If you've been net-down at DFS for a few months and you're starting to wonder whether it's just bad luck or something structural, you're not imagining it. Most casual DFS players lose money over time, and there's a reasonable explanation that has three parts — only one of which is bad luck.
This article walks through the math honestly. We're not going to pretend the structural math is friendlier than it is, and we're not going to talk you out of playing. Some readers are losing because the operator takes a cut and the math is doing exactly what you'd expect. Some are losing to normal variance on a real edge. Some are losing because of specific, fixable mistakes that compound the math problem. Each has a different fix — or, in the case of the structural math, a different framing that lets you decide whether the entertainment is worth the cost. By the end you should be able to figure out which is happening to you.
The structural math you can't beat
Every paid DFS product has a built-in margin for the operator. You have to clear it before you start winning. There are two main flavors, and they look different on the screen but behave the same way over time.
Sportsbook vig. When you bet a prop at a sportsbook at -110 on both sides, the implied probability adds to roughly 105% across the two sides — that 5 percentage points above 100 is the book's built-in margin. Player props typically run wider — closer to 6-8% margin on a -125/-105 line — meaning you need to win at least 52-54% of your bets just to break even. We walked through the math in detail in our sports betting vs. daily fantasy sports guide.
DFS operator rake. Pick'em apps build their margin into the payout multipliers rather than into the line price. PrizePicks pays 10x on a 4-pick power slip if all picks hit, but the fair-EV payout at 50/50 independent picks would be 16x — that's around a 35% shortfall versus fair, baked into the payout structure. The exact effective house edge isn't publicly disclosed and depends on how you calculate it; analyst estimates of effective hold for pick'em operators typically land in the 15-25% range, depending on contest format and how accurately the operator sets the underlying lines. Whichever number you anchor on, the structural fact is unambiguous: there's a sizeable house edge built into every contest, and you have to clear it before you start winning.
You can't make this go away. It's not a flaw in the product — it's how the operators make money. The honest framing is that DFS is structurally an operator-rake market like a casino card room or a sportsbook, and the structural math is unfriendly to the average player by design. That's not anti-DFS; it's how to read the menu.
Why even good players have bad stretches
The structural math is one source of losses. Variance is the other source you can't avoid, and it gets less attention because it's less satisfying to talk about.
Even players with a real positive edge can have losing stretches that last weeks or months. Props and pick'em slips resolve on single-game player outcomes, which are noisy by definition — even a hitter projected to clear 1.5 hits at a 60% true rate will miss in 4 out of every 10 starts. A 4-pick power slip with a real 12% true probability of cashing (above the 10% break-even rate) wins one entry in 8.3 on average, but the gaps between wins are highly variable. You can run 15 slips in a row with no winner and not have done anything wrong; you can also win three of your first five and feel like a genius for the wrong reason.
Multi-week downswings are normal at casual playing volume even when the underlying read on each pick is sound. Variance compresses with volume — a player running 50 slips a day across a season converges on their true rate much faster than someone playing two slips a night — but most casual players don't put in enough volume per month for it to compress meaningfully on a single-month timeframe. We covered per-game-outcome variance for individual props in our guide to MLB prop bets; a future piece in this series will walk through the variance math specifically.
The mistakes that compound the math problem
The structural math and variance are unavoidable. The mistakes that compound them mostly aren't. These are the patterns that turn an already-tough product into a much worse one, and they show up across players who otherwise know what they're doing.
Chasing high-multiplier longshots without the probability work. A 6-pick power slip that pays 35x looks tempting until you do the math: you need a true probability around 2.9% on the full slip to break even (1/35), which compounds to about 55% per independent pick. The marketing leans on the dollar amount of the payout; the math leans on the probability you can actually deliver. Demon picks and big parlays compound the rake/vig problem — they're not free upside.
Playing on tilt after losses. The most expensive mistakes most casual players make come immediately after a bad night. Doubling your entry size to chase the deficit. Playing twice as many slips as usual. Picking aggressively to "make it back tonight." None of this changes the underlying math — but it does change the size of the losses. A normal losing night becomes a structurally bad one because you've stacked entries when you were emotionally compromised.
Not understanding line-shopping. The same prop can be priced differently across operators. Aaron Judge over 1.5 hits might be -120 at FanDuel and -115 at DraftKings. Taking the first line you see without checking elsewhere leaves expected value on the table that compounds across hundreds of bets.
Picking based on names and teams rather than the math. Aaron Judge is a Hall of Fame talent and a fun player to bet, but "I like Judge tonight" is not a probability estimate. The market already knows Judge is good — his line is priced accordingly. The edge isn't in picking famous players; it's in finding spots where your read on a player differs from where the market has set the line.
Playing markets you don't actually understand. An NFL player who suddenly bets MLB props during baseball season because that's what's in season is operating without the baseline knowledge they have for football. The structural math is hard enough; adding "I don't really know this sport" makes it much harder.
What profitable DFS actually looks like
Profitable DFS players exist. By definition, in any contest where the operator takes a service charge, total payouts are less than total entries — meaning more than half of all dollars wagered have to come out of players' pockets net. Industry estimates have consistently suggested the profitable portion is a small minority, often cited in the 5-15% range, though the exact figure varies by study, contest type, and operator. The structural point matters more than the specific number: most players lose, and that's a mathematical consequence of the format, not a sign that players are bad at picking.
The players who land in the profitable minority share a recognizable profile. They treat DFS as work, not entertainment. They have explicit systems for selecting picks rather than vibe-based decisions. They keep records of every entry, including the rationale and the closing line. They have bankroll management rules they actually follow. They line-shop. They specialize in specific sports rather than playing everything. And they accept multi-month variance without changing their approach.
Most readers of this article won't become that player, and that's fine. If you treat DFS as entertainment within a budget you'd happily pay for the engagement of having action on tonight's games, the structural math is something to factor in — not something that makes the product bad. If you're treating it as a way to make money, the math says you'll need a level of discipline most people don't have the appetite for.
Where RunsLeft fits
RunsLeft is built around the math problem this article is about. Our model generates probability estimates for player props each night and flags the spots where our estimated probability exceeds the market-implied probability by enough to clear the rake and vig. Those are the picks where the math is actually on your side rather than against you — and they're what we surface on our tonight's DFS edges page, with the reasoning attached.
The signal is operator-agnostic — the same edge applies whether you're playing pick'em on PrizePicks or Underdog, salary-cap DFS on DraftKings or FanDuel, or betting props at a sportsbook. The math we're trying to beat is the same math; only the wrapping changes. We don't promise winners. We promise spots where the math is friendlier than the operator's default — whether you act on them well is on you.
Where to go from here
If you're new to DFS entirely and want to start with the format basics before deciding whether the math is worth it for you, our intro guide to DFS walks through the operators, contest types, and a step-by-step example. More articles on variance, bankroll, and operator-specific strategy are on the way in /learn.