Poker tracking apps love showing you 50 different stats. Most of them are noise. Six metrics carry the signal, and one of them is meta — sample-size confidence — which tells you whether to trust the other five.
1. Hourly rate ($/hr)
Hourly rate is total net profit divided by total hours played. It is the most intuitive metric and the only one that combines win rate, hand-rate, and stake into a single number. Live $2/$5 winners typically run $30-$60/hr; $5/$10 winners run $60-$120/hr; online cash winners run lower (smaller hand-rate advantages) but multi-tabling closes the gap.
Track hourly rate pause-adjusted — exclude dinner-break time from "hours played" — or your number will be artificially low.
2. BB/100 (cash games)
BB/100 normalizes win rate across stakes. A 5 BB/100 winner at $1/$2 has the same per-hand skill edge as a 5 BB/100 winner at $5/$10. Track BB/100 by stake to see where your edge actually lives.
Sample-size matters enormously: under 20,000 hands the realized BB/100 is mostly variance. Most players act on BB/100 numbers far too early. The free BB/100 calculator includes a 95% confidence interval so you can see what the realized number actually means.
3. ROI (tournaments)
For tournaments, ROI replaces BB/100. ROI is total profit divided by total buy-ins, expressed as a percentage. Strong winning MTT players run 10-30% ROI over large samples. Sample size required for reliability: 1,000+ events at similar buy-in levels.
4. ITM% (tournaments)
ITM% is the share of tournaments you cash in. Most tournaments pay the top 10-15%, so the baseline is roughly 12-15%. Winning MTT players push 15-22% but make most of their profit from deep finishes — ITM alone does not tell you if you are winning.
5. Win rate by stake
Aggregate win rate hides the truth. Win rate by stake exposes it. Most marginal players win at one stake and lose at another. Knowing your BB/100 (or ROI) per stake makes the move-up and move-down decisions almost automatic.
6. Sample-size confidence
The meta-metric. A 5 BB/100 winner over 10,000 hands has roughly a 30% chance of being a small loser (variance). Over 100,000 hands, that drops to 5%. The realized number is the same, but the confidence is dramatically different. More on sample size →
What you should mostly ignore
- VPIP, PFR, 3-bet% — only useful with a HUD on live data, not from a session log.
- Biggest pot won — fun, not informative.
- Profit by month — too short a sample to mean anything.
- Total hands played — without context, this is vanity.
How to use these six
- Track hourly rate first — the intuitive headline.
- Add BB/100 (or ROI for tournaments) for stake-normalized analysis.
- Break BB/100 / ROI by stake — usually the most useful slice.
- Check sample-size confidence before drawing conclusions.
- Use venue and game-type breakdowns to find your most profitable conditions.
- Re-evaluate quarterly.
StackEdge tracks all six automatically. The free calculators (bankroll, hourly rate, win rate, variance) handle the math if you do not have the app yet.
