Moderate costTracked stat: GTO trainer accuracy

Bet sizing tells: your sizes are talking

Last updated

A bet-sizing tell is a reliable correlation between how much you bet and what you're holding — pot with monsters, half-pot with medium strength, tiny with give-ups, overbet with draws. Observant opponents don't need to see your cards; the number does the talking.

Holdem Pro rates this a Moderate-cost leak: nearly free against oblivious players, brutally expensive against anyone paying attention.

The numbers on this leak

Cost rating
Moderate
Tracked stat
GTO trainer accuracy
Flagged at
40%
Fix target
≥ 70%

Live stat lines can't see your hole cards, so this leak is tracked with GTO trainer accuracy — how often your sizing matches solver-approved play across drilled spots. Flagged around 40%; the mastery target is 70%+, over at least 10 attempts. Measured over 10+ drill attempts.

Fix protocol (drilled hands)
45 → Improving·175 → Stable·500 + 120 live → Holding

Drills alone can never reach the top stage — your live GTO trainer accuracy has to move and hold, and stages regress if it slides back.

How are opponents reading my bet sizes?

They only need a few showdowns. Your 75%-pot bets showed value twice; your third-pot bets showed giving up once — that's enough. From then on they raise your small bets, fold to your big ones, and call your draws down. You get maximum action on your bluffs and none on your value: the exact reverse of what betting is for.

The leak compounds street by street, because a sizing tell on the flop narrows your range for the turn and river too. By the river, a good reg is effectively playing with your cards face up.

What should decide my bet size, if not my hand?

Board texture and range — the two things your opponent can also see. On a dry king-high flop, your whole continuing range can bet small, because nothing needs protection and worse hands still call. On wet, connected boards, everything bets bigger. When the size is a property of the board instead of your holding, there is nothing left to read.

How do I fix a sizing tell?

  1. 1
    One size per texture

    Pick a default sizing per board type and use it with your entire range there. Same board, same size, whether you're value betting or bluffing — that single rule deletes the tell.

  2. 2
    Run the 100-hand drill

    Choose three sizes — 33%, 66%, pot — assign each to a board type, and stick to the assignment for 100 hands. Discomfort is the point: it's the feeling of your sizes decoupling from your cards.

  3. 3
    Size for the range you rep, not the hand you hold

    Before betting, ask what size this board wants from both your value hands and your bluffs. Use that one. If a size only makes sense for one half of your range, it's a leak.

Coach's drill for this leak: Pick 3 sizings (33% / 66% / pot) and assign each to a board type. Stick to it for 100 hands.

Two hands, one number

On K-7-2 rainbow you c-bet a third of the pot — with ace-king, with pocket fives, with ace-five as a backdoor bluff. Same board, same number, whole range. Your opponent's read after ten of these: nothing. The size carries no information because it never varies with your holding.

Now the leak version: AK bets 75% "for value" and A5 bets 33% "as a cheap bluff." A decent reg needs two showdowns to decode that pattern — then he folds to your big bets, attacks your small ones, and every bet you make afterward loses money it didn't have to.

How long does this take to fix?

Longer than a tips video, shorter than you fear. Holdem Pro's mastery ladder is deliberately honest: about 45 drilled hands of focused practice reaches Improving (roughly three short sessions), 175 reaches Stable, and Holding takes 500 drilled hands plus at least 120 live hands played after the improvement shows up — because drills alone don't prove a leak is closed.

The real bar is transfer: to advance you have to move your GTO trainer accuracy from around 40% up past 70% in actual play, and the stage drops back if the stat regresses. Closing a leak for real takes weeks, not days — which is exactly why most opponents never do it.

Which player types have this leak?

The Wolf · TAGPatient, deadly, picks spots.The Surgeon · SharkPrecise, balanced, hard to read.

Do you have the bet sizing tells leak?

This page describes the leak in general. The quiz deals you 7 real hands and measures your version of it — how bad, where it shows up, and what to drill first.

Find my leak — free

Free · 2 minutes · 7 real hands · no signup

Common questions

How do opponents pick up sizing tells?

Showdowns. Every time your cards get revealed, the size you chose gets attached to the strength you had. Two or three data points is all a competent player needs to start exploiting the pattern in both directions.

What default bet sizes should I use?

A solid starter kit: about a third of the pot on dry boards, two-thirds on wet ones, with the same size used for your whole range on that texture. Consistency within a spot matters far more than the exact number.

Isn't varying your bet size good?

Vary by board and by range — never by hand strength within the same spot. Betting bigger on wetter boards is strategy; betting bigger with stronger hands on the same board is a tell.

How is this leak measured and fixed?

Live stats can't see your cards, so Holdem Pro tracks it through GTO trainer accuracy — from a flagged ~40% toward the 70% target, over at least 10 drill attempts, plus the 45/175/500-hand mastery ladder.

Keep going