Logo

Most players lose not because they pick bad games, but because they pick bad bet sizes. A 55% win rate means nothing if you're staking 20% of your roll on every spin or parlay. Bankroll management is the unsexy skill that separates the 2% who last from the 98% who reload and repeat. Here's how to actually do it.

The Math You Can't Argue With

Variance isn't a theory, it's a constant. If you flip a coin 100 times, you'll rarely get exactly 50 heads. Same with slots and poker. Your job isn't to beat variance, it's to survive it long enough for your edge to show up. That means stake sizing matters more than game selection. We found B52 CLUB to be a handy reference here.

Take a sports bettor with a 53% win rate on -110 lines. On a $1,000 bankroll, flat betting $20 (2%) gives you roughly 50 losing bets before ruin. That's not 50 bad bets in a row, that's the mathematical cushion. Bump that to $100 per bet and you're gone after 10 bad picks. Same handicapper, same edge, different outcome. The size killed him.

For slots, this is even harsher. A typical 96% RTP game has a house edge of 4%, but variance means your personal RTP over 500 spins can swing 30% in either direction. You need enough ammo to ride that out, which is why slot players should keep individual spins at 0.5% to 1% of their bankroll. On $500, that's $2.50-$5 a spin. Not $20.

Set Your Units Before You Play

A "unit" is your standard bet. Pick one number and stick to it across the board. Pros don't think in dollars, they think in units, because it removes emotion and makes losses abstract enough to keep playing correctly.

1. Calculate your true roll. Money you can lose without affecting rent, groceries, or relationships. If it hurts, it's too much.
2. Assign 1-2% per unit. $2,000 roll = $20-$40 units. That's it.
3. Set a stop-loss. Typically 3-5 units lost in a session, walk away. No exceptions. This isn't about luck, it's about not chasing.
4. Set a stop-win. Harder than it sounds. Up 10 units? Take half off the table and lock it.
5. Rebuy rules. Can't reload mid-session. Period. If you broke rule 3, you don't get to break rule 5.

The stop-loss number matters less than having one. A dealer once told me the average player doubles their average loss by chasing. The math supports that, if you ever track it.

Poker Bankrolls Are a Different Beast

Cash game poker needs 20-30 buy-ins for solid play, tournament grinders need 100+ buy-ins at their level. And that's for winning players. If you're break-even and playing $1/$2 with a $600 roll, you're one bad session from broken, no matter how many flops you hit.

Move down, don't go broke. Dropping from $2/$5 to $1/$2 isn't a downgrade, it's a pit stop. Split your bankroll into windows; if you lose 5 buy-ins, move down a stake. If you win 10, move up. Automating this removes ego from the equation, and ego is the most expensive leak in poker.

Slots and casino table games are different in that there's no skill edge to protect. Your bankroll isn't capital, it's entertainment budget. Track it like a hobby: set a monthly number, treat every spin as prepaid fun, and never top up because you're "due." There is no due. The machines pay nothing.

Bonus Money: Not Everyone Counts It

Welcome bonuses, free bets, and deposit matches pad variance but come with strings. A 100% match with a 35x wagering requirement on a 96% RTP slot means the expected loss on the bonus itself often outweighs the bonus value. Run the numbers before you accept it.

• Bonus value / (wagering x house edge) = expected profit or loss.
• $100 bonus, 35x wagering = $3,500 turn, house edge 4% = $140 expected loss. That's negative EV.
• Same bonus, 10x wagering on blackjack (0.5% edge) = $50 expected loss. Still bad, but closer.

Cash bonuses are generally better than free spins, and low-wagering is better than high. Read the terms before claiming anything that sounds generous. Generous usually means sticky terms dressed up as value.

Tracking Is Non-Negotiable

You cannot manage what you don't measure. A spreadsheet with date, game, stake, and result takes 30 seconds per session and shows exactly where you leak. Most players are shocked to find a game they love performing 15% worse than the RTP because they play it tired, drunk, or for hours after tilt.

Track by game type, not by individual session. Slots, table games, sports, and poker should each have their own line. If you're down 20 units on one and up 5 on another, you know where to cut back. Intuition lies. Records don't.

The Takeaway

You won't out-pick the house. But you can out-last it, sometimes, if your stake sizing respects variance more than your confidence does. The players who stick around for decades aren't the ones with the best systems, they're the ones who never had a bad night wipe out six good ones. Units, stop-losses, and a real bankroll ceiling are the boring tools that make that possible.

Set your unit, honor your stop, write down what you did. That's the whole edge. Everything else is noise.