It is important to test poker strategies.
ad will be displayed here
The way you test it first if possible is to run mathematical equations so you can see if it’s a sound winning strategy. However strategies that are winning in cashgames may not work out quite the same in tournaments.
You can basically figure out if a strategy is winnable just by making sure it’s sound in game theory.
For example, against optimal opponents, you have to assume they will know exactly when to reraise you. So if you are going to raise, you should do so with the intention of folding 3 times (assuming you are reraised with a raise 3 times as large as your raise) and rereraising on the 4th time. That means of the hands you raise with, you should only rereraise the top 25%. This prevents your opponent from exploiting you by reraising you too much.
If you are going to raise, you should raise based on the amount of opponents left. If you start with 10 opponents and you are first to act, the average best hand will be in the top 10% of starting hands. So you should only raise with the top 10% of hands. This varies due to the fact that certain hands stand up better against a single caller, than multiple callers, but you can still estimate one or two callers and raise enough so that you will only get 1 or 2 callers. If you get called, assume they also have a top 10% hand. If you get raised, assume they have a top 5% hand.
You can do this from every hand position. For example, if everyone folds to you, and it’s you and 3 other people, assume the top 33% of hands is the best. This isn’t quite accurate because people with an ace or king in their hands are less likely to fold, and thus since more people have folded before you, it is more likely that someone has at least an ace in their hand. However, only a slight adjustment is needed.
Now that’s the mathematical way to test a strategy, and it is only preflop. After the flop you can have strategies where you move all in or fold on the flop. You can figure out how many flops you need to move all in, what range of hands your opponent will have to put you on, and what they can call with. Then you can figure ou the equity in the hand, and determine an accurate decision. This can get complicated, but certainly can be done. Given that you get called you have to assume your opponent has a stronger hand than your pushing range, or has equity in the hand. Based on his calling range, you calculate the remaining equity. This is easier to do if you have a strategy where you have either a strong draw or an extremely strong hand.
Now you can easily deviate from this strategy, and be successful, if you can anticipate how your opponents are going to play. However, that’s a dangerous strategy, because people can easily “act dumb” only to set up a playing style which you try to exploit, then they can merely turn it around and exploit you. This is more about outplaying your opponent, and becomes more of a chess match for the really advanced players, but it’s so easy to get tricked into playing a certain way that it makes sense never to deviate too much.
Another useful strategy is equilibrium all in strategies that are discussed more in detail in the book “kill everyon” the sequel to “kill phil”.
The only real reason to have to test poker strategies is in tournament situations. Tournaments are very different in that if you lose you are done, the payout is not winner take all, and you should no be risking as much to accumulate chips, unless that can significantly reduce your risk in the future. Since the blinds cost more every single round, a blindsteal is more valuable in the future, and the potsize will be much larger later on. You want to maintain a manageble chip size if possible, but not at the cost of risking your tournament life since it’s more lucrative in the later rounds.
This means betting 6 times the big blinds early, knowing that with 1500 chips at 5/10 blinds this is only 60 out of 1500 or 4% of your chipstack, but when the blinds go up higher, you will be playing at significantly higher stakes. This amount you raise should also vary based upon the action you get after the flop. In an aggressive game, you can still make a large win even at a small blind level so you actually want to keep the pot as small as possible so you can exploit the post flop decisions to the maximum.
You need to have a strong concept of the kelly criterion in poker tournaments if you really want to dominate consistently.
Testing poker strategies is difficult because the playing environment varies so much. Really, what you should be doing is tracking your chips every game you play at every blind level, and then you can start to estimate the chance of you getting to a certain point. The place when the strategies tested becomes applicable is towards the end of the tournament if you get there often enough. The problem is, this is when you are least likely to be alive so you won’t have a sample size to test.
The conclusion is, you should test strategies against advanced AI, and if possible in sit N go situations where the chipstacks are similar to what they will be late in a tournament, which is rarely the case and the payouts are rarely ever similar. You need to calculate your equity given a certain chip stack size. Then repeat at another chip stack size. And then at another. test multiple strategies, but mainly figure out what X amount of chips is worth to you given the payout structure. That way you can test from the final 3 tables on and figure out what percentage chance you have of getting to the final table. You can end the test when you get to the final table because from prior test, you can estimate your equty based on your chips remaining. Do that enough, and you can get an idea of how much it’s worth to be alive with X amount of chips with 3 tables left. You continue this until you cover every point in the game. If you play 10,000 simulated tournaments, you will have a very good concept of whether it’s worth it to risk your tournament life to double your chips or not at any given point of the tournament.