Stop Overbetting the Turn vs Weak Players

Overbetting the turn after a small flop c-bet is a commonly discussed strategy. Players across both the GTO and exploitative spectrum tend to agree on the power of abusing “capped” ranges. There’s no flaw in the concept itself, but the most important question driving the entire strategy is: “Are they really capped?” Let's start with a textbook GTO example.

Scenario

Game: Online 6-Max 200NL Effective Stack: 100 BB

Preflop

• BTN: Raises
• BB: Calls

Flop: J♥ 3♠ 3♦

For simplicity, let's follow only what our thick value hands are doing from the button (AA-QQ & AJ).

We range bet the flop for 33%, and look how the BB is supposed to respond.

The BB should find a raise over 30% of the time. In your games are you getting raised 1 out of 3 times here? From my sample this year the entire pool is only raising me 10% of the time.

Let's continue with the standard GTO example. Here we see GTOWizard doing the text book 2x turn overbet with AA-QQ & AJ. Reason being the BB should have raised off effectively all of their trips combos on the flop.

The BB should respond with only a call or fold and never raise. Even with their infrequently slow played trips or turned boat due to the BTN's polar range/strategy. (From my sample I'm getting raised over 7% of the time when overbetting the turn.)

Now let's adjust the BB's flop raise to a 10% frequency and see how the solver adjusts.

That's quite the adjustment. Now the solver prefers a 40% sizing and it's effectively pure checking AA. Why?

The BB under raising the flop does two things. They get to the turn with too many of their strong hands and too many of their trash hands that should have bluffed the flop. The reality is that BB is the one that's polar on the turn! (Well semi-polar. They have air, strong trips, and mediocre under pairs)

Now the BTN wants to use a small bet to re-ask the question "Are you strong?" so AA-QQ, and AJ can overbet safe rivers knowing that the BB is finally capped. The turn is the last opportunity that the BB has to build the pot. They're faced with a hard choice of committing to the slow-play and risking the BTN checking back the river. Or telling on themselves by raising the turn c-bet. In my experience weaker players tend to "tell on themselves" here.

It's worth noting that the solver prefers to check back AA on the turn. This is mainly because the BB should be calling with most of their A high hands which AA blocks.

It's worth experimenting with. In my anecdotal experience players tend to just auto fold the turn with A high to any sizing. So it makes sense to let them catch up or give them a chance to monkey stab the river. It could be that the optimized exploitable strategy is to c-bet KK-QQ,AJ or better with the plan to overbet the river, bluff anything worse than a pair, and give rope with everything else.

Let's run one more experiment. It's a bit extreme, but not unrealistic for how some whales play. We'll adjust the BB's preflop range to be much wider and assume they only 3-bet AA-QQ,AK preflop. They'll donk The flop 10% of the time, and check-raise the flop 0% of the time. We're still going to range bet for a 33% size.

This genuinely surprised me. I expected something closer to my "optimized exploitable strategy" from above, but it seems it went in the other direction. We're not really bluffing and we're checking back QQ? K3?? 66???

It's tempting to believe that aggression is the recipe for beating weak players. That's how I've always seen it. It's possible these examples aren't surprising to you, but for me they were an eye-opener. I haven't been giving weak/wide players enough rope.

Conclusion

Solvers are a tool to aid thinking. None of this is an end-all-be-all conclusion. I plan to implement more rope-giving vs weak players. I'll update with another post in a few weeks with the results.