At a glance
- On the HUD
- H
- Formula
finished hands recorded with this player dealt in- Typical range
Not colored on the HUD: it has no typical range, because a good number depends on the game. See “Too low, about right, too high” below.
- Where it shows
- Every open badge: the small number at the end of the top row, such as “412 h”.
- Trust it after
- 12 chances. Below that the number is faded; reads need 12 chances and a number at least one standard error outside the typical range.
What it measures
The hands count is how much history you have on a player. Every hand they’re dealt into, at any table you have open, adds one once the hand is over. It has no good or bad value; it tells you how far to trust everything else.
The trap is reading it as the sample for every stat. It isn’t. VPIP gets a chance almost every hand, so 400 hands is close to 400 chances. Fold to 3-bet only gets a chance when the player opens and someone 3-bets them, which might happen a few dozen times in those same 400 hands. Fold to 4-bet is rarer still. A player showing “412 h” can have a rock-solid VPIP and a fold to 4-bet built on three hands.
So use hands as a rough guide to which kinds of stats are ready, and judge each number by its own count, which its tooltip shows (“9 of 31 opportunities”).
How it’s calculated
A hand counts for a player when both are true:
- they were dealt in (sitting out doesn’t count), and
- the hand is over: the next hand has started, or the table has gone quiet for 2 minutes.
A hand in progress isn’t counted yet, so the number ticks up one hand behind the action. Hands follow the player’s PokerNow ID, not their name, so hands from every game you’ve played with them add up, even if they changed names in between. When you open a table, its earlier hands come in on their own.
Every other stat counts its own chances inside those hands. Example: Jonah has 200 hands. He had a preflop decision in 196 of them (the other 4 were walks in the big blind), so his VPIP and PFR rest on 196 chances. He opened 38 times and faced a 3-bet 9 times, so his fold to 3-bet rests on 9 and is still faded. He was the preflop raiser with a chance to c-bet on 30 flops, so his c-bet rests on 30. Same player, same 200 hands, three very different samples.
Why it matters
Small samples lie, and the hands count is your first warning. Over a few hands, a tight player who picked up aces twice looks loose and a card-dead loose player looks like a rock.
The error in a percentage shrinks with the square root of its chances. For a stat around 30%, one standard error is about ±13 points over 12 chances, ±10 over 20, ±6.5 over 50 and ±4.6 over 100: four times the chances, half the error. Roughly two times in three, the player’s true number sits within one standard error of what you see.
Stats fill at very different speeds. The rates below are round illustrations, not measurements; a loose player sees more flops and a tight one fewer:
| Kind of stat | Chances per 100 hands (example) | 12 chances after | 50 chances after |
|---|---|---|---|
| Every hand: VPIP, PFR | almost 100 | about 12 hands | about 50 hands |
| Common preflop spot: 3-bet, fold vs open | say 20 | about 60 hands | about 250 hands |
| Flop spot: c-bet, fold to c-bet | say 10 | about 120 hands | about 500 hands |
| Rare spot: fold to 4-bet, 5-bet | say 1 | about 1,200 hands | about 5,000 hands |
Results need far more again. Win rate swings so much from hand to hand that thousands of hands still leave it noisy.
Too low, about right, too high
The HUD doesn’t color the hands count; it has no typical range. These bands are ours, a rough guide to what you can trust at each stage.
Thin: under 30 hands
What it says
You know little beyond the broadest preflop habits. VPIP and PFR pass 12 chances after about a dozen hands, but a 30% stat over 20 chances still carries ±10 points. There’s no player type yet (types start at 30 hands), almost every postflop number is faded, and a single memorable hand is a sample of one.
How to exploit it
- Play your normal game and treat them as a typical player at your table.
- Act only on extremes. A 60% VPIP over 20 hands is ±11 points, so even the low end is loose. A 25% vs 30% difference means nothing yet.
- Ignore faded postflop numbers, however dramatic: “100% fold to c-bet” over 2 chances is two folds.
- Pay attention to showdowns. What a player called or raised with is the best information you have this early.
Working sample: 30–300 hands
What it says
Enough for the preflop picture. At 100 hands a 30% VPIP is ±4.6 points, the player type appears at 30 hands and stops being faded at 60. Common spots like 3-bet and c-bet pass 12 chances somewhere in this band, with errors around ±10 points. Rare spots, showdown results and win rate are still mostly noise.
How to exploit it
- Trust VPIP, PFR and the player type for preflop decisions: who to isolate, whose blinds to steal, whose opens to 3-bet.
- Use flop and turn stats only when they sit well outside the typical range. A number just past the edge is still within its error.
- Let the reads guide you. A read sentence already needs 12 chances and one standard error outside the range, so it has passed a basic test.
- Read AF with its count in the tooltip, not the hands count.
Deep history: over 300 hands
What it says
Most common stats are solid, reads show up, and turn and river spots become usable. Two cautions. Rare spots can still be thin: at one chance per 100 hands, a stat has 3 chances after 300 hands. And the history spans every game you’ve shared, so a player who has changed, tightened up or picked up new habits still carries their old hands at full weight.
How to exploit it
- Use the narrower versions. Once a version by position, in or out of position, pot type or heads-up vs multiway has 12 chances, the HUD can show it, and that’s where the precise edges are.
- Hunt for street leaks: turn and river habits fill more slowly than flop stats and only now start to show.
- Watch for change. If today’s play keeps contradicting the numbers, trust what you see and give the stats time to catch up.
- Still check the count on rare stats before acting on them.
When it’s your own number
Your own badge has a hands count too, and it’s the sample anyone at the table running the HUD has on you. In a regular home game the same players see hundreds of your hands, so your VPIP and PFR are readable within a session and your postflop habits not long after.
Use the same rules on yourself. Don’t change your game because of 50 hands of numbers. After a few hundred hands, compare your VPIP, PFR and c-bet with their typical ranges and check the reads on your badge, which show what others can use against you. Leave win rate alone until the sample is in the thousands.
Using it on the HUD
Hands is on every open badge: the small number at the end of the top row, such as “412 h”. It has no color. Every other number’s own sample is in its tooltip, and the HUD fades any number with fewer than 12 chances behind it (win rate: fewer than 100 hands).
The hands count itself drives the player type, which appears after 30 hands and stays faded until 60. Reads depend on each stat’s own chances: at least 12, with the number at least one standard error outside the typical range; AF needs 20 bets, raises and calls, at least 8 of them calls. When a narrow version of a stat has fewer than 12 chances, the HUD steps back to a broader one and says so in the tooltip.
Low-percentage stats have smaller errors in points but not in meaning: a 10% 3-bet over 30 chances is ±5.5 points, the gap between a 5% and a 15% 3-bettor. Over 100 chances it’s ±3.
Questions
How many hands do you need for poker stats to be reliable?
It depends on the stat, because each one counts its own chances. VPIP and PFR get a chance almost every hand, so 50–100 hands gives a usable picture (±4.6 points for a 30% stat at 100 chances). Spot stats like 3-bet or c-bet need several hundred hands, rare spots like fold to 4-bet thousands, and win rate tens of thousands.
What does the “412 h” on a player’s badge mean?
It’s the number of finished hands the HUD has recorded with that player dealt in: 412 here. It covers every game you’ve shared, since stats follow the player’s PokerNow ID. It isn’t the sample for each stat; hover any number to see how many chances that particular stat has.
Why is a stat faded when the player has hundreds of hands?
Because that stat has fewer than 12 chances, whatever the hands count says. Fold to 3-bet only counts hands where the player opened and got 3-bet, so 400 hands can hold just a handful of those spots. The tooltip shows the exact count.
Do hands carry over between PokerNow games?
Yes. Hands are tied to the player’s PokerNow ID, so every game you’ve played with them adds to the same history, and a name change doesn’t reset it. Only tables you’ve had open count, and when you open one, its earlier hands come in on their own.
How many hands before the HUD shows a player type?
Thirty. Before that there’s no type at all. From 30 to 60 hands the type is shown faded, because VPIP and PFR can still move enough to change it; after 60 it’s shown normally. The type comes mostly from VPIP and PFR, which get a chance almost every hand, so it firms up much sooner than postflop stats.