept of differentiating between reliability and frequency, you should realize that just because a tell means something, the lack of that tell doesn’t necessarily mean anything.

Many players get confused by this one. Let me describe how confusion can happen, using the KGB Oreo tell:

Let’s say Matt Damon sees KGB perform his waiting-for-action Oreo tell two times in one hour. Both times, KGB shows a very strong hand. Matt Damon knows the tell means a good hand and that it is “reliable” (although he doesn’t have a good sense of what that word means yet).

A few hands later, Damon sees KGB make a big bet on the river and watches him carefully. KGB does not perform the Oreo tell, so Damon assumes that KGB must have a weak hand. So he makes a light call and is surprised to see KGB has the nuts.

What went wrong? Damon made the assumption that a lot of players make; that if a tell is reliable, it also has a high frequency. But as I explained, a highly reliable tell can still have a very low frequency; this just means that not seeing the tell isn’t important information. The moral of the story is that just because a tell is meaningful, you shouldn’t assume that the lack of the tell is meaningful.

Summary

Some of these concepts might seem very basic, but many players don’t have a good understanding of them. Not understanding these concepts will cause problems, even for experienced players, when it comes to interpreting tells. I think that there is value in internalizing some of these concepts, much like there is value to internalizing some of the basic mathematical concepts of poker strategy.

Zachary Elwood is the author of Reading Poker Tells, a book released in 2012. You can also find more tell and psychology related information on his blog: www.readingpokertells.com.