Poker tells: weak evidence you can test, not a mind-reading trick

Poker tells became a genre because the pictures are fun: sunglasses, shaking hands, a stare. The base rate is the problem. Strong hands and weak hands both tank. Both look nervous. Both try to look strong. If you act on the first unusual gesture, you are fitting a story to noise. The Poker guides pages on strategy will make you more money than a catalogue of faces, and poker tips is the wider habits list. This page is only the signal problem.

Treat a tell like a small piece of evidence that updates a decision you could already justify with the cards and the position. If the cards say fold and the tell says call, fold. If the cards say the call is close and this opponent has shown the same timing three times with a bluff, the tell can tip it. That is the whole legitimate use.

Bluffing in poker explains when a bet should work as a bluff. A tell does not replace that math. It sometimes tells you whether this opponent is capable of the bluff the math describes.

Worked example 1. You are on the river with a medium pair. The pot odds say a call is slightly losing if the opponent bluffs less than a third of the time. They have snap-shoved twice tonight and shown the nuts both times, and this shove was also a snap. The timing matches the value hands, not a movie stare. You fold. You did not “read their soul.” You used two showdowns.

These are patterns to write down, not reflexes. Any of them can be reversed by a player who has read the same list.

1. Betting faster with strong hands and stalling with bluffs, or the reverse, but stable for that person.
2. A bet size they use only when value-betting, and a different size when they give up or bluff.
3. Glancing at their chip stack before a value bet, as if pricing the raise.
4. Covering hole cards with a hand, the old “protecting a monster” cue, which many players now fake.
5. Speech: a sudden story, or a sudden silence, that differs from their baseline chat.
6. Breathing that changes when the river card hits their draw, then a bet.
7. Looking back at hole cards when a straight or flush completes, because they forgot what they held.
8. Hands that shake while betting a big value hand, which is adrenaline, not always fear. Shaking while bluffing happens too. Do not quote the proverb as science.

None of these works on a stranger for one orbit. They work, a little, on a regular whose showdowns you have watched. If you cannot name the last hand they showed, you do not have a tell. You have a vibe.

1. Snap-checks on a missed flop and longer checks when they are trapping, if the pattern survives showdown.
2. Bet sizing that jumps only on the river, a common bluff or a common thin value, and you will not know which until they show.
3. The same size for every bluff and a rounder size for value, or the opposite. The sameness is the tell.
4. Timing that collapses when they are playing several tables and only slows down in the big pot.
5. Chat that appears only when they are comfortable, which often means a made hand, and chat that vanishes when they are bluffing. Again, reversible.
6. A tank that happens only when they are facing a bet, never when they are the bettor. That player is deciding a call, not designing a movie.
7. Changing the timing after they have been caught. A player who was snapped off may start tanking with the nuts to look weak. Update the file. The old tell died.

Online you also get the absence of a tell. A player who uses one size on a timer is giving you sizing data, not a mood. Poker for beginners covers the streets if those are still fuzzy. Tells sit on top of that structure.

Worked example 2. An online regular bets 33 percent on every dry flop and 75 percent on every river, in under three seconds. You have seen the river size with a missed draw twice and with a strong hand once. The size is not a tell yet. The speed is not a tell. You need more showdowns or you play straightforward. Collecting the data is the skill. Guessing on hand four is not.

A tell is a change from a baseline. If you do not know how long this player normally takes to call a flop bet, a ten-second tank means nothing. Watch a full orbit with the intention of doing nothing clever. Note their average snap, their average tank, and the bet sizes they like. Then wait for a showdown. Write one sentence: “Snap river bet was the nuts,” or “Tank-shove was a missed flush.” One sentence is a hypothesis. Three sentences in the same direction are a pattern. One sentence in the opposite direction deletes it.

Do this for opponents you play often. A tourist you will never see again is entertainment, not an edge. The same method catches your own leaks. If your bluffs take longer than your value bets, the table will notice. Time yourself for an hour.

Physical tells in a loud room are mostly worthless. Hoodies, headphones, and a card cap are not morality. They are a way to have fewer moving parts. The moving part that remains is your bet size. Make that the thing you choose on purpose.

Pick a routine and keep it when you are strong, weak, and indifferent. Look at your cards once. Put a chip on them. Look at the board. Decide. Act in the same motion. Use a similar amount of time for a value bet and a bluff, which usually means slowing the value bet down rather than speeding the bluff up. Speeding up is what makes the bluff obvious.

Do not narrate. “I guess I have to call” is a sentence weak hands love. Silence is not a tell if you are always silent. It is a tell if you are chatty except when you bluff.

Keep bet sizes tied to the plan, not the pulse. If your bluff is always a weird number and your value bet is always a round one, you have published a legend. Choose sizes before the session, the way you choose opening ranges, and let the hand pick from the list.

Online, the routine is the clock. Use a two-to-four-second floor even when the fold is obvious, if you can do it without ruining the game, or accept that your snap-fold gives away nothing of value because the hand is over. Do not snap the nuts and tank the bluff. That one is free money for anyone paying attention.

The goal is not to become unreadable in some mystical way. The goal is to remove the extra information so the opponent has to play the cards.

If you cannot check these off, ignore the gesture.

• You have a baseline for this player from earlier hands, not from a movie.
• At least one showdown lines up with the pattern you think you see.
• The card decision is already close. The tell is a tie-break.
• You are not ignoring a bet size that clearly means value.
• You could explain the tell in one sentence without the word “energy.”
• You have a plan for your own timing on the next bluff so you are not the easier mark.

Then act. If the showdown contradicts you, delete the tell in public, at least inside your own notes. Stubbornness about a read is more expensive than the read was ever worth. Poker is still the cards, the position, and the price. Tells are a thin extra column.

Ignore them against someone you just met, and ignore them when you are tilted. Ignore them when the board already says the range is strong. A feeling you will not write down is not a tell.

A pure chance game does not offer this hobby. Coinflip has no timing tell that changes the next flip, and no face to study. That is a feature when you want rest from reads. When you return to poker, go back to baselines. Fifteen tells on a list are fifteen hypotheses. The ones that really work are the ones the same opponent repeats after the cards are turned up, and those are usually the clock and the bet size.