To estimate how fast you can go broke, define what "broke" means for your game, quantify your edge and variance, then model many realistic sequences of bets to see how often (and how quickly) your bankroll hits that ruin line. This turns bankroll management from vibes into measurable risk you can act on.
Core Principles of Bankroll Survival
- Define a concrete ruin point (not just "$0"), then measure everything against it.
- Risk is driven by three inputs: edge, variance, and how long you keep playing.
- Use conservative assumptions; your model should survive "bad-but-plausible" runs.
- Bet sizing is the main lever you control; small sizing changes can drastically shift ruin risk.
- Separate "math says OK" from "cashflow says OK" (fees, minimums, life expenses).
- Monitor drift: odds, win rate, and stake discipline change over time-update the model.
Quantifying Ruin: Defining 'Going Broke' for Your Strategy
This approach fits intermediate players who already track results and want a practical estimate of blow-up speed under a specific staking plan. It's useful for poker, sports betting, and other repeated-bet games where you can approximate average edge and volatility.
Don't do this if (a) your bets are highly correlated (e.g., many parlays tied to one outcome) and you can't model that correlation, (b) your "edge" is unknown or unstable (new market, tiny sample), or (c) you're using the bankroll for essential living costs-then you need a cashflow-first plan, not just a ruin model.
Quantitative example: If you start with 10,000 THB but your platform requires a 1,000 THB minimum stake, your practical ruin point might be 1,000 THB (you can't place your intended bets below that). In that case, "going broke" happens at 1,000 THB, not at 0.
Variance, Edge and Timeframe: The Three Drivers of Risk
Before you use any bankroll risk calculator or build a simple risk of ruin calculator yourself, gather inputs you can defend. For intermediate users, "good enough" inputs beat perfect theory.
What you need
- Bankroll definition: starting bankroll and your ruin threshold (e.g., stop at 20% of start, or when you can't place minimum stakes).
- Bet schedule: fixed stake, proportional stake, or tiered stakes (and any caps).
- Edge estimate: expected value per bet (or per 100 hands for poker). Use a conservative estimate.
- Variance estimate: typical swing size per bet/hand; if you can't compute it, approximate from historical results.
- Timeframe: number of bets (sports) or hands/sessions (poker) you expect to play under this plan.
- Tools: spreadsheet (Google Sheets/Excel) or a simple script; optional: an online bankroll risk calculator for quick cross-checking.
Quantitative example: Sports betting: if you place 200 bets per month and want "survive 3 months," your timeframe is 600 bets. Poker: if you play 20,000 hands per month, "survive 2 months" is 40,000 hands.
Modeling Survival: Simple Simulations and Analytical Approaches
The safest, most user-friendly method is Monte Carlo simulation: you generate many plausible sequences of results, apply your staking rule, and record whether/when you hit the ruin threshold. This is the core of most credible bankroll management tools, whether branded as a bankroll risk calculator or a risk of ruin calculator.
-
Lock your ruin definition and stake rule
Write down (1) starting bankroll, (2) ruin threshold, and (3) how stakes change after wins/losses. If you change rules mid-model, the output becomes storytelling.
- Example rule: "Stake = 2% of current bankroll, capped at 500 THB."
- Example ruin: "Stop when bankroll ≤ 2,000 THB."
-
Estimate edge and variance conservatively
Use a pessimistic edge and a realistic (or slightly worse) variance. If you only have a small sample, shrink your edge toward zero.
- Sports example: assume your long-run ROI is +1% even if recent results look like +4%.
- Poker example: use a lower win rate than your best month.
-
Choose a result model that matches your game
For sports betting, model each bet as win/loss with payout odds; for poker, model per-session (or per-1,000-hands) profit as a random draw around your win rate with observed swing size.
- Sports: profit per bet = stake × (odds−1) on win, or −stake on loss.
- Poker: profit block = mean ± noise; use your historical distribution if you have it.
-
Run many simulated timelines and record "time to ruin"
Simulate the full timeframe repeatedly. For each run, stop early if the bankroll crosses the ruin threshold, and record the bet/hand count when it happened.
- Track: ruin yes/no, time-to-ruin, worst drawdown, ending bankroll if not ruined.
-
Stress-test the assumptions (edge down, variance up)
Re-run the same simulation with worse inputs. Your plan should not collapse when edge is slightly lower or variance slightly higher than expected.
- Stress-test idea: cut edge in half and increase swing size modestly, then compare ruin frequency and speed.
-
Translate the output into a decision rule
Decide in advance what level of ruin risk is acceptable for you, then adjust stakes (or timeframe) until you meet it. This is practical bankroll management, not academic modeling.
- Example rule: "If stress-test runs show frequent early ruin, reduce stake size and re-test before playing."
Quick mode: a fast-track algorithm
- Set start bankroll, ruin threshold, and a simple stake rule (fixed % with a cap).
- Use conservative edge and realistic swings from your log (or default to lower edge if unsure).
- Simulate many runs for your target timeframe; record how often you hit ruin and how fast.
- Stress-test: edge down, variance up; if ruin becomes common, cut stake size and repeat.
Sizing Bets: Kelly, Fractional Kelly and Practical Alternatives
Kelly sizing is a growth-optimal formula under strict assumptions; in real play, estimation error and changing conditions make full Kelly fragile. For most intermediates, fractional Kelly (or simpler caps) is safer and easier to keep consistent-especially if you're mixing markets like sports betting bankroll management with occasional higher-variance plays.
Quantitative example: If your model suggests a "full Kelly" stake of 4% of bankroll, then half-Kelly is 2%, and quarter-Kelly is 1%. If you notice your edge estimate is noisy, quarter-Kelly is often more survivable than full Kelly under the same variance.
Sanity-check checklist before you trust the bet size
- Your stake rule is written and repeatable (e.g., "1% of bankroll, cap 500 THB").
- The ruin threshold matches platform minimums, fees, and your withdrawal constraints.
- Edge inputs are conservative and based on enough history to be credible.
- Variance inputs reflect real swings (including losing streaks), not just average results.
- Your model includes the actual payout structure (odds, rake, commissions, promotions excluded if uncertain).
- You ran a stress-test where edge is lower than expected and swings are worse than expected.
- Your staking plan still works if bankroll drops (no forced "must-bet" minimum that breaks the rule).
- You can execute it without emotional overrides (no doubling to "get even").
Managing Downswings: Stop-losses, Rebalancing and Liquidity Plans
Downswings don't kill bankrolls by themselves; uncontrolled responses do. Build rules that prevent bet-size creep and keep you liquid when variance spikes.
Quantitative example: If your stop rule is "pause when down 10% from peak," then with a 10,000 THB peak you pause at 9,000 THB. If you instead stop at "down 10% from start," you might keep playing deeper into a downswing after a temporary peak-be explicit.
Common mistakes that shorten bankroll life
- Using full Kelly off a noisy edge estimate and calling it "optimal," then blowing up during normal variance.
- Increasing stakes after losses (chasing), which is the fastest way to invalidate any poker bankroll management strategy or sports plan.
- Ignoring correlation (stacking bets on the same team/market outcome), making your "number of bets" less diversified than it looks.
- Changing the ruin threshold midstream ("I'll stop at 2,000 THB... ok maybe 1,000..."), which guarantees worse outcomes.
- Confusing short-term luck with edge and raising stakes before the advantage is proven.
- Not reserving liquidity for fees, minimum stakes, and withdrawals-practical ruin happens earlier than mathematical ruin.
- Not re-estimating inputs after market conditions change (line quality, limits, rake, rule changes).
- Using a black-box "bankroll risk calculator" output without verifying the assumptions match your staking rule and bet type.
Translating Models to Action: Rules, Checklists and Monitoring
Your model is only useful if it becomes operating rules. Pick an approach that matches your discipline and data quality, then monitor and adjust on a schedule.
Alternative operating modes (when each is appropriate)
- Fixed-percentage staking with caps
Use when your edge is uncertain or changing; it's robust and easy to execute. Good default for sports betting bankroll management with mixed odds and bet types. - Fractional Kelly (quarter- to half-Kelly)
Use when you have stable sizing inputs and good records; it's closer to "optimized" while controlling estimation error. - Unit system with periodic rebase
Use when simplicity matters: define 1 unit as a small % of bankroll and only update unit size weekly/monthly to reduce emotional resizing. - Hard drawdown stop + reassessment
Use when tilt risk is high or when bankroll is also life-critical: stop at a predefined drawdown, then re-run your risk of ruin calculator assumptions before resuming.
Common Practical Concerns About Bankroll Longevity
How many simulations do I need for a useful answer?
Enough that results stop changing materially when you re-run them. If outputs swing a lot between runs, increase the number of simulated timelines or simplify your model assumptions.
Can I use a bankroll risk calculator instead of building my own?
Yes, as long as you can input your real stake rule, odds/payout, and ruin threshold. Treat any tool as wrong until you confirm its assumptions match your game.
What's the most common bankroll management mistake for intermediates?
Overestimating edge and then sizing too aggressively (often full Kelly or "confidence-based" sizing). The second is changing stakes impulsively during a downswing.
How does a poker bankroll management strategy differ from sports betting sizing?
Poker usually models variance over hands or sessions with rake and table conditions; sports betting often models discrete win/loss payouts with odds. Both still reduce to edge, variance, and volume over time.
Is a risk of ruin calculator meaningful if my results aren't independent?
It can be misleading if correlation is strong (same event exposure, same market driver). In that case, you must model correlation explicitly or reduce simultaneous exposure as a safety rule.
Should I set a stop-loss per day or per session?
Use a stop-loss when it prevents rule-breaking (tilt, chasing), not as a superstition. Tie it to a drawdown threshold and require a short reassessment before resuming.
How often should I update my inputs?
Update on a fixed schedule (e.g., monthly) or after a major change (limits, odds source, rake/fees, strategy shift). Frequent ad-hoc updates tend to smuggle emotion into the model.
