Game flow

It is impossible to talk about poker without including the psychological aspects. But at the same time, the language we use to try to talk about psychology is very imprecise. If you ask random players why they do what they do, you will usually hear people talking about scans, gut feelings, the amount of time they have or the flow of the game. But if you ask them to explain these concepts, all you get is some philosophical musings and attempts to get away with it. And how can you blame them? A lot of things that happen on the psychological battlefield are usually more intuitive than rational. But that doesn't mean that psychology is impervious to rational analysis. It simply means that it will be harder and we will have to use much more precise words than we are used to.

We can start with the slipperiest term in psychology: game flow. It is a well-known term, because of the difficulties in trying to define it. However, those who have played many hands at the poker table usually know intuitively what it means.

A typical example of it contradicting the definition: 'According to the flow of the game, you should bet or check here', 'Bet or check, depending on the flow of the game'. What phrase could be used to replace "game flow"? We could say match flow, but that would not help at all. We could try "What do you think he's going to do?", but that is not accurate. Game flow is a more specific thing than these phrases.

Let's try a little experiment to help bring some clarity. Historically, poker notes were designed to record hands synchronously. This is how a language was invented to denote how each hand was played, by writing down each action, on each street. Eventually, this evolved into modern hand history. But we don't have any universally accepted notation for marking hands diachronically. We can only show our session in one way, by cutting and pasting the history of how many hands were played in that period of time. Why don't we try inventing a new language for marking?

Let's first agree on the marking. If someone makes a valuebet, we will mark V, the hand where B bluffed, F folded and C called. So, if the sequence of hands was such that on the river (or some other fixed point of interest), first the opponent made a valuebet, then bluffed on the next hand, then made a valuebet again, then checked/folded, and finally called, we would write V.B.V.F.C. This would be value, bluff, value, folded, called.

You may have noticed that there is very little information in this marking. We don't know what size the pots are or whether they are similar, we don't even know who the raiser was. So we will use this to mark similar situations that we want to look at. This means that the pots will be similar in size, the same person but'ins, in very similar places.

Following the principles discussed above, let's imagine the following situation: we are heads up, the opponent makes a 4-bet preflop. We will use our agreed marking after our own 3-bet. The opponent is therefore considering whether to 4-bet for value, 4-bet as a bluff, call our 3-bet, or fold. So we have a sequence of 20 hands that looks like this (which, by the way, is taken from a real match):

F.F.C.F.V.F.B.V.F.C.F.F.B.F.F.F.V.F.C

Let's read this sequence carefully. The first thing you should have noticed is that this guy 4-bets quite often (add Vs and Bs and you get 5/20, or 25%), but that's only a small fraction of the hands. Nevertheless, seeing all these 3-bet hands recorded in this way, we can make valuable observations.

First of all, remember that we don't choose to do a valuebet or not. We do it when we have a good hand. The same can be said for counter-hands to 3-bets. Although there is a slight variation in players' 3-bet call ranges, most players usually call with almost the same range, with a slight variation at the end of the range. This does not affect the flow of the game in 4-betting at all, as players almost always call with a predetermined range and do not tend to play with it too often. Our opponent simply calls when he has a certain hand and thinks he has to call with it. He doesn't choose whether to call it or not. Calls are usually clear in certain places too - the opponent knows exactly what you are doing and what your range looks like when you call.

With all this information, we can say that a call is not a decision-based event in the flow of the game. It is a systematic choice. Removing all calls from the sequence would give us a clearer picture of the psychological factors. The sequence would look like this:

F.F.F.V.F.B.V.F.F.F.F.B.F.F.F.V.F

The argument we made about the opponent not choosing to call can also be applied to 4-betting for value. Although there is a small variation between players in the value ranges for 4-betting, most players will use the same range more often than not (some players will use 4-bet/call from AJ, 77+ heads up if there is a very aggressive 4-betting dynamic).

So, in a sense, this player has no control over his value bets either. If we were faced with the same sequence, our V would be in exactly the same places.

However, this does not mean that we can eliminate value betas from the sequence, because in fact these betas are indistinguishable to us from his bluffs. Both actions are simply 4-bets to us. The V of the player affects the psychological aspect of the game, even if he cannot control his valuebets. Each V and each B is for us evidence that the player is bluffing more and more. Even if he 4-bets ten hands in a row with value hands, we will only see one or two showdowns and we will probably conclude that the hands we didn't see were bluffs.

Here's what the sequence looks like when we highlight V and B, the 4-bets:

F.F.F.V.F.B.V.F.F.F.F.B.F.F.F.V.F

Now that F looks like the spaces between 4-bets, we can analyse this dynamic.

The player first makes three consecutive folds, then receives value hand. Then he folds and finally bluffs. He follows up with another value hand and finally, realising that we have just seen a very concentrated sequence of betas, decides to cool things down a bit and folds four times. Then one bluff, followed by three folds, another value hand and another fold.

So it seems that the player usually chooses to fold about three times between bluffs. After the V.F.B.V. sequence, he pauses a little longer, folding four hands, presumably to restore his image in the eyes of the other player. At the end of the sequence, a value bet was made, following a bluff. We would bet that this player folded the other two or three hands.

So what is the purpose of this exercise?

Basically, we analysed the flow of the game by analysing this sequence of 20 hands. We could describe it as follows: game flow is the pattern of decisions implemented over time, which influences subsequent decisions. There are two main elements that make up the flow of a game: the simulated randomness and the emotional dynamics (we will talk about them another time).

And finally: game flow is a human phenomenon. If two computers were playing with each other (and both knew they were playing with a computer), game flow would not exist.

Where is the best place to play poker?