Bankroll Management: Staying Solvent
Published Sat Sep 06 2025
Bankroll management ensures you survive variance long enough for your edge to materialize.
It transforms mathematical advantage into realized capital growth while limiting probability of ruin. Poor bankroll discipline converts good edges into bankruptcy.
Core Principles
- Define a bankroll separate from living expenses.
- Use consistent stake sizing logic (units, % of bankroll, fractional Kelly).
- Reassess sizing after meaningful bankroll changes, not every win/loss.
- Separate tactical bankroll (active risk) from reserve (not in play) if stakes or limits escalate.
Units System
1 unit = base stake (e.g., 1% of bankroll). Larger confidence edges maybe 1.5–2 units (cautious). Avoid 5–10 unit emotional spikes.
Determining Base Unit
| Bankroll | Conservative (0.5%) | Moderate (1%) | Aggressive (2%) |
|---|---|---|---|
| 1,000 | 5 | 10 | 20 |
| 5,000 | 25 | 50 | 100 |
| 10,000 | 50 | 100 | 200 |
Aggressive only if edge size & liquidity proven.
Fractional Kelly Overlay
Blend qualitative confidence with fractional Kelly suggestions. Use the lower of (unit cap, fractional Kelly stake).
Confidence Tiers (Example)
| Tier | Criteria | Stake |
|---|---|---|
| A | Strong model agreement + positive CLV historically | 1.5–2u |
| B | Standard edge, baseline confidence | 1u |
| C | Small edge / higher uncertainty | 0.5–0.75u |
| PASS | Edge below threshold or correlated overexposed | 0u |
Stop-Loss / Review Points
Instead of emotional stop-losses, set analytical review triggers (e.g., 30% drawdown OR every 500 wagers) to evaluate edge validity.
Segregate Bankrolls
Different strategies (e.g., live betting vs pre-game models) can have separate tracking to avoid cross-subsidizing unclear performance.
Takeaway
Consistency beats spasmodic staking. Protect the bankroll; it is the engine that converts EV into realized profit.
Rolling Review Framework
Every 500 bets or monthly (whichever later):
- Update equity curve & maximum drawdown.
- Segment performance (sport, odds bucket, time of day).
- Compare actual yield vs model EV & CLV trend.
- Adjust unit size only if: bankroll changed by >25%, or volatility materially different.
Risk of Ruin (Conceptual)
Exact calculation requires distribution assumptions; heuristically, ruin probability increases sharply if typical stake > edge / variance comfort band. Staying below ~1–2% flat stakes (unless using fractional Kelly) keeps ruin probability low for modest edges.
Common Pitfalls
- Chasing losses with oversized “get-even” bets.
- Unlogged bankroll adjustments (hard to audit performance).
- Mixing personal finances with betting float.
- Ignoring correlation: stacking multiple bets on same underlying event inflates risk.
FAQs
When to increase unit size? After sustained bankroll growth (e.g., +50%) and confirmation edge persists (positive CLV, calibrated probabilities).
Shrink during drawdowns? Only if analysis shows edge deterioration—not just variance.
Multiple strategies different stakes? Yes: assign each a % allocation and track separately. Rebalance periodically.
Use credit instead of cash bankroll? Increases psychological stress & ruin risk; avoid.