Downswing is variance — not a curse
Every player has a horror story — fewer have a histogram. Downswings feel personal because brains map losses to causation automatically. Variance is unpersonal math: given your winrate, format, and volume, a −20 buy-in month can be inside normal 95% bounds — annoying, not evidence of a curse.
Variance anchors by format
Confidence intervals beat gut feel
A 3 bb/100 winner over 25,000 hands might show −5 bb/100 over the next 10,000 — still consistent with skill. Plot rolling 10k bb/100; if it wanders inside ±2σ, you are watching noise, not collapse. Strategy changes on noise are how players spew during variance trying to «fix» randomness.
The audit fork: variance vs leak
Export flagged hands
Tag spew, thin call, missed fold during downswing — not every loser.
Compute spew rate
Spew bb lost / total bb lost — if < 15%, variance dominates.
Compare to baseline
Same spew rate as heater month? Process intact — variance.
Act on bankroll, not mood
Drop stake at 70% bankroll rule; do not widening ranges to «get even.»
Psychological tax — real, but separate line item
- Outcome attachment — reviewing results before decisions — fix with decision-first ledger.
- Stake creep — moving up to «escape» downswing — RoR explosion, not escape.
- Volume panic — 3× tables to «recover faster» — variance amplifier.
- Superstition loop — new avatar, new stake, same ranges — noise cosplay.
I remember quitting a winning strategy during a 4σ month
I remember abandoning a BTN steal chart after −18 buy-ins in six weeks — spew review showed 4.1% of losses flagged, identical to prior +12 buy-in quarter. True issue: MTT volume during same window added correlated variance I never separated in the ledger. Reverted chart; next eight weeks +9 buy-ins with same process. I almost «fixed» variance by deleting edge.
Plot your last 10k hands or 200 tournaments, compute spew %, and only then change a line on NUTS POKER. Downswing is not a curse — it is statistics billing you on schedule. NUTS POKER rewards the player who respects confidence intervals, not the player who rewrites charts every bad month.
