In January 2017, a program called Libratus, built at Carnegie Mellon University, spent twenty days playing heads-up no-limit hold’em against four professionals at Rivers Casino in Pittsburgh. It won by a wide margin. Two years later Pluribus, developed by Carnegie Mellon with Facebook AI, beat elite players in six-handed games. Those were research projects, but the same idea was already sitting on players’ laptops: software that calculates strategy from the rules of the game instead of from human habit.
That software is what players call a poker solver. If you have ever watched a high-stakes stream and heard a commentator say a bet “looks like a solver line”, this is what they mean.
What a poker solver actually calculates
A solver searches for a Nash equilibrium, the concept named after mathematician John Nash: a pair of strategies where neither player can gain by changing theirs alone. In poker this is usually called GTO, for game-theory optimal. A true equilibrium for full no-limit hold’em is far too large to compute, so solvers work on a simplified version of one specific spot.
You define that spot. The inputs are:
- the range of hands each player can hold when the spot begins
- the board cards (the flop, or the flop plus turn)
- the effective stack and the pot size
- the bet and raise sizes each player is allowed to use
From those, the solver builds a game tree and runs repeated iterations until the strategy stops improving. It reports how far it is from equilibrium as exploitability, often shown as a percentage of the pot. The result tells you, for every hand in each range, how often it should bet, check, call, raise or fold, and what each option is worth in expected value.
One point surprises beginners. Solver strategies are mixed: the same hand might bet 40 percent of the time and check 60 percent. That is not indecision. It keeps the overall range balanced so that an opponent cannot read the action.
PioSOLVER: the professional standard
PioSOLVER, released in the mid-2010s, became the tool most professionals used to study postflop play. It runs on your own computer, which means you control every input, and that is both its strength and its learning curve. You build the trees, paste in the ranges and decide how much accuracy you want before stopping the calculation.
The editions differ in what they handle. Cheaper versions cover postflop spots, while the higher tiers add preflop solving and scripting tools that let players run hundreds of flops in a batch overnight. Large trees with many bet sizes need a lot of memory, so serious users often run Pio on desktops with far more RAM than a typical laptop.
Pio’s best-known feature is node locking. You can fix one player’s strategy at a point in the tree, for example forcing a player to never bluff the river, and the solver recalculates the best response. That turns a theory tool into a way of studying how to exploit specific leaks.
GTO Wizard: the pre-solved library
GTO Wizard took a different route. It runs in a web browser and gives you access to a very large library of solutions that have already been calculated, covering cash-game and tournament formats at common stack depths. You pick a format, follow the action, and the strategy appears in seconds, with no tree building and no waiting.
That makes it the easiest starting point for most players. Beyond the library, GTO Wizard added a trainer that deals you spots and scores your decisions, hand-history analysis that compares your plays to its solutions, and a custom-solving feature for spots outside the library. Subscription tiers decide how much of that you can use, and the details change often enough that it is worth checking the current plans on the site itself.
Simple Postflop and other alternatives
Simple Postflop, from the Simple Poker suite, is a desktop solver that has long been seen as a more affordable alternative to PioSOLVER. Its interface is friendlier to newcomers, and the same company offers preflop and training tools alongside it. GTO+ is another low-cost desktop solver that many players use for quick postflop checks.
For Pot-Limit Omaha and multiway pots, MonkerSolver has been a common choice, since the large number of possible PLO hands makes those spots much harder to calculate. Preflop and push/fold work in tournaments is often done in separate tools such as HoldemResources Calculator or ICMIZER, which deal with payout structures in ways postflop solvers usually do not. Those charts connect directly to the ideas in the guide to short stack and big stack strategy.
How beginners should study with a solver
The usual mistake is to open a solver and try to memorize what it says. A single flop solution contains thousands of numbers, and nobody can remember them all. Coaches tend to suggest a different approach.
Start with spots you actually play often, such as a single-raised pot where you opened from the button and the big blind called. Look at the overall pattern first: how often the range bets, which bet size it prefers, and which board textures change that. Then look at a few hand classes, like top pair, flush draws and complete air, and ask why each one takes the line it does.
It also helps to know the terms used in solver output (range advantage, polarized, merged, blocker), and the poker terms glossary covers many of them. If you still need the basics of the game itself, start with Texas Hold’em rules for beginners before you spend money on any of these tools.
Finally, keep solvers off the table. Every major site, GGPoker and PokerStars included, bans real-time assistance software, and GGPoker has publicly reported closing accounts for it. Their security teams look for decision patterns and timing that match solver output. A solver that would take minutes to calculate one spot on a fast desktop was never meant to be consulted with fifteen seconds on the clock.