What GTO Zealots Don't Want You To Know

Hey Crushers!

Today we'll look at how you can make your life so much easier by doing something that feels like it isn't allowed. Sounds too good to be true? It's not!

Do you often find yourself stressing about being balanced when making a cbet? If so, you will like this issue of the newsletter. Because today we'll explore that playing unbalanced doesn't automatically mean you'll lose a lot of EV.

I know, GTO obsessed zealots won't like to hear this, but let's look at it one step at a time.

Let's say we're in the BB and face a BTN open raise to 2.5bb. We 3bet a GTO range to 13bb and get called by the BTN who we also assume to be playing a GTO preflop range in this case.

The flop comes KQ6r and you're debating what to do with your range. The flop is dry and K high, so can you just cbet range? What about hands like JJ or 88 that don't gain too much by betting? Or hands like 9s8s that almost don't have any equity at all? Don't these just want to check, which means you would have to implement a checking range?

Well, let's have a look at the solver.

Going with a selection of 33%, 66% and 130% pot, GTO Wizard tells us a small bet is preferred, but not being range bet:

However, if we use GTO Wizard AI's feature of selecting the single best bet size for us, it turns out a 25% pot size would be the best one. When it comes to practice, using a single bet size, especially on the flop, is going to be your best approach in most cases. It simplifies your life a whole lot while rarely costing you a lot of EV.

In this case, the 25% pot bet size would be used very frequently, but not quite as a range bet:

We can see that the solver would implement a checking range across a number of hands. Some being pairs, but most being air hands.

If we take a closer look at the hands that would check the most, we'll see that all of those hands fit at least one of these categories:

• They are spades, which don't have backdoor flush draws

• They don't have good backdoor straight draws

• They block our opponent's auto folds (mostly A♠️x; identified by the high "Trash Removal" score)

In other words, they have tiny equity and/or block our opponent's folds. Simply said: They are bad hands to cbet.

So what would happen if we just decided to cbet ourentire range nonetheless? How would it effect our EV? Well assuming our opponent adjusted perfectly, these would be our EVs in each case:

GTO cbetting

Range cbetting

14.11 vs 13.97, which is actually quite a big difference. You lose 14bb/100 by cbetting range instead of a perfectly balanced 85% range.

Of course that assumes that our opponent picks up on us range betting and adjusts their strategy perfectly. Then they would almost never fold:

The BTN's perfectly adjusted reaction against our range cbetting

If they don't do that and we assume they'll play the same GTO range the solver would play against a perfectly balanced 85% cbetting range, our EV would be much higher:

The BTN's response to a bet if we cbet the GTO frequency (85%)

Our EV if BTN didn't adjust against our range cbet

What does that tell us?

It tells us that only if our opponent is fully aware of our strategy, which they will never be in practice, AND they implement the correct counter strategy perfectly, cbetting range will cost us a big amount of money. Otherwise it won't.

But to make it even more interesting, let me show you this:

I transferred the simulation to Piosolver and it also results in the GTO strategy being around 85% cbet for a 25% bet size.

I then node locked the cbetting range to not cbet range (100%), but only 95% of hands. And Piosolver has this cool function where it automatically constructs a 95% cbetting range that includes the 95% of hands that are best to cbet, while still being balanced. In other words, it removes the 5% of hands that are best to leave out from the cbetting range.

And first of all, I noticed that the difference in EV going from 85% cbets to 95% cbets was much lower than going from 85% to 100%:

Our EV when cbetting GTO

Our EV when cbetting 95% of hands

Now instead of 14bb/100, the EV difference is only 7bb/100 (chip numbers in Pio are 10x of big blinds, as the software cannot calculate half big blinds. So betting 6.4bb becomes 64chips and EV of 14.37 is 143.7 in Pio).

So not bad.

It means that only checking 5% of hands instead of betting them can win us 7bb/100 more than just betting every hand. So what hands are these that Pio wants us to check?

As we can see they are mostly spades without a good backdoor straight draw (9♠️8♠️, 9♠️7♠️) and Axo with the Ace of spades, which blocks our opponents A♠️x which they would always fold to a cbet.

Of course the solver also mixes in some pairs into the checking range to be balanced and to protect the checking range.

But this made me wonder. What if we ONLY ever checked the worst of the worst hands, namely exactly the hands I mentioned above, 9♠️8♠️, 9♠️7♠️ and Axo with the Ace of spades?

So I nodelocked the strategy:

...And discovered that the EV doesn't drop a lot! It's only 4bb/100 worse than playing a perfectly balanced 95% cbetting range. And that already takes into account that our opponent knows we only ever check our worst hands! But still, they wouldn't be able to profit for more than 4bb/100:

Meanwhile, you could mindlessly bet close to 98% of your range - basically everything but the worst of the worst hands, which you'd just check.

Isn't that what we all dream of? In the current age of balanced GTO strategies it almost seems like a sin to even think about playing a strategy that only checks bad hands. But as we can see, there are situations where this isn't a big mistake, even in theory.

So what can we take away from this?

• Simplifying your cbetting strategy to one bet size will make your life easier

• Simplifying your cbetting strategy to a range bet when the GTO solution would be a high frequency cbet, will make your life easier, but can cost you quite some EV in theory in some cases

• If there's a small number of hands in your range that really like to check, consider only ever checking those hands and cbetting all others. As we saw, it is possible that the EV doesn't drop a whole lot, even in theory. In practice it likely won't drop at all, but maybe even increase. As your opponent will not be able to pick up on your strategy that quickly, if at all. And even more likely, they wouldn't be able to correctly adjust against it, but that is a whole different topic...

I hope you got value out of this issue and can implement some of these findings into your own game to make more money.

Thanks for reading!

Talk soon,

Victor

2 Card Confidence

Latest news:

I'm the middle of producing Preflop Xploits, a course that equips you with all necessary tools to exploit your opponents preflop. Including data analyses of a real midstakes player pool on Pokerstars and the average player's tendencies. Sign up to the waitlist to receive updates about the release and be the first to profit from the exclusive insights.

Save 10% on your first purchase of GTO Wizard by using this affiliate link (only works for new accounts).