I have been fortunate to work with poker psychologist Jared Tendler for many years and even co-authored his book The Mental Game of Poker. Of all his advice, I especially remember some he gave when I did not know how to start working on the theory. This was before the solvers. He said, "If you don't know what to do, play your C-game."

His idea was to find his biggest mistakes and work on fixing them. Blunders cost the most, hurt the most, and are usually the easiest to spot. In terms of time efficiency, this is the most promising approach.

Already in the solver era, my other co-author Dara O'Kearney showed me how to sort errors by EV. Some types of errors cost more than others, and solvers know which ones. Since we can find out the EV of our decisions, we have no excuse not to work on the most expensive faces.

In the GTO Wizard, you can look at the cost of errors in the big blinds. Let's see how this can be used when working on a game.

### Preflop Mistakes

Let's start with a simple example. Here is a chipEV raising range from UTG with an effective stack of 40 blinds for multi-table tournaments.

The expectation of each hand is indicated in the matrix. It is very profitable to raise aces – each hand will bring an average of 10.05bb, but a pair of fours plays almost zero – the profit with it does not exceed 0.01bb. Hands such as Q8s have an EV of zero.

If you hover over the combination, the program will show the expectation of each action.

As you can see, an all-in with AA is also profitable and will bring in an average of 8.01bb, but this is significantly less than expected from a regular raise. It makes sense: with such a strong hand, we want to provoke action, not stop it.

Some of the hands we open, such as Q9s, can become negative if we replace the standard open with a shove.

An open raise earns us 0.07bb on average, and an immediate all-in loses 0.87bb. The difference between these actions is almost one whole blind! Q9s play well post-flop, but a 3-bet will probably fold. When we shove, only those hands much stronger than us will call us, and the expectation will turn out to be negative at the exit.

This is, of course, pretty obvious. The most interesting thing here is to dig into the border hands, to see how expensive it will be for us to make mistakes with going out of range, even if by one division.

In this example, the weakest offsuit ace we can open up is ATo:

All-in with ATo is also negative, but a standard raise will bring in 0.06bb. Let's compare this to waiting for a hand just one space down, A9o:

If you find yourself in this situation, open with A9o, and then look at the chart and make sure that you made a mistake, you can try to console yourself with the fact that since the hands are close, there is almost no difference between them.

Alas, there is a difference, and a very significant one.

Raising with A9o costs us 0.08bb, and the difference between raising with ATo and A9o is 0.14bb. It seems to be just a little bit outside the range, but in the long run, it's as much as 14 bb/100 – an indecent loss rate for a regular, any cash player will confirm. If there are many such mistakes, they accumulate over the course of the race and take a lot of money from you, much more than you think.

If UTG opens with an optimal range and everyone folds to the big blind, you need to defend like this:

The BB continues with over half of his hands as he has excellent pot odds and closes the action. Take a mediocre hand like J6s:

The difference between calling and folding is very significant – 0.26bb! A mistake in this spot will cost a lot more than a wrong open with a marginal hand UTG. Folding looks like a safe bet, but it actually burns money as you lose 26bb/100 compared to calling.

But what about if the player has not invested in the pot from the blinds? Let UTG raise, and let the hijack come in:

The weakest offsuit ace to raise with is AJo with an expectation of 0.07bb to call.

Interestingly, the hand right below it, ATo, never calls, but 3-bets lightly.

This is because ATo plays poorly against an opening range from UTG, but blocks strong hands containing an ace, allowing for a 3-bet bluff. Bluffing with ATo is near zero, but calling loses 0.11bb. The gap in waiting for a call with AJo and ATo is 0.18bb. Have you noticed that the difference in EV between adjacent hands increases? A bad call hurts your wallet more than a bad raise.

For the sake of the experiment, let's try to move one more space and check the call with A9o – a hand that should not be played in these positions under any circumstances:

The call with A9o loses 0.42bb, i.e. the difference with the most borderline GTO call has reached 0.49bb. This is already very serious – the loss rate is 49 bb/100. Going out of range a little bit is already unpleasant, but this is almost a disaster.

### Postflop mistakes

In the following example, UTG opens with a 40bb effective stack, gets a 3-bet from the small blind, and calls. Flop – . On this flop, the small blind bets almost his entire range in one size – 25%.

This is a classic range bet – the widest small bet that can be made due to the large advantage of the small blind's range over UTG on this texture. We can look at the expectation of other sizings with different hands.

Whether we have the nuts, a medium-strength hand, or air, a quarter-pot sizing is the most profitable. It's tempting to bet big with a monster like KJs for more value, or with air like T8s to knock out more hands with higher hands, but that would be a mistake. With KJs, the difference in expectation between a half pot and a quarter bet is as much as 0.31bb – 31bb/100. Huge gap! But we just played a little with the sizing...

In response to a c-bet in a quarter of the pot, UTG acts like this:

Most hands should continue since the bet is small and the board isn't very dynamic, so preflop hand strength still comes into play. One of the weakest hands that UTG can continue with is KQs, the expectation of calling with which is 0.07bb.

Among the few hands that do fold are KQo with a negative call expectation of 0.71bb.

If you know you have KQs to call, intuitively it would seem that KQo is not that far behind, but the lack of a backdoor flush draw makes this hand completely unplayable.

Let's say UTG called with the correct range and we saw the turn – . Here's how the small blind plays:

The king is a good card for him, the range advantage is still there, and the strategy, by and large, does not change – range betting 25% again. Let's take a look at other sizings:

The last combination of KJs is almost the nuts, but it's still best for us to make a small bet with it. If we follow our emotions and increase the sizing to half the pot, we will lose as much as 0.87 bb – a mistake worth 87 bb/100! Almost three times more expensive than making a mistake with exactly the same action on the flop. It's just that the pot has grown, and it has become more painful to make mistakes.

### Mistakes at the end

Let's now see how the situation changes as we get closer to the money and move to the ICM model instead of chipEV.

Under ICM conditions, the GTO Wizard program begins to measure expectation not in blinds, but in percentages of tournament equity: 100% is the total value of table stacks – just the table and not the entire tournament. If the total stacks of all the players on our table are worth \$1,000, the loss of one EV is 1% of \$1,000, or \$10.

In the following example, we approach the bubble:

The player on UTG+1 is in danger due to a very short stack, the position of the big blind also inspires concern.

If UTG folds, the UTG+1 strategy should be:

KTs – the weakest hand with a king, the push from which brings money:

We've seen examples before when an open raise is positive and an all-in is negative. The opposite situation arose here: an all-in with KTs is profitable and gives us 0.07% of the table equity, but an open raise loses 0.03%. The reason is obvious – shoving works because we have pretty strong blockers and decent equity when called. The open is disadvantageous because our 3-betting range dominates us, and even if our opponents call, we risk missing the flop. This is a very important factor when playing with a strong ICM influence. The last thing we want to do is to implement a significant part of the stack and find ourselves in a difficult situation, under pressure, where we are almost forced to give up. It is much better to use your stack for aggressive action, setting difficult tasks already for opponents.

KTs shove up, what about K9s?

And this hand loses 0.03%, a 0.1% difference from KTs.

If UTG+1 pushes the correct range, how should the BB respond? Have you forgotten that the BB is short stacked too? Losing all-in almost guarantees he will bust.

The bottom of the calling range with unpaired hands is AJs and we are making 0.38% table equity with this hand.

And here's what happens to the hand one space down – ATs:

There is a whole abyss between these two hands! The difference between a bad call and a good call was 0.41%.

This is an important and valuable lesson. In the previous example, a wrong border push cost us 0.1%. Going out of the range to call – and also by just one space – costs us more than four times as much! A bad call is always worse than a bad push. After all, by pushing you can win money both when you hit your hand and due to fold equity – opponents can give up. When we call all-in, the only way to win is to win the showdown.

That is why it is so important to study ICM carefully. Identical errors cost us differently at different stages of the tournament. If we are 3% wrong around the \$11 Sunday Storm bubble when the total table equity is \$200, the real money value of our mistake is 60 cents. The exact same mistake on the final table, where the prize pool is approaching \$50,000, is already worth \$150 – almost 14 buy-ins! Deviation from theory by one division turns out to be a local catastrophe.

### Conclusion

We studied the difference in EV between the weakest continuation range hands and the strongest fold range hands and saw how costly making repeated mistakes in situations that look innocent and marginal can become. We have also seen that some mistakes are worth more than others. With a bad fold, we lose the opportunity to win chips. A bad active action – betting or all-in – loses much less than a wrong call, because by calling we deprive ourselves of the opportunity to win the pot immediately with the help of fold equity.

That is why I advise you to first learn the calling ranges, and only then proceed to pushing ranges. A bad call at the most expensive final table of the year takes away dozens or even hundreds of buy-ins.

Post-flop mistakes become more expensive with each new street. The most expensive mistakes are made on the river. Therefore, post-flop correction is best done in the opposite direction – first work on the river, then on the turn, and then on the flop.

Picking some lower fruits is easier – they hang right above you, and you don’t need to climb anywhere. Those are the ones that should be dealt with first. Preflop these are calling ranges – sharpen them up before moving on to open raises and shoves. Do not forget to adjust the conclusions taking into account ICM. Smart calls deep in the money and at the final tables are your most expensive decisions of the year and the key to MTT success.

You can buy a subscription to GTO Wizard in the GipsyTeam store . We accept Russian cards and all major payment systems and give up to 7% cashback.

GTO Wizard
from \$39

Efficient GTO study and practice for most disciplines in your browser