Bankroll management for LoL betting, without the hand-waving
The math behind flat staking, Kelly criterion, and why most bettors go broke even when they pick winners. Includes a ruin probability table for common stake sizes.
Most LoL bettors who go broke are not bad at picking winners. They are bad at sizing bets. This article is about sizing bets. It is the least exciting piece on this site, and the one that will save you the most money.
The setup
You have a bankroll. Call it 1,000. You want to bet on LoL matches. Two questions:
- How much of that 1,000 should you risk per bet?
- How do you adjust when you win or lose?
The answers depend on your edge. The tricky part is that most people do not know what their edge is. They estimate high. They size accordingly. They go broke.
Flat staking vs percentage staking
Flat staking: every bet is a fixed dollar amount. 20 per bet, regardless of bankroll size.
Percentage staking: every bet is a fixed percent of your current bankroll. 2 percent of whatever the bankroll is today.
Percentage staking is mathematically better when you have a genuine edge. It grows your stake as you win and shrinks it as you lose, which reduces ruin probability and captures more upside.
Flat staking is simpler and more psychologically robust. It is also fine if you rebase your “stake unit” every month or after any large swing.
The ruin table
Here is what happens at different stake sizes, assuming you pick winners at 54 percent at -110 lines (which is above-average skill, for context):
| Stake size | Ruin probability over 1000 bets |
|---|---|
| 1% of bankroll | < 0.1% |
| 2% of bankroll | 1.0% |
| 3% of bankroll | 5.8% |
| 5% of bankroll | 22.4% |
| 10% of bankroll | 62.1% |
| 15% of bankroll | 82.7% |
That last row is worth staring at. A skilled bettor who stakes 15 percent per play has an 83 percent chance of going broke inside a thousand bets. The math is brutal.
At 2 percent, a skilled bettor goes broke once in every hundred parallel universes. That is the price of staying in the game.
Kelly criterion, explained without calculus
Kelly criterion is the “optimal” stake size given your edge. The formula for even-money bets:
Kelly stake = edge / odds.
If you have a 5 percent edge at 2.00 decimal (even money), Kelly says bet 5 percent of bankroll. If the edge is 10 percent, bet 10 percent.
Problem one: you do not know your edge. You estimate it. If you estimate 10 percent and the real edge is 3 percent, full Kelly on the 3 percent is fine but full Kelly on your 10 percent estimate over-bets by more than 3x. That gets you ruined fast.
Problem two: even with a correct edge estimate, full Kelly produces enormous variance. A Kelly-optimal bettor faces 50 percent drawdowns regularly. Most people cannot psychologically survive that.
The pragmatic answer: use quarter Kelly or eighth Kelly. Or just stake 1 to 2 percent flat and stop worrying about it.
The standard deviation problem
Even at 55 percent hit rate at -110, you will have losing streaks of 7 or 8 in a row. Multiple times per thousand bets. This is not bad luck. It is how independent 55/45 bets behave.
If a streak of 8 losses at 2 percent staking hurts you emotionally, imagine 8 losses at 5 percent staking. Now imagine at 10. This is why staking discipline matters more than pick quality for 99 percent of bettors.
What to do with bonuses
Deposit bonuses are real money if you meet the rollover. The math is simple.
A 100 bonus with 10x rollover requires 1,000 in wagering at an average hold of 4.5 percent. Expected loss on that rollover: 45. Net value of the bonus: 55 profit, assuming you do not tilt into bad bets to meet the rollover.
A 100 bonus with 20x rollover requires 2,000 in wagering. Expected loss: 90. Net value: 10 profit, assuming perfect discipline.
Rollover requirements above 8 to 10x are generally not worth chasing unless the edge of your regular bets is genuinely positive.
Unit size and split budgets
Practical approach I recommend to people who want structure without obsession:
- Determine your bankroll. Call 1 unit = 1 percent of bankroll.
- Default stake per bet: 1 unit.
- Maximum stake per bet: 2 units (only on your highest-confidence plays).
- Rebase the “1 unit” definition at the start of each split based on current bankroll.
That is it. Two-sentence staking plan. The people who do this outperform most bettors who obsess over Kelly fractions.
When to increase stakes
Do not. Not until you have at least 500 tracked bets with positive CLV across multiple event types. Once you have that data, increase cautiously, from 1 unit default to 1.25, then 1.5, measured over another 500 bets.
Most bettors increase stakes after hot streaks. Hot streaks regress. Increasing stakes after a hot streak produces the steepest drawdowns.
When to decrease stakes
You should decrease stakes if:
- Your recent 100 bets have negative CLV. Not negative results. Negative closing line value.
- Your written reasoning, compared a month later, reads as tilted or sloppy.
- You catch yourself betting for reasons outside your model (“the match is on now and I want action”).
Decrease means halve your unit. Not skip. Betting with less skin lets you keep evaluating your process without compounding losses.
The emotional layer
I have one friend who is a competent quantitative trader professionally. He cannot bet LoL profitably, because he cannot bet 1 percent on matches where his “feeling” is strong. He pushes to 3 percent, gets unlucky, doubles to 6 to “make it back,” gets unlucky again, and gives up for 6 months. Repeat annually.
The math is not the hard part. The discipline is the hard part. If you know you are the type of person who will push stakes when emotional, pre-commit to a maximum stake per bet and do not bet without it. Many books let you set a per-wager limit in account settings. Use it.
Event-level budgets
For Worlds and MSI, I set an event budget. Usually 5 to 10 units total across the whole event. If I hit the budget, I am done for the event regardless of what is left on the schedule.
This does two things. It caps my exposure to any one event, and it forces me to pick which matches deserve plays. I bet less at majors now than I did five years ago. My ROI has gone up.
The one-paragraph version, in case you skip the rest
Bankroll of X. Bet 1 to 2 percent per play. Track every bet in a spreadsheet including closing line value. Rebalance the percentage each split based on current bankroll. Never increase stakes after wins, never chase losses, never bet a match just because it is on now. Pre-commit to per-wager and monthly limits. If you cannot follow these rules, bet smaller or stop.
None of this is clever. All of it is what people who are still in the game after five years actually do.