Expected Value Doesn't Care What You're Betting On: Why the Math Travels Across Sports
A bettor grinding NFL spreads and a bettor hunting promotional value at online casinos look like they're doing different things. Mechanically they aren't — both are estimating a probability, comparing it to a price, and acting only when the gap is in their favor. The market changes; the decision rule doesn't.
The rule itself is the easy part. The hard part — the part that separates winning bettors from losing ones — is producing the probability estimate that goes into it.
The One Formula Underneath Everything
Expected value is calculated the same way whether you're betting on the Chiefs to cover, Real Madrid to win La Liga, a CS2 map handicap, or claiming a no-deposit promo:
EV = (Probability of winning × Amount won) − (Probability of losing × Amount risked)
Plug in real numbers. A +150 underdog you believe has a 45% chance of winning:
EV = (0.45 × $150) − (0.55 × $100) = $67.50 − $55 = +$12.50
Per $100 risked, this bet returns +$12.50 in expectation. The bet would be -EV — a money-loser long-term — if your true probability estimate dropped below roughly 40%, the implied probability of +150 American odds.
That's the decision rule. Point spreads, totals, moneylines, Asian handicaps, esports map winners, casino bonus terms — all reduce to the same comparison between an estimated probability and an implied one. What the rule does not do is tell you how accurate your estimate is. The formula treats 0.45 as a fact; in practice it's the output of a model, and the quality of the bet is bounded by the quality of that model. Two bettors using identical EV math but different probability estimates end the season in very different places.
Why the NFL Is the Worst Place to Find an Edge — and Where That Logic Sends You
Sharp money moves the line within minutes of the opener, and by kickoff the closing number reflects almost every piece of public information available.
Some of that information is directly observable in the market itself — for example, publicly available betting splits and consensus picks that show where the majority of bets are landing. The market hold on a standard -110/-110 spread is 4.55%, but the practical edge available to a recreational bettor is far smaller, because the line itself is already very accurate.
Edges live where pricing attention is thinner. Mid-tier European soccer leagues and prop markets are priced by smaller analyst teams and trade at lower limits than NFL sides. Esports markets — CS2, Valorant, League of Legends — historically had even less pricing depth, with closing lines that moved sharply on small action.
Thin doesn't stay thin. Books adapt, sharp action makes soft markets less soft, and lines exploitable in 2020 are often closer to efficient by 2024. The work of finding edge is partly the work of finding the next inefficiency before the rest of the market does.
Bankroll Math: The Other Half of the Job
Knowing a bet is +EV doesn't tell you how much to risk. That's a separate calculation, and skipping it has bankrupted plenty of mathematically correct bettors.
Most disciplined bankroll systems land in the same range:
- Flat staking. Risk a fixed percentage (typically 1–2%) of bankroll per bet. Slow, reliable, hard to blow up.
- Percentage staking. Risk a fixed % of current bankroll. Compounds during winning streaks, contracts during losing ones.
- Kelly criterion. Risk a percentage equal to your edge divided by the odds. Mathematically optimal for long-term growth but punishingly volatile; most professionals use fractional Kelly (one-quarter or one-half) to soften variance. The original 1956 derivation by John Kelly Jr. at Bell Labs concerned information transmission rates, not gambling — the proof that proportional staking maximizes long-run growth rate applies identically.
Whichever method you pick, it has to survive losing streaks. A bettor with a 55% true win rate at -110 odds will still see runs of eight or nine losses in a row over a long enough sample. A 10% per-bet stake gets eaten alive by that variance even when the underlying edge is real. For practical mechanics on how variance interacts with positive EV, our guide on successful sports betting risk management covers the details.
Where Promotional Value Sits in the Same Framework
A no-deposit bonus has a calculable expected value, but it's calculated differently from a sports bet, and the popular shortcut is wrong often enough to cost people money.
Six variables determine bonus EV: bonus amount, wagering requirement (typically 30–50× the bonus), game contribution (slots usually 100%, table games often 10% or less), game RTP / house edge, max cashout, and game volatility.
The mechanically correct way to think about EV during the wagering phase is in terms of total turnover. To clear a 40× requirement on a €10 bonus, you wager €400. On a 96.5% RTP slot, the house edge is 3.5%, so expected loss across that turnover is €400 × 0.035 = €14. Net of the €10 credit:
EV ≈ Bonus − (House edge × Wagering × Bonus)
EV ≈ €10 − (0.035 × 40 × €10) = −€4
Negative — and that's the typical case. Most no-deposit bonuses with 40×+ wagering on standard-RTP slots are slightly −EV in raw terms; their value comes from edge cases — lower wagering multipliers, lower-house-edge games where they qualify, or asymmetric max-cashout structures where one high-variance hit produces an outcome the average doesn't capture. A 25× wagering on a 97.5% RTP slot flips positive: €10 − 0.025 × 25 × €10 = +€3.75.
This is why the popular Bonus × RTP^Wagering shortcut — which would estimate the same €10 / 40× / 96.5% bonus at roughly +€2.40 — is misleading. It models the bonus as if you re-wager the full balance every spin and treats the surviving balance as expected value, which isn't how a wagering requirement works.
Whether a specific promotion is worth claiming depends on inputs that change every offer, so the practical move is to read the terms carefully or use a source that already has. European jurisdictions have run some of the better-structured no-deposit terms — the Netherlands' regulated market operates under the Kansspelautoriteit (KSA) since 2021. Aggregators that track current terms across operators — for example, listings of dutch casino no deposit bonus offers — exist for the same reason soccer pick aggregators do: most users aren't going to read forty T&Cs a week.
What This Looks Like in Practice
A disciplined bettor's workflow has roughly the same shape across markets:
1.) Estimate true probability — and know your confidence in the estimate. For NFL spreads: power ratings adjusted for injuries, weather, rest. For soccer: expected-goals models calibrated against historical results. For a casino bonus: the house-edge-on-turnover calculation above. The estimate always carries error; the bettor who pretends it doesn't goes broke faster.
2.) Compare to implied probability. Convert the price to a probability and check whether your estimate is meaningfully higher than implied — meaningfully meaning "by more than the noise in your model."
3.) Size the bet for variance, not just for edge. Apply a staking rule that survives losing streaks. Kelly for the aggressive, fractional Kelly or flat staking for everyone else.
4.) Track closing line value or post-clearance return. Did the price move toward or away from your bet after you placed it? CLV is one of the few metrics that distinguishes a real edge from a hot streak, because it's measured against the market rather than against outcomes.
What's not in the list: hot takes, public sentiment, "feel," parlay-of-the-day picks. The four steps above can produce repeatable edge across NFL, soccer, esports, and promotional inventory — but only to the extent the underlying probability estimates are calibrated.
The decision rule is portable. The skill is in the inputs.
For model-driven numbers to compare against your own estimates, see our AI sports picks.
FAQ
How do you turn American odds into an implied probability?
Does a positive expected value bet always win?
Why do many bettors use fractional Kelly instead of full Kelly?
What does the vig do to expected value?
Build a streak today
15 in a row = $100 · 21 in a row = $1,500 · Free to play.
The Psychology of Winning: How Ambition, Competition, and the Desire to Succeed Shape Human BehaviorNovember 12, 2026
How Soccer Inspired Zoom SoccerNovember 9, 2026
The 90-Minute Trap: Why a Correct Football Pick Can Still Lose Your BetOctober 22, 2026