As someone that has spent a lot of time studying and playing chess and especially chess engines, I am aware of many of the criteria that they use to select the “optimal” move in any given position. The list of desirable criteria is fed into the engine and the ranking and importance of each factor. These are weighted and so certain factors are given more importance than others when evaluating a position. So the chess engine knows what factors to weight and by how much and so within a very small amount of time “evaluates” the position.
The end result of the best programs in the world now is that they find a solid line at the very least within a few seconds. Their analysis goes very deep but this is less relevant in chess then many people think. Top grandmasters understand intuitively when there are no tactics in a position and they analyse between 3-5 moves deep on average. So if there is no forcing sequence after 3-5 moves then the probability is exceedingly high that none exist.
This is a valuable analogy with poker for several reasons because many people over-analyse in poker in situations where the solid line can be found in a fraction of the time. So the next question is if we had to build a poker engine then what factors would we program into it to enable it to find good solid poker lines? I think this boils down to how we make money in poker and what errors our opponents make that we can exploit or how our opponents can exploit us.
In short then we can look at two key factors in our ability to evaluate a poker hand and its potential profitability. These are the probability of winning the pot without a showdown and the probability of winning the pot at showdown. So the combined level of fold equity and pot equity make a weighted figure but what they figure is can be very difficult to ascertain within a poker game. The best players are able to do this by experience and knowledge.
The next key question is “which factor is more important, pot equity or fold equity”? To answer that question isn’t easy because it all depends on the particular game dynamic and this is probably why it is so difficult to program a computer to play poker in numerous different games and stake levels at the same time. This is because the game dynamic is different. However let us take full ring NLHE as an example at say the NL50 level.
These games are very tight and aggressive and so showdowns are seen far less frequently than if the game were very loose and passive. So fold equity becomes the dominating factor and so position is obviously very important within this game dynamic. This can be highlighted with the following hand example where you have the 5c-5d in early position.
In a weak loose passive game then limping is profitable because there will be numerous weak limpers enter the pot behind you and who are weak enough to pay you off when you flop a set. However in a solid tight-aggressive game then what will happen is that nearly all of your opponents to your left will fold but one that will raise with the blinds folding. So you are left with a position where if you fold then you lose a chance to hit a set and lose one big blind but if you call then your chances of hitting the set are not good either and the flop with overcards will look scary…….either position isn’t good and so the best play would have been to fold pre-flop.