RangeSharp
RangeSharp
What GTO means in poker, how solvers find optimal strategies, when to follow GTO, and when to deviate.
"GTO" gets thrown around a lot in poker. Most players have a vague sense it means "play like a solver." Few can explain what it actually is or why it matters.
Here's the short version: GTO stands for Game Theory Optimal. It's a strategy that can't be exploited. No matter what your opponent does, you don't lose more than you have to. In practice, that means playing a mixed strategy with specific ranges and frequencies that balance each other.
Imagine you play rock-paper-scissors. If you always throw rock, a smart opponent will always throw paper. You're exploited. If you randomize 1/3 rock, 1/3 paper, 1/3 scissors, no opponent can do better than break even by adjusting. You're unexploitable.
Poker is more complex, but the idea is the same. A GTO strategy has the property that if your opponent plays perfectly, you break even (or lose only the minimum dictated by the rules, like the blind structure). If your opponent makes mistakes, you profit. You never get crushed by someone who "figured you out."
That doesn't mean GTO is the most profitable strategy. An exploitative strategy (one that targets a specific opponent's leaks) can make more money. But it can also get crushed if the opponent adjusts or if you misread them. GTO is safe. It's a floor. You might not maximize vs a particular player, but you don't get run over.
Solvers are programs that compute Nash equilibria. They simulate the game, enumerate strategies, and find the strategy profile where no player can improve by changing their play. That's GTO.
The solver doesn't "know" poker. It iterates. It tries different actions, computes expected value, and adjusts until nothing improves. The output is a set of ranges and frequencies that form an equilibrium.
Preflop is simpler than postflop. Fewer decision points, fewer branches. So preflop GTO solutions are more stable and easier to compute. Postflop, you need to simplify (e.g., limit bet sizes, restrict certain lines) to get tractable solutions. Preflop, you can get pretty close to "true" GTO for common configurations.
Some players say: "My opponents don't play GTO, so why should I?" That's a confusion.
GTO is a baseline. When your opponent deviates, you can exploit them. But you need a reference. If you don't know what balanced play looks like, you can't identify when someone is over-folding, over-bluffing, or playing too many hands. GTO tells you what "normal" is. Deviations from that are opportunities (or warnings that you're the one being exploited).
Also, in tough games, good players will probe for leaks. If you're way off GTO, they'll find it. Learning GTO shores up your weaknesses so you don't get targeted.
GTO: play the equilibrium. Ignore who's across from you. Maximize robustness.
Exploitative: play the best response to what you think your opponent does. If he folds too much to 3-bets, 3-bet more. If he calls too much, value bet thinner. You're targeting his mistakes.
In practice, you blend both. Default to GTO. When you have a read (from stats, history, or live tells), shift toward exploitative adjustments. The better your information, the more you can deviate. In an anonymous pool with no history, GTO is usually the best you can do.
GTO often uses mixed strategies. A hand might raise 70% and call 30%. That's not "sometimes raise, sometimes call randomly." It's "at equilibrium, you should raise 70% of the time and call 30%." The mix makes your opponent indifferent. If you always raised, he could exploit that. If you always called, same thing. The mix removes his edge.
For memorization, mixed strategies are annoying. You can't just say "raise" or "fold." You have to know the frequency. Most players round: if it's 80% raise and 20% fold, they just raise. The EV loss from that simplification is small. If it's 55% raise and 45% call, rounding is sloppier. Those spots need more care.
Several situations justify deviations:
Softer games. If your opponents fold too much, open and 3-bet more. If they call too much, value bet more and bluff less. GTO assumes a competent opponent. Against fish, you can push the edges.
Different stack depths. Solvers are run for specific stack sizes (often 100bb). At 50bb or 200bb, ranges change. Shorter stacks favor all-in or fold. Deeper stacks favor more postflop play and wider ranges.
Rake. High rake shrinks ranges. Low rake (or no rake) widens them. Solver outputs usually assume a rake structure. If yours is different, adjust.
Table dynamics. If the table is passive, you can open wider and steal more. If it's aggro, tighten up and let them bluff into you.
Your own tendencies. If you know you over-fold in certain spots, you might consciously over-correct by calling more. That's meta-exploitation of yourself.
"GTO means never bluff." Wrong. GTO bluffs a lot. It just balances bluffs with value. The ratio depends on bet size and board texture. Preflop, "bluff" means opening or 3-betting hands that aren't for value. GTO does plenty of that.
"GTO is only for pros." No. GTO gives you a coherent structure. Beginners who learn GTO ranges avoid the chaos of "I'll just play what feels right." It's a faster path to competence than trial and error.
"Solvers killed poker." They changed it. Everyone has access to similar information now. The edge comes from implementation (actually playing the strategy), exploitation (adjusting to opponents), and postflop skill (where GTO is harder to compute and apply). Preflop is more solved. Postflop still has room for creativity.
"I need to memorize every frequency." You don't. Approximations work. Know the big blocks (these hands raise, these fold, these mix) and refine the edges over time. Tools like FreeRangeLab let you study GTO ranges and train on them. You can start with simplified versions (e.g., rounded to 25% steps) and add nuance as you improve.
Use GTO as your default. Learn the ranges for your game (6-max cash, 100bb, etc.). Train until they're automatic. Then play. When you notice an opponent doing something strange, consider an adjustment. When you're unsure, fall back to GTO.
Don't let GTO paralyze you. It's a tool, not a master. The goal is to make good decisions. Sometimes that means following the chart. Sometimes it means deviating. The chart gives you a place to start.
Preflop has fewer decision points. You get dealt two cards. You choose: fold, call, raise (and maybe 3-bet, 4-bet, etc.). The tree is manageable. Solvers can compute near-exact equilibria for common configurations (100bb, standard open sizes, typical rake). The output is stable across different solver brands. PioSolver, GTO+, Simple Postflop, they'll give you similar preflop ranges. The differences are in rounding and minor assumptions.
Postflop is different. Every street multiplies the branches. Bet sizes, check-raising, multiple players. Solvers need simplification: limited bet sizes, restricted lines, sometimes heads-up only. The "GTO" you see postflop is an approximation. Preflop, you're closer to the real thing. That's why preflop study has such a high ROI. You're learning something that actually holds up.
You can read GTO ranges and agree with them. Then you sit at the table and fold A5s from UTG because it "feels" wrong. Or you call with KJo from the BB when the chart says 3-bet or fold. The gap between knowledge and execution is where most players bleed.
Training fixes that. Decision drills: see the hand, pick the action, get feedback. Wrong answers create a correction. Right answers reinforce. Over hundreds of reps, the chart becomes automatic. You stop thinking "what does GTO say?" and start thinking "this is a raise" or "this is a fold." That's when GTO pays off. Not when you've read it. When you've internalized it.
Tools like FreeRangeLab are built for this. Add GTO ranges from the library, run spaced repetition drills, and track what you get wrong. The algorithm surfaces your weak spots. You study those more. The cycle compounds. That's how you close the gap.
GTO assumes your opponent is also playing GTO (or close). In a heads-up match, that might be true. In a 6-max game, you're playing against a field. Some will be tight, some loose, some aggressive, some passive. Your "best" strategy is the one that maximizes EV against the mix you're facing. GTO is a reasonable default when you don't have reads. As you get reads, shift. The baseline stays useful. It anchors your adjustments.
GTO is unexploitable play. Solvers compute it. You use it as a baseline, deviate when you have information, and train until it's automatic. Preflop GTO is reliable and learnable. Don't overthink it. Learn the ranges, practice them, and adjust when the situation demands it.
Master your preflop ranges.