How to Survive Variance as a +EV Bettor
An edge is real, but it only shows up across hundreds of bets, so you have to be built to last that long.
This lesson walks through the no-vig math behind a +EV bet, shows how large the normal swings are around a small edge, and explains how a price move, oversized bets or too few bets can break a sound strategy.
Bet a flat 1% of your bankroll, track the price you actually got, and set a minimum edge that matches how many bets you can realistically place.
What makes a bet +EV
A +EV (positive expected value) bet is a price that pays more than the outcome's fair chance deserves. In the video's example, a sharp sportsbook prices the Knicks at -130 and the Spurs at +110, which convert to implied probabilities of 56.5% and 47.6%. Those add up to 104.1%, and the extra 4.1% is the vig, the book's built-in cut.
Removing the vig gives the no-vig (fair) probability, which puts the Knicks at about 54.3% to win. A softer book still hanging the Knicks at -110 only needs 52.4% to break even, meaning that is the win rate where the price neither makes nor loses money over time. On a $100 bet, that gap is worth about $3.60 on average every time you place it.
One bet, two prices
From the sharp book's -130 / +110 market
About +$3.60 per $100 on average
About -$0.50 per $100 on average
Why good bets still lose so often
No single bet ever pays the $3.60. You either win about $91 or lose $100, and the $3.60 only appears as an average over many bets. In the video's math for a 54% winner at -110 with flat $100 bets, the expected profit after 100 bets is about $360, but the standard deviation (the typical swing around that average) is about $950.
Because the edge is smaller than the noise, the chance of being below zero after 100 bets is about 35%, and after 500 bets it is still roughly one in five. You can estimate that probability before you bet from two inputs, your average edge and your bet size, so a red stretch becomes something you expected rather than a reason to quit.
Price beats hit rate, and prices move
+EV is a property of the price, not the game. If the same Knicks bet moves from -110 to -120 ten minutes later, the break-even rises to 54.5%, above the 54.3% fair number, and a bet worth about $3.60 is now worth about -$0.50. Bettors working by hand often find good bets too late, and across 100 bets that can turn about $360 of expectation into roughly zero.
The same logic explains why a low win rate can still be profitable. When you take plus-money prices (odds that pay more than you risk) on underdogs that win more often than the price implies, you can lose far more bets than you win and still come out ahead. Hit rate tells you how often you cash, but price decides whether you profit.
A spreadsheet that records the price you spotted will still call a bet +EV after the line has moved. Record the number you actually bet at, and if the price has moved past your threshold, pass, even when it stings.
Bet a flat 1% so a cold run cannot end you
Bankroll sizing is the threshold that sinks the most +EV bettors. The rule in the video is roughly 1% of your bankroll on every bet, the same percentage whether you love the play or barely like it.
Most people do the opposite and size up on the plays they feel best about, but confidence is exactly where variance hides. At 1% per bet, an ordinary cold run costs you a slow week, while betting 5% or 10% on your favorite plays lets the same cold run end your account.
Match your minimum edge to your volume
Volume is the second threshold, because the edge only shows up in bulk. Under a few hundred bets you have not really run the experiment yet, which is like grading a coin after six flips. If you cannot realistically log 300 to 500 quality bets, your results are noise in either direction.
The third threshold is an edge floor, the minimum EV you require at the price you can actually get. An edge under 1% can work when it is spread across thousands of bets on an exchange, but at normal manual volume the video's floor is roughly 3% EV. Pick a lane: a thin edge with heavy volume, or a thicker edge with lighter volume.
Find a +EV bet by hand
Pull the sharp line
Find the same market at a sharp book and note the prices on both sides.
Strip the vig
Convert both prices to implied probabilities and scale them so they add up to 100%. That is your fair probability.
Convert the soft price
Turn the price at the book you can bet into its break-even percentage.
Compare
If the fair probability is above the break-even and the EV clears your edge floor, the bet qualifies.
Size it and recheck
Stake your flat 1% unit, and confirm the price has not moved before you place the bet.
Let +EV Sniper run the same math live
Doing this by hand takes a few minutes per bet, which is often longer than the edge lasts. +EV Sniper runs the same steps live across books: it pulls prices from sharper books, strips the vig, finds the fair number and sorts every row from most valuable to least. In the video's example, the top row was a baseball total at Fanatics showing about +7.5% EV, or about $7.50 per $100 on average.
Filters turn your thresholds into settings. Requiring at least three books means you never trust a fair price built on a single stale line, and a game-line weighting option sets how the fair value is calculated. The math stays the same, so learn the method first and automate only what you understand.
Variance survival checklist
- The edge is real: you can run the no-vig math yourself on any market.
- Variance is bigger than intuition: being down after 100 +EV bets about one time in three is normal.
- Bet a flat 1% of your bankroll on every play, whatever your confidence.
- Log real volume, and set an edge floor that matches it.
- Track your next 100 bets at the price you got before you let a red stretch tell you anything, then aim for 500.
Article FAQs
Why do +EV bettors lose so many bets?
Each bet still wins or loses on its own, and the edge is small compared with the normal swings. In the video's example, a sound strategy has about a 35% chance of being down after 100 bets and roughly a one-in-five chance after 500.
How many bets do I need before judging my results?
Under a few hundred bets, results are mostly noise. Log at least 100 bets at the price you got before reading anything into a red stretch, and 300 to 500 quality bets before judging the strategy.
How much should I bet on each play?
The video uses a flat 1% of bankroll on every bet, regardless of how confident you feel. Sizing up on favorite plays is how an ordinary cold run turns into a lost bankroll.
What minimum EV should I require?
It depends on your volume. An edge under 1% only adds up across thousands of bets, so at normal manual volume the video uses a floor of roughly 3% EV at the price you can actually get.
Open +EV Sniper
Scan books live with the vig already removed and every row sorted by EV, so you can stop at your edge floor.