Poker is, perhaps somewhat paradoxically, a game of words. We spend our time using words to describe mathematical concepts, and it’s not easy. There are a lot of words that get thrown around in poker without the correct application; there are a lot of words that get used way too much, to the point where their meaning is eroded; and then there are the words that aren’t used often enough, to the point where people overlook their significance. One of the words in this is last category, in my opinion, is the word ‘frequency’. Let me tell you why.
Flaws in the way people discuss hands
I’ve been participating in discussions on poker forums and talking hands with people for around six or seven years now, and I’ve seen a lot of evolutions in the way people talk. But one thing has been fairly constant throughout this time – even once it became more or less second nature for people to talk in terms of ranges, very few people have really yet learned to talk in terms of frequencies.
What I mean by this is that when people state their opponents’ ranges for certain spots, they state them in terms of what they believe the range is or isn’t, without attaching a number to their degree of confidence in that range – in other words, without attaching a frequency. This means they end up operating in a world where everything is either completely binary – 0% or 100% – or it’s a vague, poorly-defined quality described only by an adjective.
How to discuss a hand
Let me give an example. When asked about an opponent’s likely 3-betting range, a player might say something like, “they’re definitely 3-betting 99+ AQ+ here, probably AJs, maybe 88 or even 77…I guess they have quite a lot of bluffs too, maybe some suited aces, suited connectors…something like that”. This is a really non-specific description! What they’re saying here is that there’s a 100% chance they’re 3-betting 99+ AQ+, a 0% chance of 22-66, and most other hands could be anywhere between 0.1% and 99.9%, since we don’t know what “probably” or “maybe” mean in this person’s estimation.
Instead, this person would be better served to say, “they’re 3-betting 99+ AQ+ 100% of the time, AJs 75%, 88 50%, 77 25%…A2s-A5s 75%, A6s-ATs 50%, 65s-T9s 50%…”, etc etc. Can you see how much more useful it is to assign a frequency, and therefore a likelihood of a certain hand being in the range? Assigning frequencies helps us to be significantly more specific with how we discuss hands, and it teaches us not to operate in a binary universe where something is either happening or not happening – after all, this is results-oriented. We want to be process-oriented. Assigning numbers also gives us the chance to plug this hand into a calculator with a much greater degree of accuracy.
What does GTO have to say about it?
The short version of the answer to this question is that GTO play will almost always involve variable frequencies to some degree. Very few things are done with 100% or 0% frequency when you run a situation through a GTO calculator, even a preflop calculator like HoldemResources Calculator – in order to achieve perfect balance and unexploitable play, you have to ‘mix it up’ to a degree in many spots, to use a less scientific parlance.
In fact, when running postflop GTO spots through Simple Postflop or an equivalent piece of software, you’ll find that the software is so specific that the true solution to a certain spot will usually involve each specific combination of each specific hand being played with a variable frequency. That is to say, it’s not only specific enough to say that your opponent should be check-calling the river with 45% of his King-Jack hands, but that he should be check-calling with 96% of the combinations that include the Jack of hearts, or that he should check-fold 24% of the time he has KhJc specifically.
We obviously can’t think in terms this precise while actually playing, but that doesn’t mean we shouldn’t strive for a greater degree of specificity while analyzing hands. What we do away from the table rubs off on our at-the-table thinking, and it’s important that we put in place a thought process that will drive us closer to being able to think as precisely as possible.
The evolution of your thought process
Teaching yourself to always express yourself in terms of frequencies is not easy – it’s tempting to slip back into more simple modes of expression and just use terms like ‘maybe’, ‘probably’, or say things like “I’m not sure if that hand is in his range”. But if you want to help yourself take your thinking to the next level, it’s necessary to take that uncertainty and try to define it mathematically, so that we can start to eliminate the confusion caused by putting ourselves in the dark.
The main difference between “I’m about 50% sure TT is in villain’s range”, and “TT is in villain’s range 50% of the time” is that the former is more results-oriented – the former is based in what’s happening right now, while the latter is based on all the possible outcomes of that same situation over time. It’s an even more big-picture way of thinking than simply putting your opponent on a range.
In general, anything that moves us away from short-term thinking is going to be beneficial. In particular, thinking in terms of frequencies is going to move us out of a binary universe and teach us to recognise the flaws in our own perceptions, as well as the variation in our opponents’ choices over time – it’s simply a more all-encompassing way to talk about a poker hand. So the next time you venture onto the TPE forums to talk through a hand, try to make sure you put some numbers to those adjectives. It’s the next step to an evolved way of thinking.
papanica
Thanks Matt…