Stats Are Nothing, Analysis is Everything (Part 1)
What are the purpose of stats? They only indicate how often a player performs a particular action, but the situations in which he acts, and the opponents that he acts against remain unknown – in order words, statistics show a game's pattern, but never the games content.
For example, say a soccer team lines up with three forwards – that is the statistic. This does not, however, mean that they will automatically get in more shots on target (the content) as opposed to if the team used one or two forwards, because, maybe:
a) all three forwards have an off-day
b) the ball hardly ever reaches them as the opponent team's midfielders are superior
Because of the over-estimation of the usefulness of statistics, many low-limits players make two typical mistakes:
1) Having “over-communicated” with “more experienced” players at the same limit, they begin to correct the stats artificially
2) They evaluate the opponent's strategy and range incorrectly
So, let's look first at the first variant. Consider the unfortunate soul who plays 6-max with 16/13 stats. Doing so, by the way, is highly inadvisable, because the blinds will eat up a significant part of your profit without appearing to do so – plus, players who play tightly and keep their strategy as “wait until something is dealt” are a problem in two ways:
1) When a good hand happens along, you will not always get the full worth from it. But then what else can you expect? Here is an opponent who did absolutely nothing for a long while, then suddenly he raises pre-flop and bets on all three streets. Certainly, more or less reasonable opponents would run away, horrified.
2) Paying coolers for opponents. It's certainly clear, when you have been dropping patiently, waiting for aces in the deal, the aces came and you got the raise of a contbet in the main pot on the flop 678ss. It doesn't matter if one is raised, as your opponent will almost always be lower-ranked here, as he was waiting for those aces, and not looking to drop them. In general, it's a bit of a nightmare.
So, the player realizes, either by himself or with the help of others, that it does not work this way anymore, and based upon data mining, most regular players with a positive balance on this limit play with 21/17 stats. This means he should play the same way. This is the same as buying a car and a driving license, but forgetting to acquire any actual driving skills.
But how can it be raised? There are many variations. One can open four percent wider in all positions, or raise wider only at the button, or at the button and small blind, and so on. Incidentally, many people do not know that if you raise 3-bet from four percent up to seven percent, then VPIP/PFR will increase up to the same three to four percent.
Let's assume we have decided that we will steal on the button more often. A number of problems arise immediately. For example, many players will raise on the button with 22+, Ax+, J9o+, 56-57s+. The player thinks that he will add it to the raise range of J7s+, J8-K8o and 36s-T7s. But is there any sense in this? Not always. This is why the fact of who you are raising is more important than what you are raising.
In this case, if you've decided to steal more, you have to do it for the players who do not give up, neither for the steal or for the contbet. The same works for raising (maybe a bit tighter), but to those who give up for a steal or call and then fold for a contbet (for example, fold to a contbet OOP less than 60 percent of the time), one has to raise half the deck and even more, up to Qxs and cards in general, any one-suited cards that make a straight. In other words, if stealing at the button equals to 35 percent and you want to increase it up to 45 percent, it does not mean you have to raise ten percent more then everyone else, it means that you should raise maybe 30 percent against loose players, and between 60 and 70 percent against tight players. You will then get about 45 percent on average.