← Home Guide

Bankroll management

You can pick winners at a 55% clip and still go broke with bad staking. Bankroll management is the least exciting — and most important — skill in betting.

Bet in units, not dollars

A unit is your standard bet size — typically 1% of your bankroll. With a $1,000 bankroll, 1 unit = $10. Thinking in units does two things: it scales your risk to what you can afford, and it lets you compare results honestly over time ("I'm up 12 units this season" means something; "I'm up $340" doesn't).

Your bankroll is money set aside only for betting — separate from rent, bills and savings. If losing it would hurt your life, it's not a bankroll, it's a problem. See responsible gambling.

Flat betting — the professional default

Bet the same amount (usually 1 unit, occasionally 2 on your strongest plays) on every wager, regardless of confidence. It sounds boring because it is — and it works because it removes emotion from sizing.

The math that matters. At −110 you need to win 52.38% just to break even. A bettor winning 53% betting flat $100 units nets about +$8 per 100 bets after juice... wait, let's be precise: 53 wins × $90.91 − 47 losses × $100 = $4,818 − $4,700 = +$118 per 100 bets. Small edges compound; big swings kill. Flat betting keeps the swings survivable.

Avoid progressive systems (Martingale and its cousins): doubling after losses feels logical until a 6-game losing streak asks you to risk 64 units to win 1. The math doesn't care about your system.

The Kelly criterion — sizing to your edge

Kelly sizes your bet proportionally to your edge: bigger edge, bigger bet; no edge, no bet. The formula:

f* = (bp − q) / b  ·  b = decimal odds − 1, p = your win probability, q = 1 − p
Worked example. You estimate a team wins 60% (p = 0.60) at +100 (b = 1.0). Kelly says f* = (1.0 × 0.60 − 0.40) / 1.0 = 0.20 → bet 20% of bankroll. That's full Kelly — aggressive. Most practitioners use half Kelly (10%) or quarter Kelly (5%) because your probability estimates are never as precise as you think.

Kelly's dirty secret: it's only as good as your probability estimate. Overestimate your edge and Kelly happily tells you to overbet. Use the Kelly calculator with fractional Kelly, and be conservative with p.

Line shopping — free money, legally

The same bet is priced differently at different books. Chiefs −6.5 at one book and −7 at another is not a rounding error — in the NFL, that half point is worth real win probability, and −110 vs. −105 on the same line is pure savings.

Worked example. You bet $110 to win $100 fifty times a season. At −110 everywhere, going 26–24 nets −$60. Getting −105 on those same 50 bets: 26 wins × $104.76 − 24 × $105 = $2,724 − $2,520 = +$204. Same picks, $264 swing — from price alone.

Keep accounts at multiple legal books and check our odds board — it highlights the best price per outcome automatically. This is the single easiest edge available to a recreational bettor.

Track every bet

Memory lies; spreadsheets don't. Log every wager: date, sport, bet type, line, odds, stake, result, and — critically — the closing line. If you're consistently beating the closing line, your process is sound even during a losing week. If you're not, your process needs work even during a winning one. (More on this in closing line value.)

Rules that keep you solvent


Next: Advanced strategy → — expected value, closing line value, and how the pros think about price.