Stephen Chidwick Explains Top 3 GTO Misconceptions

25 Apr 2025

Intermediate

This material is for medium-skilled players

This is Steven Chidwick (currently ranked #3 on the All-Time Money List), and today I want to talk about what I believe are the most common misconceptions around GTO poker.

One of the things I hear all the time is that people think playing GTO, or learning to play in a GTO style, is all about memorization. The idea is that you’re supposed to study a ton of poker charts, remember what to do in every situation, and then just recall that information at the table to copy the correct strategy. But I think that’s a huge misconception.

First of all, it’s simply not possible. Just take one single hand on a single board — say, the button versus the big blind. Even in that one spot, there are so many different hands you could have. Then think about every decision point: I could bet small, bet big, or check. My opponent could check-raise small or big. Each of those is a new node in the game tree.

Now imagine trying to memorize the optimal strategy for every single combo at every one of those nodes — hundreds of combos in each case. Then multiply that by every possible position matchup, by every stack depth, by every stage of a tournament, and by different ICM pressures. Trying to memorize all of that is just not realistic.

That’s why, instead of trying to memorize everything, I think the real value in studying GTO lies in learning the principles. It’s about recognizing patterns in the data that deepen your theoretical understanding of the game. These broader concepts, or these heuristics, can then be applied flexibly to new situations. 

To me, that’s one of the most fascinating things about poker. When a poker solver analyzes a hand, it plays out billions and billions of hands, experimenting with different strategies. Eventually, it converges on an equilibrium — a state where both players understand each other’s strategies and know that deviating would result in a worse outcome.

As humans, who play poker, we don’t have the luxury of those billions of trials. We have 5, maybe 10 or 30 seconds to come up with a strategy on the fly — just one shot, using only our brain. That’s the strategy we end up playing.

They don’t understand the psychology behind the game. They don’t factor in individual player styles of opponents or pick up on physical tells. But our brains do. Ideally, we use that real-time information and blend it with our theoretical knowledge to create a unique solution in each hand.

That’s what I find really fun about poker. For me, GTO, or the theoretical side of poker, adds even more complexity and enjoyment to the game I already love. Poker has always been about solving problems and figuring things out as you go.

Now, the second biggest misconception I often hear is that GTO poker is boring. I think the idea that GTO poker is boring is just completely false.

First of all, no one is actually playing true GTO. Even the players who are perceived as playing close to game theory optimal — well, they’re really not. They’re not putting in nearly as much volume or precision as a solver does.

When you run training simulations against a solver — say, against George — you constantly find yourself in incredibly tough spots that most human opponents just won’t create. You’ll see overbets, unusual lines, and really creative plays that go far beyond standard human decision-making.

Personally, I’ve learned a ton of novel and interesting poker strategies just from studying solver outputs. One example is what we call a demi-bluff — a concept I hadn’t even considered before using solvers.

Let’s say you have a bottom pair. You make a bet that might get a fold from hands like a middle pair, but still gets called by ace-high flush draws. Then on the river, you check back. So was your turn bet a bluff or a value bet? In reality, it was both.

Before I studied solver outputs, that kind of hybrid line wasn’t even on my radar. I didn’t realize a bet could be both a bluff and a value bet at the same time. So yeah — I actually find solver-based poker really fun to play and watch. It’s not boring at all.

The third common misconception I hear a lot is that GTO doesn’t apply to the games people actually play. The logic usually goes: “No one plays like that, so why even bother studying these strategies?”. But that mindset misses the point.

In fact, identifying situations where people deviate from solver-approved strategies is the first step to exploit them. If you don’t know a particular play is a mistake, you won’t know how to best respond to it.

The core principles you learn from studying GTO can be applied broadly. Like I mentioned earlier, we as humans take in much more information than solvers do: recent history, opponent tendencies, physical reads, and more. While solvers simplify the game, we can combine solver-based insights with real-world reads to make smarter decisions.

To give an example, I pulled up a common spot:

Board:    | Situation: Lojack vs. Big Blind | Single-raised pot (SRP)

The big blind checks their entire range, which makes sense: it’s a bad flop for the big blind and a strong flop for the lojack. The solver recommends a larger c-bet size from the lojack, with a fair amount of checking mixed in.

But what I often see is people just auto c-betting their whole range small in this spot. Understanding that a bigger c-bet captures more EV in theory helps you recognize when opponents are misplaying, and that gives you room to exploit.

For instance, notice how pocket pairs below the king prefer to check. If your opponent is checking with those hands too frequently, you can plan for that — check back the flop and target them with well-timed bluffs later in the hand.

Also, it's important to remember that eventually, you will face players who are using solver-informed strategies. When that happens, you want to be ready — you’ll already know how to respond, and you won’t be caught off guard.

On this same board, if the turn comes the , the solver now has the big blind leading with almost half of their range. There’s not much raising, but this lead frequency is something many players miss entirely. The reason for the big blind leading so often on the turn here is simple: they have 18% flushes. 

With the ace and king on the board, those cards block a large chunk of the suited combos the lojack could have opened with. So even though the overall equity of the big blind’s range might be lower, they can still carve out some strong leads with their flushes.

In fact, the solver shows the big blind still has a pretty decent equity advantage on this turn. But here’s the thing — people will often say: “Okay, no one in my games leads the turn like this. Everyone just checks”. And if that’s the case, it’s actually a huge opportunity.

Let’s say you c-bet the flop, they check-call, and now the flush completes on the turn. If they check again, you now know they’re likely underleading despite holding three times as many flushes as you do — they have 18%, you have only 6%.

The solver’s strategy here shows what to do if they do lead with half their range and have the flush advantage. But against real players who just check, their check-back range is likely stronger than the solver assumes, because they’re not protecting it with leads.

So what’s the adjustment? You slow down on the turn. Some of your thinner value bets, like   or  , might not be worth betting anymore. Instead, you might just pure check those hands.

Hopefully walking through this board and the strategy has shown how you can take solver-based ideas and use them in real games, even when your opponents are far from optimal. And if you had any of these GTO misconceptions before, I hope this helped clear them up. Thanks for joining, this has been Steven Chidwick — see you next time!