What a poker solver prints: ranges, frequencies, and EV

A study program for Hold'em takes the game you describe and searches for a strategy pair neither player can improve inside that description. PioSOLVER's viewer manual treats the range as the basic object: the hands a player might hold, stored as a relative probability rather than as the two cards dealt.

GTO Wizard's 2023 article calls these programs EV-maximizing routines that stop when neither side can improve, and it names that stop equilibrium. PioSOLVER's quick start shows a free install and a commercial Pro or Edge install that asks for a licence code. This page will not supply a code, a crack, or a step that skips payment.

Simple Postflop's product page describes a two-player Hold'em program that shows a strategy, an EV, and equity for a tree you build. The page calls that product the most accurate of its kind. That line is marketing we did not test.

You can read equilibrium play notes if you want the idea of an unexploitable mix. This page does not reteach that idea as a betting system. The poker topic hub collects the rest of the cluster.

Adults only. If a study session turns into chasing a loss, stop and call 1-800-GAMBLER. Responsible gambling is the longer limit.

PioSOLVER's FAQ says a range weight is the relative chance of holding a hand. If aces have weight 1 and kings have weight 0.5, aces are twice as likely as kings in that range.

A strategy is a different object. It is the mix of actions for one hand, and the actions at a single node add up to 1.

Their FAQ walks a check-raise without a chart. AJs at weight 0.75 that checks 70% and then raises 40% has a check-raise weight of 0.21. That 0.21 is a weight after two decisions, not a claim you win 21% of the pot.

The viewer can show combo counts, such as 60 combos checking and 40 betting, using only one player's range. Real frequency also uses the other range and card removal. In their sketch, hero checks every king of spades and bets every ace of spades against a range that is only the ace, so the check is half the combos and the whole real line.

GTO Wizard's Study Mode help says EV is the average return of a hand or an action. Their example is a $1 coin flip with an EV of $0.50, which is not the same figure as equity if the hand is checked down. Their longer article writes that average as each ending's value times the chance of reaching it, and an ending is a fold or a showdown.

PioSOLVER's FAQ adds a unit warning. Calling $100 and then folding can show as −$100 in global EV, while a viewer shows 0. Average EV also has to be weighted by matchups, their word for one weighted hand against another, because aces are dealt more often than sevens.

Two-player cash decisions are where frequencies get discussed as play. This page stays on the print itself.

PioSOLVER's quick start says the browser can show EVs, strategies, and ranges after you build a tree and stop it yourself. That page's timing notes are theirs. We did not time a run on our machine.

GTO Wizard's article lists the inputs: betting tree, target accuracy, pot, stacks, ranges, board, and any rake or ICM change. A size you leave out is a line the program cannot choose, and they say a tree can also be too big to finish. They do not claim every chip increment of no-limit.

Simple Postflop's page covers strategy, EV, and equity for two players in no-limit or fixed-limit Hold'em, with a free turn-and-river option and a paid standalone licence. The dollar price was not in the text we read, so it is not published here.

The page lists Windows, at least 8 GB of RAM, and a Windows layer on Mac. We did not install it.

Calling any of these a poker solver is fair only when you mean that print: ranges, frequencies, and EV for a tree you specified. A crypto-casino primer is a different product. The tool does not hold a balance.

The print ends where your inputs end. A different stack, rake, or open size is a new tree, and one flop's colors do not move onto the next board.

Combo percent is not the rate the line happens once the other range removes cards. The EV is the average inside the paired strategies from that node forward, not the cash in your pocket after one session.

GTO Wizard's article says practical runs stop short of a perfect equilibrium, and that their solves target 0.1% to 0.3% of the starting pot depending on the solution type. That target is theirs. We did not reproduce it.

A two-player tree does not describe a pot with more callers. Simple Postflop's page is explicit that the solve is two players, so a multiway pot is outside that page.

Equilibrium is a statement about both sides playing the print. If the other person folds too often, a best response can leave the mix. Adjusting for that leak is a different study, not a script that turns a node's EV into a night's profit.

Studying one poker solver node is a reading exercise with a stop rule. You should be able to say the stack, the pot, and the bet sizes out loud. A screenshot missing those three is an incomplete quote.

Hardware and a licence are part of the limit. Pio's commercial quick start asks for a licence code. We will not describe a pirated build, a shared key, or a torrent.

A house game charges you inside the price of the bet. A PvP pot charges a fee on the combined money. A study program charges a vendor, when it charges anyone, and then hands you a grid.

A poker solver is the left column. It does not settle the middle column or the right column.

Two $50 Coinflip stakes make a $100 pot. The fee is $5, and the winner is paid $95. Four $50 Jackpot entries make a $200 pot, the fee is $10, and the winner is paid $190.

Roulette on Purple or Silver is about 7.88% after the fee. Slott takes 10% of the winning pot. Those figures are in the terms, and a node's EV does not refund them.

The built-in cut is the house column as one number. A book's margin is the same idea on a sports price. Seed-checked play is how a hashed game is audited, and none of those pages is a solve.

Poker is the table on this site. Fairness checks a seed. A vendor grid does not audit that seed.

Example one is theirs. GTO Wizard's Study Mode help says a coin flip with a $1 price has an EV of $0.50, because half the time the result is $1 and half the time it is $0. You do not receive fifty cents on the flip that loses.

Example two is PioSOLVER's unit warning. Calling $100 and then folding is −$100 in global EV, while a viewer can show 0. The minus one hundred dollars is money the line gave away inside the tree, not a fee the program adds.

Example three is our pot, not a solve. Risk $50 on Coinflip against another $50 and the winner receives $95, because the missing $5 is the 5% fee. A small positive node EV does not pay that $5 back, and it does not forecast the next session.

Before you pay a vendor, or trust a screenshot, walk this list. A poker solver screenshot without the tree beside it is an incomplete quote.

• [ ] I can name the stack, the pot, the board, and the bet sizes.
• [ ] I know whether the percent is combos or real frequency.
• [ ] I know whether the EV is the global figure or the viewer's number.
• [ ] I am not treating a mix as a promise of profit.
• [ ] The spot is two players, or I have written that it is not.
• [ ] Rake or ICM is either inside the tree or marked absent.
• [ ] I will not use a cracked copy, a shared key, or a torrent.
• [ ] Any licence price I remember is on the vendor's own page.
• [ ] Money I later play with can go to zero. 1-800-GAMBLER is saved.
• [ ] I opened the responsible-gambling page before I sized a real bet.

Definitions for the three outputs sit on the vendors' own pages, including the split between combo frequency and real frequency. A free Pio install and a free turn-and-river option are named by the vendors.

A missing bet size is a missing answer, a two-player solve skips extra callers, and we did not test the accuracy targets they state. A cracked copy is outside this page, and one node's EV is not a session result.

Who should stay away: minors, anyone who wants a winning script, anyone who will pirate a binary, and anyone who cannot separate a node's average from money they can lose. If study turns into chasing, stop and call 1-800-GAMBLER.