dfs · tools
DFS Variance Calculator: Monte Carlo Bankroll Simulation
Last Updated: March 1, 2026
DFS variance determines whether a profitable strategy survives long enough to realize its edge. This Monte Carlo simulator runs thousands of bankroll trajectories based on your win rate, ROI, contest mix, and entry sizing to project the range of outcomes you should expect — including your probability of going broke before your edge compounds.
Last Updated: March 2026
Key Takeaways
- Variance in DFS is driven primarily by contest type — GPPs produce 3-5x the standard deviation of cash games per dollar entered.
- Risk of ruin drops exponentially as entry size decreases relative to bankroll. Keeping per-slate exposure below 5% keeps ruin probability under 3% for most profitable players.
- A Monte Carlo simulation reveals the full distribution of outcomes, not just the average — the bottom 5th percentile trajectory matters more than the median for survival.
- Even at 8% ROI, a GPP-heavy player can experience 50+ slate drawdowns. The simulator quantifies how deep and how long those drawdowns last.
- Pair this tool with our DFS bankroll management guide and the DFS contest EV calculator for a complete risk framework.
How Does a Monte Carlo Simulation Work for DFS?
A Monte Carlo simulation generates thousands of independent bankroll paths by randomly sampling outcomes from your estimated win rate distribution. Each simulated slate draws a result — win or loss, with payout sized to your contest type — and updates the running bankroll. After 10,000 simulations across your specified number of slates, the tool aggregates the results into percentile bands, median trajectories, and a risk-of-ruin estimate.
The critical output is not the average outcome. The average is simply your ROI multiplied by entries multiplied by slates — basic arithmetic. The value of simulation lies in the tails: what happens in the worst 5% of scenarios, how deep drawdowns run, and what percentage of paths hit zero before recovering.
| Metric | What It Tells You |
|---|---|
| Median final bankroll | The “most likely” outcome, accounting for compounding |
| 5th percentile path | Your worst realistic trajectory — plan for this |
| 95th percentile path | Best realistic case — do not plan around this |
| Risk of ruin (%) | Probability your bankroll hits zero before the simulation ends |
| Max drawdown (median) | Typical deepest trough you will experience |
Why Does Contest Mix Drive Variance More Than Win Rate?
GPP (guaranteed prize pool) tournaments concentrate payouts at the top. A typical large-field GPP pays 20% of entrants, with the top 1% receiving 40-60% of the prize pool. This payout structure means most entries lose, and wins are large but infrequent. Cash games pay roughly 50% of the field at 1.8x, producing a much tighter distribution.
Our analysis of contest structures across major DFS platforms shows that a player entering 100% GPPs at 8% ROI experiences a standard deviation per slate roughly 4.2x higher than a player entering 100% cash games at the same ROI. The practical consequence: the GPP player needs a substantially larger bankroll to survive the same number of slates with the same ruin probability.
| Contest Mix | Std Dev Per Slate (Relative) | Risk of Ruin at 5% Entry Size (500 Slates) | Risk of Ruin at 10% Entry Size (500 Slates) |
|---|---|---|---|
| 100% cash | 1.0x | ~1% | ~6% |
| 70% cash / 30% GPP | 1.8x | ~3% | ~14% |
| 50% cash / 50% GPP | 2.5x | ~5% | ~22% |
| 100% GPP | 4.2x | ~12% | ~41% |
Estimates assume 6% ROI. Actual ruin rates depend on payout structure, field size, and win-rate calibration.
How Should You Use These Results?
The simulator answers one question: given your estimated edge and your contest mix, how much can you enter per slate without an unacceptable probability of ruin? The answer is a function of three variables you control — entry size, contest allocation, and total bankroll — and one you estimate: your ROI.
Start by entering conservative ROI assumptions. If you have tracked fewer than 300 slates, discount your observed ROI by 30-50% for the simulation input. Overestimating edge is the primary cause of false confidence in low ruin probabilities.
Compare your simulated ruin probability against your risk tolerance. Most professionals target a ruin probability below 5% over a full season (roughly 200-300 slates for NFL, 800+ for NBA). If your ruin probability exceeds that threshold, reduce entry size or shift allocation toward cash games.
Track your actual results over time using the Odds Reference dashboard and re-run the simulation quarterly with updated inputs. Your edge estimate improves as sample size grows, and the simulator becomes more reliable.
For a deeper treatment of how to size entries relative to bankroll, see our DFS strategy guide.
FAQ
Q: What is variance in DFS?
A: Variance measures how far your actual DFS results deviate from your expected value over a set of slates. A player with 5% ROI will not profit exactly 5% every night — some nights produce large wins, others produce losses. High variance means wider swings. GPPs have far higher variance than cash games because payouts are top-heavy, so your results can deviate dramatically from expectation even over hundreds of entries.
Q: How many slates does it take to know if I’m profitable?
A: Most DFS players need 500 to 1,000 tracked slates before their results reliably separate skill from variance. At 5% ROI with a typical GPP mix, the 95% confidence interval on your true ROI does not tighten below plus or minus 3% until roughly 800 slates. Cash-game-only players converge faster — around 300 to 400 slates — because lower variance narrows the distribution sooner.
Q: What’s a realistic ROI for DFS?
A: Profitable DFS players typically sustain 3-8% ROI over a full season. Elite grinders with ownership modeling and optimizer stacks reach 10-15% in select sports and contest types. The median DFS player loses money after platform rake, which runs 10-15% of the prize pool. If your estimated ROI is above 15%, your sample size is likely too small to be conclusive.