Translated with the help of AI. We apologize for any errors and would appreciate your help in correcting them.
Translated by order of the educational portal university.poker
Original source: GTO Wizard
Many players find it difficult to cope with aggression on the turn, and therefore it is useful to include it in your strategy. The decision when to make a check-raise, check-call or check-pass on the turn can significantly affect EV, since it is here that it is determined whether we will play for an average or a very large pot. But it is not always easy to determine which hands are best used for check-raising, especially given the influence of blockers and the texture of the board. Considering all these factors, let's look at a few basic principles that will help build range for check-raising on the turn. The easiest way to divide check-raises into two types: when we go all-in and when we don't. Let's start with situations where we go all-in -in with the help of check-raising, which happens mainly when the villain puts both on the flop and on the turn.
1. check-raise all-in
Checking the entire stack is an important part of the arsenal, especially in 3-bet sweats or in situations with short stacks in tournaments. But before choosing hands for such a game, you need to take into account some points.
There are situations when we do not have an important component for check-raising on the turn.
Even if we technically have a check-raise turn range, it may be small enough to make sense to use it fully. The main determining factor for rejecting the check-raise on the turn will be that the bet range on our opponent's turn is highly polarized — the reason is that the polarized bet range will not have too many difficult decisions on our all-in. His bluff will fall off, and his Velho hands will continue. Therefore, the first principle is to avoid check-raising on the turn if the opponent makes large bet. Often this is due to the release of a blank card on the turn, which allows the opponent to put the polar range.
- Here's an example of what I mean:
In the single-raise preflop sweat between the button and the big blind (deep 30bb in MTT), the opponent often puts a quarter of the pot on the flop
, and we call. Then on the turn comes out
, we wait, and the opponent makes a bet of sized in the pot, which is actually a little more than the geometric size of the bet for this situation, leaving only 71% of the pot on the river. Or, if you look at it differently, the opponent actually allows us to check out 71% of the pot.
2. blank thorn in a shallow stack (30bb) turn
The turn betting range of a BTN player mainly consists of a top pair and strong hands, as well as a lot of weak bluff. At the same time, he does not put his draw, because these hands do not like check-raise, and many of the draw on this turn have become much weaker. In fact, 85% of the hands that the opponent puts on the turn do not have a draw, that is, only 15% of his bet is a draw.
Breakdown of the bet range on the BTN turn
The more polar nature of the opponent's betting range (in the sense that his bluff has a lower equity) means that there are very few value hands in our range that need any protection (very few river cards that could scare us). Therefore, our strategy against a sized bet in the pot consists of only 2% of check-raises (about 8.5 combinations).
If we completely eliminate the check-raise option, it won't actually take away any EV from us, because all the combinations that could go into a check-raise still behave normally through a check-call. However, one important point that is worth noting is that in some cases we should think about dock beta on certain thorn maps. But fortunately, this is not one of those cases.
3. Dynamic turn in shallow depth (30bb) turn
Now let's compare this situation with another, where the turn comes out with a completely different card. Instead 
, two flush draw appear on the board, and some straights are closed. Here we have some dunk betas on the pot, about 9% of our range. We do not remove this option, as it affects the opponent's range of bet and our range of check-raises. After we wait, the opponent no longer puts so much on the turn, choosing the bet in the pot, because the turn is no longer a blank card, and there is no significant advantage in nats on the BTN.
Faced with a rate of 67% of the pot, we now have a fairly substantial check-raise range of 11% of the BB range. Interestingly, we don't do it purely as a bluff.
As you can see below, the bulk of our check-raises (which are shown in dark red below in the “Manhattan” chart from the Breakdown tab) are on the side of stronger combinations, with only a few very specific low equity combinations that follow this path:
If we look closely at our check-raise range using the range breakdown graph below, we can see that the most common hand category in this range is two pair. In fact, almost half of our range consists of two pair that are vulnerable (e.g., K8, K5, 87, and 85). We also put all our low straights, and higher straights are more likely to collide, because the low straight (64) is more vulnerable to bad cards on the river. To complete the range, we also check-raise several combodros (for example, T4s, 62s) and draw pair (for example, 86o, 76o).
BB Check-Raise Range Breakdown
It may seem counterintuitive that we do this without having a large number of bluff in our range, but there is a very specific reason for this. It's all about what part of the opponent's range we're actually trying to attack. In the previous situation with the turn, the opponent's bet range was highly polarized, and therefore we could not put any part of his range in an indifferent position with the help of a checkraise, but in the described situation this is not the case.
Breakdown of the BTN bet range on the turn
This chart is a breakdown of the opponent's bet range on the turn. We immediately notice that it is quite strongly focused on several specific types of hands — there are many top pairs in its range, and since 48% of the range do not have draw, this means that 52% of the range have draw, which means that many draw hands are put on the turn.
This has two specific implications for our check-raises:
- Now we have an incentive to protect our strong but vulnerable hands from possible bad cards on the river.
- We can put a lot of enemy draw in a difficult situation with our check-raise.
The best example of hands that we target with checkraises is AKs. As you can see in the graph,
and 
are indifferent to the push on the turn. We add to our check-raising range sufficiently weak hands such as a pair with draw, so that the top pair with the top kicker of the opponent has an incentive to call, while we leave in the range enough two pair and straights for him to have an incentive to fold. Many top pairs of the enemy without draw find themselves in a difficult situation when he puts and we push.
However, our check-push on the turn performs another function — the fact that we will occasionally check-push on the turn, actually restrains the opponent from bet with certain hands, because such hands will be in a difficult situation if we make Push. Here I highlighted the range area of the opponent on the turn with top pairs, pairs of lower rank and the second pair, as well as the entire range with draw.
when BB can check-push on the turn
As you can see, both parts of the BTN betting bet range can be bet a little more than 50% of the time if BB has a check-raise range on the turn. This is because the opponent is forced to be very selective in choosing the combinations that he puts and that he waits for, for fear of getting into a difficult situation after the raise.
However, if we remove the check-raise option from our strategy, these two parts of the opponent's strategy look like this:
in the absence of check-raises at BB on the turn
Obviously, in this situation, the opponent can play much more aggressively if he does not need to fear our raise. He can fine-tune extract value with weaker equity and bluff with a wider range of hands with draw. This is another key reason why it is important to build a range of checkraises on the turn. We want to prevent our opponent from betting all his delicate Velu-hands and draw with impunity. In fact, we force BTN to introduce some polarization into its turn betting range — forcing it to find bluff with low equity hands, which is difficult to do in such situations. We can not put in a difficult position a hand like oesd + flush draw, since such hands have too much equity, but we can make many flush draw and straight draw more often play the check if the opponent is afraid of our check-raise.
The presence of check-raises in our strategy prevents the enemy from betting all his thin valley hands and draw with impunity.
Thus, it can be concluded that our check-raising strategy on the turn is aimed at two specific parts of the enemy's turn range:
- Middle part of its valley range (here: strong top pairs without draw).
- Medium draw (here: stronger flush draw and straight draw).
In general, in situations where the opponent's bet range on the turn is initially more polarized (as, for example, on the turn, which we considered earlier), we have practically no check-raise range on the turn. Therefore, in such cases, it is quite reasonable to completely exclude this part of the strategy. A more detailed analysis of this approach will be discussed in the section on deep stacks. Now let's move on to situations where we do check-raise not push, but less sizing, which leaves us to play the river if our check-raise is colored.
4. check-raise is not All-In
We will deal with the same situation: a flop with two suited cards and the same cards on the turn — and . The difference is that now we are considering a stack of 100bb instead of 30bb, that is, the SPR at the turn will be much higher.
Empty thorn in deep structure (100bb) turn
Let's start with. BTN again chooses a betting strategy of sized in full pot, despite the fact that the SPR here is significantly higher than before. The main difference is that we have the possibility of a check-raise sized to a full bank, while 73% of the pot remains on the river, that is, we are still far from the situation with the check-push.
BB's response in this situation includes slightly more checkraises than in the case against full-bank bets at 30bb pot stacks, although the total number of combinations increased from only 8 to 14.
In this situation, I decided to conduct a more thorough analysis to check whether the complete exclusion of the check-raise option significantly affects the expected profit, given that in deep stacks, removing options from the decision tree may have a greater effect. Instead of giving BTN the only betting option in 100% pot, I added five betting bet sizes: 50%, 100%, 150%, 200% and 250%, with the option to check-raise BB against any size. Sizing of the check-raise was determined dynamically by the solver. I measured both the total EV when reaching the turn and the frequency of bet with a dock bet of 25% and 75% of the pot. Then I changed the decision tree, removing the possibility of a check-raise against any bet size.
The objectives of this experiment were twofold:
- Track how the absence of a check-raise on this turn affects the overall EV.
- Check if we will need to significantly change the strategy of dock-bets to compensate.
Here are the results, starting with donk-bet strategies:
Strategies donk bet BB on the turn in the presence of a check-raise (top) and its absence (bottom)
As you can see, the conclusions are quite unambiguous. The lack of a check-raise range against any bet size here did not affect the need for donk-bets and reduced the total EV by only 0.2% of the pot, which at this stage is equivalent to 0.0182 big blinds, or 1.82 bb/100.
This is an acceptable sacrifice for the sake of simplicity:
- We no longer need to worry about check-raises in this situation.
- Instead of learning the solution node with check-raises on the turn, we just need to refresh our understanding of the river game when the check-raises range on the turn is missing.
We have now established that the presence of a range of check-raises on this blank turn is not at all mandatory.
Dynamic turn in deep structure (100bb) turn
Now let's look at what happens when the turn is not a bank card, but effective stacks deeper. Let's go back to the turn. This time BTN chooses sizing in 67% of the pot, but uses it with a lower frequency — only 39% of its range. With deeper stacks, its possibilities for bet are slightly limited, because the likelihood of encountering a strong part of our range is much higher. For example, hands like AK and AA are put much less often, and TT-QQ is not put at all.
Interestingly, the frequency of BB check-raises is about the same as for shorter stacks (about 42 combinations versus 40), but the structure of the range is significantly different.
At first glance, it may seem that the differences are not so significant, but let's compare the Manhattan charts from the Breakdown tab. At the top, the strategy for stacks 30bb with the inclusion of check-raise-push is shown, and below — the strategy for stacks 100bb using check-raise without all-in.
The differences between these graphs are highlighted in detail:
Upper part of the range:
- For short stacks, we prefer slow play with the top of our range (e.g. with straights) to protect the overall range. This is justified, since there are few chips left on the river (low SPR), and we can easily stack-off.
- With deep stacks, it is much more important for us to increase the pot with the top of the range. If we don't, we won't be able to make the most of our range of strong hands often, as the stacks are too large.
Bottom of the range (bluff):
- With deep stacks for check-raises, we choose hands with less equity than with all-ins. The reason is simple: in the case of a call, we do not go to all-in, and we have another street where you can continue bluffing or find an opportunity for selection.
Conclusions:
- When playing in deeper stacks, our strategy becomes more focused on building up the pot with strong hands to maximize EV pot.
- Bluff at deep stacks become less demanding on equity, as we keep the possibility of continuing the game on the next street.
Another interesting trend is that there is more variety in the deeper stacks in the check-raise range. Some hands of average equity begin to use check-raise with a certain frequency.
To better understand which hands fall within this range, let's take a look at the hand distribution:
BB check-raise turn range: all-in (left) vs no all-in (right)
The main difference in the velly value of the BB part of the range is the strong focus at the very top of the range. As we saw earlier, straights make up the bulk of our valley range, and some combinations of the top pairs are used as bluff. The part of the BB range containing draw emphasizes combinational draw and open straight draw (open-enders), in contrast to weaker draw or gatshots. Although such hands have less equity against the opponent's bet range at this depth, flush draw still occupies a very small part of the range.
One feature of this strategy is that while BTN can technically go all-in over our check-raise, it does not have much incentive to cut the decision tree by one street, as it has a positional advantage. Moreover, its Push against our sizing of the check-raise (equal to the pot) is also quite large, about the size of the pot, so it pushes only 4% of the time (8 combinations). This means we don't have to worry too much about protecting our check-raise range from these kinds of pushes.
Target BTN range
The final part of our analysis is to determine what part of the opponent's call range we are actually aiming at on the turn. Recall that with shorter stacks, we mainly focused on the king without draw. Given that such hands are put on this turn less often, perhaps the goals have changed? In fact, they've really changed. Those hands that were previously indifferent (for example, a top pair without draw) are no longer in this state. Now they become pure folds.
- For example, a hand like AKo here does not even come close for a call — it is a clean fold.
Example:
As can be seen from the graph, the call with AKo loses about 4% of the pot, which is equivalent to a significant loss of about 2 big blinds or 200bb per 100 hands. For comparison, a great example of a hand that is on the verge of indifference is 99.
Depending on the specific combination of EV call with this hand varies from -0.8% of the pot to +0.2% of the pot.
In the pot, which by this point is as much as 43bb, this is a pretty good result — it is difficult to get closer to an equivalent solution than these values show.
The same goes for AA:
With them, call and fold are equivalent. AA in this situation turns out to be a special case, since it is the only hand with one pair without draw, which can call to at least with some frequency. In the previous spot, we targeted hands with one pair without draw to make them indifferent to pass and call, but now our main goal is to get them to fold.
Hands that remain indifferent are uncomfortable draw such as T9o, or pair + gutshot such as 99. We also partially aim for the top of the BTN range to make his hands indifferent, but in a different way. The opponent may want to run two pair or sets for protection, but due to the fact that there are many streets in our check-raise range, they do not get enough benefit from this. Thus, we force them to choose between a call and a push without the significant advantage of one of the solutions.
5. Conclusion
Given all of the above, we can formulate general heuristics (that is, recommendations) for different situations.
Heuristics to simplify decisions
When the turn is a blank blank, and the range of bet in the position is highly polarized – simplify the strategy by completely eliminating check-raises on the turn. This will result in a very slight loss of EV, regardless of the depth of the stack.
Heuristics for check-raises depending on sizing and stack depth
In shorter stacks:
- Use check-raise all-ins.
- Priority is given to vulnerable Velu hands, while the very top of the range can be slammed.
- Bluff mainly consist of weak combo draw and pair + draw hands.
- The main goal is to make hands with one pair without draw indifferent to fold and call.
In deeper stacks:
- Use check-raises with sizing less all-in.
- Priority is given to the very top of your Velé range, as these hands must be played on the entire stack.
- For bluff, a more diverse range is used, including draw with high equity and pair hands + blockers.
- Expect almost all hands with one pair without draw to fold. Therefore, the main goal is to make uncomfortable draw indifferent between the call and the fold, and strong, but not nuts velya hands, indifferent between the call and the push.
These heuristics will help you better navigate similar situations in the future. They have already been of great benefit to me. Good luck at the tables!
