If you've read our math behind DFS losses piece and our variance deep-dive, you've seen us forward-reference bankroll management twice. This is that article. Bankroll discipline is the unsexy part of DFS — there's no signal to find, no edge to identify, no slate to analyze — and it's the part that separates the players who survive long enough to see their edge materialize from the ones who blow up before they get there.
The accessible content here is short: have a separate bankroll, size entries as a small percentage of it, don't tilt-resize after wins or losses, keep records. That's most of what you need. The optional Kelly section at the end walks through the math behind those rules — skip it if math isn't your thing. The actionable content is in the first five sections.
What a bankroll actually is
Your bankroll is the total amount of money you've set aside for DFS — separate from rent money, separate from your emergency fund, separate from whatever budget category you'd normally pull entertainment spending from. It's a defined pool, and it has a single job: absorb the variance of DFS contests long enough for your edge to show up in the results.
The size of your bankroll should be an amount you'd happily lose for the entertainment value. Not "could afford to lose if I had to." Happily lose — meaning, if you ran the bankroll to zero, you'd feel like you got reasonable entertainment for the money, even if you'd rather have won. This is the same psychological frame that works for a $200 night out: you spent it on the experience, and getting it back was never the point.
If you can't size your bankroll at a level you'd happily lose, the right answer isn't to play smaller stakes — it's to not play DFS at all. The math of operator-rake markets is unkind to players who can't tolerate losing.
The two main sizing rules
Once you have a bankroll, the question is how much to put on each entry. Two conventional ranges cover most of the practical answer:
Beginner rule: 1-2% of bankroll per entry. A $100 bankroll plays $1 or $2 slips. A $500 bankroll plays $5 or $10. This is conservative sizing that survives long variance — you can absorb a 20-slip cold streak without seeing your bankroll go anywhere alarming. Growth is slow on real edges, but survival probability is high.
Aggressive rule: 3-5% of bankroll per entry. Same $100 bankroll plays $3 or $5 slips. This is what you'll see recommended by players who treat DFS as a serious income source. Growth on real edges is faster than at 1-2%, but so is the speed of ruin. A 20-slip cold streak at 5% sizing puts you down 100% of bankroll — you're broke.
These ranges assume the kind of small edges most DFS players actually have. If your edge is larger, Kelly math (in the optional section below) says you can size larger. If you don't know your edge size, default to the conservative end. The cost of being too conservative is slow growth; the cost of being too aggressive is going broke before your edge materializes. The asymmetry favors caution.
Why flat sizing matters
Once you've picked a percentage, you stick with it. Same percentage after a win as after a loss, after a hot week as after a cold one. Flat sizing is the foundational discipline of bankroll management, and most casual players don't follow it.
The temptations both directions are real. After a winning night: I'm running good, I should press while I'm hot. That's not Kelly math, it's tilt math. After a losing night: I need to make this back tonight. Also tilt math, in the more dangerous direction.
The math behind flat sizing is what our variance guide walked through. Single-game player outcomes are noisy. Even a real positive edge produces multi-week losing stretches that look terrible mathematically and feel worse emotionally. The point of flat sizing isn't optimal growth — full Kelly does that. The point is survivable variance. You stay in the game long enough for your edge to show up.
Pick a percentage, write it down, and don't change it because of how last week went.
How fast bankrolls actually grow
Even with a real positive edge, bankroll growth is slow. Concretely slow.
Take our hypothetical player from the variance article — 12% true cash rate on 4-pick 10x slips, slightly above the 10% break-even. Size at the aggressive 5% of starting bankroll: $5 per slip from $100.
The math per entry:
- Win (12% probability): get back $50 → net +$45 (+45% of starting bankroll)
- Loss (88% probability): lose $5 (-5% of starting bankroll)
- Expected value: 0.12 × 45% + 0.88 × (-5%) = +1.0% per entry
To double your bankroll, expected entries needed: roughly 100. That's about a baseball season for someone playing 1-2 slips a night. One full season to double on a real positive edge at the aggressive end of conventional sizing.
At the conservative 1% sizing on the same edge, EV per entry drops to +0.2%; expected entries to double rises to about 500 — three baseball seasons. The trade-off is much longer survivable cold stretches.
Two caveats. These are expected values — actual results land on either side, sometimes dramatically. And the math assumes the 12% rate holds. Overestimating your real edge is one of the most common ways players overbet relative to their actual Kelly fraction.
For comparison: full-Kelly sizing at 2.22% on this same slip (derived in the optional section below) gives EV per entry of +0.45%, so expected entries to double rises to about 225. Slower nominal growth than 5% sizing — but much lower ruin probability across the cold streaks that will inevitably happen along the way. That trade-off is exactly what Kelly is optimizing.
This isn't a reason to avoid DFS. It's a reason to size honestly and treat entry fees as entertainment expenses you'd be happy to spend even on a losing night.
Common bankroll mistakes
Most blown bankrolls trace to one of four patterns, none of which are about pick accuracy.
Tilt-resizing after wins or losses. Increasing stakes when you're hot, decreasing when you're cold, doubling-up to chase deficits. Covered above and in our math behind DFS losses — the most common bankroll killer.
Using non-DFS funds when bankroll is low. The bankroll exists as the boundary between DFS spending and the rest of your finances. Reaching into rent or savings or credit to refill it breaks the separation principle. The right move when the bankroll is gone is to stop playing for a while.
Not tracking results. Logging every entry (stake, contest, outcome, one-sentence rationale) makes you a better picker and gives you actual data on your win rate — so you're sizing against a known edge instead of a guess.
Treating short-term results as signal. A 15-slip cold streak doesn't mean your read is broken. A 5-slip hot streak doesn't mean you've solved the game. Variance dominates short samples. See our variance guide.
The math behind Kelly sizing (optional reading — skip to "Where RunsLeft fits" if math isn't your thing)
The 1-2% and 3-5% conventional rules aren't arbitrary — they're approximate fits to Kelly criterion sizing for the size of edges most DFS players actually have. The Kelly criterion formula:
f* = (bp − q) / b
Where b is the net payout per unit risked (decimal odds minus 1), p is your true probability of winning, and q = 1 − p. Three worked examples:
Pick'em (4-pick 10x slip, 12% true win rate). b = 9. f* = (9 × 0.12 − 0.88) / 9 = 0.20 / 9 = 2.22%. Full Kelly on a 12%-true slip is close to but slightly above the conservative end of the conventional sizing range.
Sportsbook (-110, 53% true rate). b = 0.91. f* = (0.91 × 0.53 − 0.47) / 0.91 = 0.012 / 0.91 = 1.3%. Small edge, small Kelly fraction.
Sportsbook (-110, 55% true rate). f* = (0.91 × 0.55 − 0.45) / 0.91 = 0.050 / 0.91 = 5.5%. Note what just happened: a 2-percentage-point increase in win rate (53% → 55%) quadrupled Kelly's recommended sizing (1.3% → 5.5%). Edge size matters non-linearly — a small accuracy improvement in your true win rate has a large effect on optimal stake.
Fractional Kelly. Full Kelly is theoretically optimal for long-run growth, but it tolerates drawdowns most humans psychologically can't. Bankroll drops of 30-50% are normal at full-Kelly sizing even on a positive-EV strategy. Half-Kelly (1.11% on the pick'em case) cuts the drawdown roughly in half while sacrificing only ~25% of growth rate. That's why most practitioners use fractional Kelly: bankroll discipline includes sizing within what you can emotionally tolerate, not just what's theoretically optimal.
The big caveat. Kelly only works if you actually know your true win rate. Overestimating your edge — easy to do, especially after a hot streak — leads to overbetting and faster ruin than the math suggests. When in doubt, defer to fractional Kelly and the conservative end of the conventional ranges.
Where RunsLeft fits
RunsLeft surfaces edge picks — spots where our model's probability estimate exceeds the market's by enough to clear the operator's rake and vig. What we don't do is tell you how to size bets against those edges. The bankroll math is yours; the edges are operator-agnostic. Our tonight's DFS edges page works for PrizePicks, Underdog, DraftKings Pick6, salary-cap DFS, and sportsbook props — but whichever wrapper you choose, the sizing decisions in this article apply unchanged.
Where to go from here
If you're new to DFS entirely, our intro guide covers the format basics. The full series — math, variance, props, operators, legality, glossary — is at /learn. The bankroll discipline this article covered is the foundation everything else builds on; the math is unsexy but it's load-bearing.