A positive EV bet finder works out a fair price for each outcome from a reference market with the bookmaker's margin taken out, then compares it with every bookmaker's odds. When a price is higher than fair by more than a set amount, it gets flagged as positive expected value (+EV). The flag is only as good as the fair price behind it, and that fair price depends on choices the finder makes for you: which market it trusts, how it removes the margin, and how fresh its prices are. This guide shows how those choices work with an invented NBA example, then the checks to run before you take a flagged bet.
How a positive EV bet finder works
Every finder does the same four jobs.
- Picks a reference market. This is the price the finder treats as closest to the true chance: a sharp market, an exchange, or a consensus of many bookmakers.
- Removes the margin. The reference prices add up to more than 100% because of the bookmaker's cut. The finder strips that out to get fair probabilities. This step is called de-vigging, and our guide to what vig is covers where the margin comes from.
- Compares every bookmaker. For each outcome, expected value is fair probability × decimal odds, minus 1.
- Flags anything over a threshold. Usually a minimum edge, such as 2% or 3%, so tiny edges don't clutter the list.
BetBank's positive EV finder does this across Australian bookmakers and lets you compare the edge under different fair-price methods on the same bet. If you're new to the idea itself, start with what positive EV betting is.
A worked example (invented numbers)
Take an NBA head-to-head market. Tied NBA games go to overtime (NBA Official Rulebook), so there are only two outcomes. Say the reference market prices it like this:
| Outcome | Reference odds | Implied probability |
|---|---|---|
| Favourite | 1.30 | 76.92% |
| Underdog | 3.60 | 27.78% |
| Total | 104.70% |
The extra 4.70% is the margin. To get fair odds, the finder has to decide how much of that 4.70% to take off each side. That decision is the first thing to check.
Check 1: which de-vig method set the fair price?
There are four common ways to remove the margin (LSports: Margin Methods Overview):
- Multiplicative scales every probability down in proportion, so the favourite carries most of the margin.
- Additive takes the same amount off each outcome.
- Power raises each probability to a power that makes them add to 100%, which takes relatively more off the longshot.
- Shin models the margin as protection against better-informed bettors, and also takes more off the longshot.
The reason the last three exist is the favourite-longshot bias: on average, bettors overvalue longshots and undervalue favourites (Wikipedia: Favourite-longshot bias). So bookmakers tend to load more margin onto the longshot, and a method that takes the margin off evenly can make the longshot look fairer than it is.
Here's what each method does to our example:
| Method | Fair odds: favourite | Fair odds: underdog |
|---|---|---|
| Multiplicative | 1.361 | 3.769 |
| Additive | 1.341 | 3.933 |
| Power | 1.330 | 4.028 |
| Shin | 1.341 | 3.933 |
In a two-outcome market like this one, Shin and additive land on the same answer. Now suppose one bookmaker has the underdog at 3.90 and another has the favourite at 1.38. Here's the EV of each, depending on the method:
| Method | Underdog at 3.90 | Favourite at 1.38 |
|---|---|---|
| Multiplicative | +3.47% | +1.39% |
| Additive | -0.83% | +2.91% |
| Power | -3.18% | +3.74% |
| Shin | -0.83% | +2.91% |
The same underdog price is a +3.47% flag under one method and a losing bet under the other three. The favourite is positive under all four. A finder set to multiplicative with a 2% threshold would show you the underdog and hide the favourite. One that uses power would do the opposite.
What to do: look at which method the finder used. Trust flags that stay positive under every method. Be wary of longshot flags that only appear under multiplicative.
Check 2: is the reference price fresh?
A finder can only be as current as its reference market. If the reference hasn't moved after news but a bookmaker has, or the other way round, the gap between them looks like an edge when it's really a delay.
In the NBA, the main news is player availability. Teams must name any player whose participation may be affected by 5 p.m. local time the day before a game, then file a game-day report, usually between 11 a.m. and 1 p.m. local time, with earlier deadlines for the second game of a back-to-back (NBA Injury Report). For Australian bettors, US afternoon deadlines tend to land overnight or early in the morning, so a price you saw the night before may be out of date. The 2026-27 regular season runs from 20 October 2026 to 11 April 2027 (FOX Sports), so this comes up most days.
Australian rules make timing matter more. Online in-play betting on sport with licensed operators is banned under the Interactive Gambling Act, apart from bets placed wholly over the phone (Mediaweek), so +EV betting with Australian bookmakers happens before tip-off.
What to do: before you bet, check the price is still there on an odds screen and check for any late team news. A big edge that turns up just after an injury report usually means one market has moved and the other hasn't caught up yet.
Check 3: is the reference market good enough for this bet?
The fair price is only reliable if the reference market is busy and well informed. Head-to-head, line and total markets on major leagues are the strongest. Player props (points, rebounds, assists, threes) are thinner, have fewer bookmakers pricing them, and often carry bigger margins. A finder can still put a fair price on a prop, but there's more guesswork in it.
What to do: ask for a bigger edge on thin markets before you act. A 2% flag on a main NBA line and a 2% flag on a bench player's rebounds are not the same quality of signal.
Check 4: is the edge bigger than the error?
Look back at the underdog in Check 1. The four methods disagree on its EV by almost 7 percentage points, from +3.47% to -3.18%. That spread is from method choice alone, before any stale prices or thin markets. A 2% edge sits well inside that range.
This is why finders use thresholds, and why a threshold that suits main markets may be too low for props or longshots.
What to do: set your minimum edge with the market in mind. Ask for more on longshots and props, where method choice and thin markets move the fair price most.
Check 5: how many bets before the results mean anything?
Even a genuine edge looks like noise for a long time. Say the multiplicative edge on the underdog were exactly right: +3.47% at odds of 3.90, with a $10 flat stake every time. Using a normal approximation:
| Bets placed | Average result | Typical swing (1 standard deviation) | Chance of being behind |
|---|---|---|---|
| 100 | +$35 | ±$172 | about 42% |
| 500 | +$173 | ±$385 | about 33% |
| 1,000 | +$347 | ±$544 | about 26% |
| 3,000 | +$1,041 | ±$943 | about 13% |
These numbers assume the edge is real. If the fair price was wrong, more bets only make the loss bigger. After 1,000 bets you still can't tell the two apart from profit and loss alone.
What to do: track whether your prices beat the closing line, not just whether you're up. Closing line value shows whether you're beating the market much sooner than results do. Keep stakes small enough to ride out swings like the ones above, and read our bankroll management guide before you size anything.
Check 6: can you actually get the bet on?
The edge only counts if the bookmaker takes your stake. Bookmakers watch for accounts that keep beating their prices, and they cut maximum stakes, sometimes heavily (Wikipedia: Arbitrage betting). In May 2026, Tabcorp blocked a number of professional gamblers, saying they no longer fit its risk management framework (The Drop).
What to do: plan for limits. Use only bookmakers licensed in Australia, and record the stake you were actually allowed, not the one you asked for. A +EV flag at a price you can only bet $5 on is worth much less than it looks.
What to look for in a positive EV bet finder
| Feature | Why it matters |
|---|---|
| Shows its fair odds and method | So you can run Check 1 yourself |
| More than one de-vig method | So you can see if a flag holds up |
| Fresh prices | Delayed feeds show edges that have already gone |
| Australian bookmaker coverage | More real prices you can actually bet |
| Minimum edge filter by market | So props and longshots need a bigger edge |
| Closing line and bet tracking | So you can measure your edge before results catch up |
A free finder with delayed prices is fine for learning the maths. Before you pay for one, check the first three rows: they decide how many flags survive the checks above.
See the NBA board for yourself
BetBank's NBA odds board shows head-to-head, line and total prices side by side across Australian bookmakers, with player props on Silver, and Gold adds line movement and expected value picks. You can try the positive EV finder on a free account with delayed odds. For the NBA season, first-time members on a weekly or monthly Silver or Gold plan pay half price for their first 2 months after a 7-day free trial, if they start by 30 November 2026. The details are on the BetBank NBA season page. Whatever tool you use, run every flag through the six checks before you bet.
Positive EV betting doesn't promise a profit: fair prices are estimates, results swing for a long time and accounts get limited. Only bet what you can afford to lose. If gambling stops being fun, you can block yourself from all licensed Australian online and phone betting with BetStop. You must be 18+ to bet. Please gamble responsibly.