LOADING
Vinny · The book
“One second. Counting the house money.”
“One second. Counting the house money.”
Variance is real. A hot streak proves nothing. A cold streak proves nothing. You need hundreds of bets before the signal separates from the noise.
Sample size is the number of bets behind a result, and small samples are dominated by variance. A genuine 55% bettor can easily go 5-10 over 15 bets, and a losing bettor can go 10-5 — so you generally need 200–500 bets at consistent stakes before a win rate means anything.
I love touts who go on a hot streak and start selling picks. "10-2 this month, join now!" Ten wins? Over two weeks? I'd flip a coin against their picks and be slightly nervous. But the public doesn't know what sample size means, and the tout is charging $49.99 a month.
Small samples are a gift to the fraudulent. They can always find a recent stretch that looks meaningful. "Last 30 days." "Last 12 picks." Whatever window flatters them. Twenty good bets isn't an edge. It's an afternoon.
I convinced myself I had a system based on about 15 picks. FIFTEEN. I was 11-4. I doubled my unit size.
The next 20 picks went 8-12. I called it a bad stretch. It wasn't a bad stretch. It was regression to the mean. I had no system. I had a lucky start followed by normal results. I couldn't tell the difference.
Variance is not your enemy — it's just the cost of operating in probability. Even a strong 55% bettor will run 5-10 over 15 bets without anything being wrong. And a 45% bettor can run 10-5 over 15 bets and look like a genius.
Rule of thumb: you need 200–500 bets at consistent stakes before you can say anything statistically meaningful about your win rate. Below that, almost any result — hot or cold — is probably just variance.
This is why EE's public ledger goes back to the beginning, every bet included. Not because every bet was perfect, but because the only honest record is the full record. Anyone selling picks off a 30-bet hot streak is selling variance, not edge.
Roughly 200–500 bets at consistent stakes. Below that, hot and cold stretches are usually variance rather than proof of skill or a broken system.
Because normal variance produces streaks. A true 55% bettor will sometimes go 11-4, and so will a true 45% bettor. A short record can't separate luck from edge.
It is the tendency for extreme short-term results to move back toward the true long-run rate. A lucky start is usually followed by ordinary results, not a continued streak.