Reading a Solver's Frequencies: What the Mixes Actually Mean

Open any solver output and the first thing you notice is that it rarely says "always." Instead you get a hand that bets 68% of the time and checks 32%, or a call that's really 55% call, 30% raise, 15% fold. To a lot of players this looks like the solver hedging — or worse, like noise to be ignored in favor of the one "biggest" action. Both readings are wrong, and both cost you EV. A mixed frequency is not indecision. It's a precise instruction about how a hand should behave inside a range, and learning to read it is one of the highest-leverage study skills you can build.

This article unpacks what those percentages actually mean, why the solver mixes at all, and — most importantly — how to translate a screen full of frequencies into decisions you can make at the table without a computer in front of you.

What a Frequency Actually Describes

The first mental correction: a solver frequency almost never describes what to do with a single combo on a single occasion. It describes how a strategy behaves across your whole range over the long run. When the solver says a hand bets 70%, it is not saying "this exact hand is 70% of a bet." It's saying that for the strategy to stay balanced, this category of hand needs to appear in your betting range 70% of the time and your checking range 30% of the time.

Why does that distinction matter? Because balance is a property of ranges, not individual hands. Your opponent doesn't see your cards — they play against your whole range. If every hand that could bet always bets, your checking range becomes a puddle of weak hands that any thinking opponent can attack relentlessly. Mixing exists to keep both of your ranges — the betting one and the checking one — protected. If this idea is new, our primer on thinking in ranges, not hands is the foundation this whole article stands on.

Why the Solver Mixes at All

A pure strategy — always bet, never check — is only correct when one action dominates so completely that the opponent can't punish its predictability. That happens sometimes, and when it does, the solver will show you a clean 100%. But on most interesting spots, two actions are close in value, and the solver mixes for a specific reason:

• Indifference. At equilibrium, the solver makes the opponent indifferent between their options. To do that, it has to show up with the right blend of strong and weak hands in each of its own lines. Mixing is the mechanism that builds those blends.
• Protecting both ranges. Sending some strong hands to the checking line stops that line from being pure air. Sending some weak hands to the betting line gives your value bets cover. Mixing populates both.
• Near-equal EV. When betting and checking a hand are worth almost the same, the exact split barely changes that hand's own EV — but it changes how exploitable your range is. The solver spends its "free" decisions on hands where the choice is close, buying range protection at almost no cost.

That last point is the key to using frequencies in practice, and it leads directly to the most important reading skill.

The Skill That Matters: Read Magnitude, Not Decimals

No human reproduces a 68/32 split at the table, and you shouldn't try. Chasing the exact decimal is the single most common way players waste solver time. What you actually need is the direction and the strength of the mix. Read frequencies in three buckets:

• Near-pure (90%+ one way): treat it as "basically always." A hand that bets 93% is a bet. Don't get cute.
• Heavy lean (roughly 65–90%): this is your default action, with the minority action available as a deliberate mix-up or when a read pushes you. Bet this hand most of the time; occasionally take the other line to stay unpredictable.
• Genuine coin-flip (40–60%): the solver is telling you both actions are fine. Here you're free to deviate — pick the action your opponent's tendencies favor and lose almost nothing in EV.

Once you see mixes this way, a wall of percentages collapses into a handful of usable rules. You're no longer memorizing numbers; you're reading how committed the solver is to each action.

Turning a Mix Into a Table Decision

So you're in a spot, and study told you this hand class mixes around 60% bet / 40% check. What do you actually do with one specific hand, one time? Three practical approaches, in order of how most strong players operate:

1. Break the tie with a read

A near-even mix is the solver saying "I'm indifferent — you decide." That's an invitation to be exploitative. Facing a station who never folds? Push the value hands in the mix toward betting. Facing someone who over-folds to aggression? Bet your bluffy candidates more. The solver hands you the equilibrium; your reads tell you which way to lean off it. This is exactly the handoff we cover in GTO vs exploitative.

2. Use a real randomizer for the close ones

When you have no read and the mix is genuinely close, you can randomize — the second hand on a watch, the suit of your hole cards, a spot on the felt. In practice this matters far less than people think, because if the two actions are near-equal in EV, guessing wrong costs you almost nothing. Save the mental energy for spots that aren't close.

3. Simplify to a heuristic

The endgame of studying frequencies is not carrying them to the table — it's compressing them into a rule you can recall mid-hand. "On this texture I bet my top pairs and my best draws, check the middling stuff" is a heuristic distilled from a hundred individual frequencies. That compression is the whole point of study; if you want the full method, see how to study with a solver without wasting hours.

Frequencies Across a Range vs a Single Hand

There's a second layer worth understanding: aggregate frequencies. When a report says "the aggressor bets this whole board 76% of the time," that's a different number from any individual hand's mix — it's the sum of every hand's decision, weighted by how often each appears. Aggregate frequencies are how you spot big-picture patterns: which board classes get bet often, which get checked, how sizing shifts by texture. Individual mixes tell you what to do with this hand; aggregate frequencies tell you what your strategy looks like from across the table. You need both, and good reports let you move between the two — from "how does my range play this board type" down to "what does exactly this combo do here."

Common Misreads to Avoid

• "Just take the bigger number." Collapsing every mix to its majority action strips out the range protection the mix was buying. Fine for near-pure spots; costly when you do it to every 55/45.
• Memorizing decimals. 68 vs 62 is noise at the table. The bucket — heavy lean vs coin-flip — is the signal.
• Treating a mix as one hand's hedge. The percentage is a range-level instruction, not a probability that this specific hand is good.
• Ignoring why the minority action exists. The 15% raise in a mostly-call spot is usually your strongest and your bluffiest hands. Knowing which hands take the minority line matters more than the percentage itself.

The Bottom Line

A solver's frequencies aren't the tool being unsure — they're the tool being exact about something humans find unintuitive: that individual hands exist to serve the range they belong to. Read mixes by magnitude, not decimals. Treat near-pure numbers as "always," heavy leans as defaults with an escape hatch, and true coin-flips as free space to exploit a read. Then compress what you learn into heuristics you can actually recall. Do that and a screen full of intimidating percentages turns into a short, usable set of rules — which is exactly what solver study is supposed to produce.

Ready to go deeper? The natural next steps are understanding value-to-bluff ratios (the theory behind why those minority actions are populated the way they are) and node locking (deliberately unbalancing the solver to attack a specific opponent once you understand the balanced baseline).