Heads-Up Poker Reads: Bet Patterns and Timing Beat Body Language

The defining feature of heads-up poker is information density. At a nine-handed table you face the same opponent once every ten hands. Heads-up, you face them every hand — an hour builds a bigger sample than a full night at a full table.

The problem is that most players spend that advantage in the wrong place: watching a face for expressions instead of recording how the person actually bets.

Build a baseline first, or everything is noise

“His hands are shaking, so he is nervous” carries a hidden assumption: that you know his hands do not normally shake.

Observation without a baseline is guessing. Some people naturally talk fast, some have hands that always shake, some take a long time over every decision. Reading a personal trait as a live signal is worse than not watching at all, because it gives you confidence in a wrong conclusion.

Heads-up, the baseline builds quickly. Spend the first twenty to thirty hands recording rather than judging:

• What size does he open to? Is it always the same?
• Roughly how often does he fold on the button?
• After he checks and faces a bet, what does he usually do?
• Does his decision speed change noticeably between spots?

Those answers are the baseline. From then on you are not looking for a particular gesture — you are looking for how far he has moved from his own normal.

Most reliable: betting patterns and sizing

Behaviour patterns beat physical tells for a practical reason: they leave a record, they accumulate, and they can be checked.

The pattern deviations that matter most heads-up:

Sizing that tracks hand strength. Many players bet bigger with strong hands without noticing, and smaller when bluffing because they want to lose less. If his flop continuation bet is normally half pot and one hand it jumps to eighty percent, the difference itself is information.

Check-raise frequency. Most recreational players check-raise almost exclusively with strong hands, and it is one of the easiest tendencies to exploit heads-up. Conversely, an opponent who never check-raises lets you bet at a high frequency in position with confidence.

Response to a raise. When raised, does he fold quickly, call, or re-raise? The distribution of those three answers says far more about his range than any expression.

One caveat: an opponent using a mixed strategy will deliberately take different actions in identical spots. Against that player, pattern reading loses value, and the answer is to fall back on ranges and maths rather than hunt harder for clues.

Second most reliable: decision time

Timing is the hardest signal to hide completely, because thinking genuinely takes time.

The useful read is not “fast means strong, slow means weak”. It is the relative difference within one repeated spot:

• An instant check: usually means he gave up on this street some time ago, and leans weak.
• An instant bet or call: normally a pre-planned action. It can be a very strong hand or a planned bluff, so read it together with the sizing.
• A long think, then a call: typically a marginal hand doing the arithmetic. Those hands rarely survive pressure on the next street.
• A long think, then a big bet: heads-up this is often a genuinely strong hand working out how to extract the most, not hesitation.

Physical tells: less reliable than you think

The standard list — shaking hands, dilated pupils, changed breathing, leaning forward — has three problems.

First, most of these are stress responses, not strength responses. Shaking comes from adrenaline, and adrenaline arrives with big hands and big bluffs alike. It tells you the hand matters to him. It does not tell you which way.

Second, the environment interferes. Pupil size follows the lighting; breathing follows temperature and mood. Treat these as signals and you will often be reading the room rather than the player.

Third, the moment an opponent knows about them, they get used against you. Experienced players deliberately look strong while bluffing. The more weight you put on physical tells, the more a reversal costs.

The sensible position is this: use physical tells as corroboration, never as the primary basis. When the betting pattern already points somewhere and the behaviour happens to agree, take a little extra confidence. Overturning a mathematical read on the strength of one gesture loses money over time.

Online heads-up leaves two dimensions

Online there are no faces and no chip handling. Only timing and sizing remain.

That sounds like a loss, but it actually makes reading cleaner, because the two that survive are the two that were reliable in the first place. What to do online is specific:

• Record the sizing distribution. What sizes does he use for opens, continuation bets and river bets? The fewer distinct sizes an opponent has, the easier he is to play against.
• Watch for timing anomalies. Online timing is measurable to the second, but rule out connection lag and multi-tabling first — someone playing six tables may be “thinking” on another one.
• Use the hand history. Reviewing his actual ranges after the session is more accurate than any live read. Long-term reading ability is built in review, not invented at the table.

The other direction: what you are giving away

Observation runs both ways. While you build a sample on him, he is building one on you.

The three easiest things to get caught doing: tying bet size to hand strength, tying decision time to hand strength, and only taking a particular action in a particular spot — check-raising only with strong hands, for instance.

The fix is not acting. Acting is hard to sustain and eats the attention you need for decisions. The practical answer is to standardise your behaviour: use a small set of bet sizes, give yourself a consistent thinking rhythm, and let the action itself carry no extra information.

Common mistakes