讀懂撲克分析報告：HUD 數據、樣本量與 leak triage
2026年6月2日
VPIP-PFR gap、WTSD triangle、positional BB/100、EV loss triage。data 可靠所需 minimum hand。五手 verification drill——避免 single-stat panic adjustment。
- poker stats
- HUD
- VPIP
- PFR
- WTSD
- BB/100
- leak analysis
- poker report
- hand analysis
VPIP 34% 不一定是 leak——可能是 40-hand loose live sample。Tracker dashboard、session summary、AI leak flag 需要 sample calibration 與 cross-stat read。
RiverOdds、nlh.poker、TPP Academy、Seeker Start 共識：cluster read——VPIP with PFR、WTSD with W$SD、c-bet with fold-to-c-bet——加 minimum hand threshold。
Primary Stats
| Stat | 含義 | 常見 misread |
|---|---|---|
| VPIP | 入池 width | 無 position context 判「高=壞」 |
| PFR | raise vs limp/call freq | 忽略 limp pool |
| VPIP − PFR | passivity index | gap >15 = call/limp heavy |
| WTSD / W$SD | showdown freq / win rate | 高 WTSD + 低 W$SD = station |
| BB/100 | profit rate | 單 format 20k+ hand 才 stable |
Online 6-max reg 大致 22–28% VPIP / 18–22% PFR。同 format 比你自己。
Sample Size Threshold
| Stat | 最低 hand | 原因 |
|---|---|---|
| VPIP / PFR | ~300 | 每 orbit |
| 3-Bet | ~1,000 | sparse |
| C-Bet | ~1,500 | 需 split position |
| WTSD / W$SD | ~3,000+ | showdown rare |
Positional split mandatory。
WTSD Triangle
| WTSD | W$SD | Profile | Fix |
|---|---|---|---|
| High | Low | call too wide | value bet thinner |
| Low | High | fold too much | bluff more |
Five-Hand Verification Drill
stat → filter 10 → review 5 → one sentence rule → recheck after 2,000 hands。
總結
Analysis report 是 diagnostic data，非 autopilot。Calibrate sample size before repair。Sort EV loss by repeatability，verify one hand with pot math。