Expected value, usually shortened to EV, is the average result you would get from a bet if you could place it over and over. Positive EV means the price is better than the real chance of it happening. Negative EV means it is worse. Almost every bet offered to you is negative EV, which is how bookmakers pay for stadium naming rights.

The idea in one example

Suppose a match is genuinely a 40% chance for the underdog. The price that exactly matches 40% is 2.50, because 1 divided by 2.50 is 0.40. That is the fair price. At 2.50 you would break even forever.

Now one bookmaker is offering 2.65 on that same team. Here is what a $100 bet is worth on average.

Six dollars on a hundred, so a 6% edge. There is a shortcut worth remembering: divide the price you are getting by the fair price and subtract one. 2.65 divided by 2.50 is 1.06, which is your 6% without any of the arithmetic above.

Where the fair price comes from

The whole thing rests on knowing the true chance, and nobody knows it exactly. What you can do is estimate it well, and there are two common ways.

Ask a lot of bookmakers at once

Any single book's price includes its own margin and its own opinion. Average the price across every book quoting the market and the individual quirks cancel out, leaving something close to the market's collective view. When one book is well outside that consensus, either it knows something the others do not, or it is slow.

This is what the Fair Odds figure on every Bonus Basher card is. It is the average back price for that selection across all the books in the scan. If you are backing at 2.65 and fair odds says 2.50, that price is doing something the rest of the market is not.

Take the margin out of a sharp price

The other approach is to trust one market that is usually right, typically an exchange like Betfair where people bet against each other rather than against a bookmaker, and strip out the commission and spread to get a clean probability. That is where the term devigging comes from, since the vig is the margin. The scanner's alert engine uses this method for its +EV picks and labels them Betfair Devig or Consensus depending on which was available.

How much to stake

Knowing a bet is good does not tell you how big to make it. Bet too small and a real edge earns nothing. Bet too large and an ordinary losing run takes the bankroll before the maths gets a chance to work.

The standard answer is the Kelly criterion. In plain terms, it says stake in proportion to how big your edge is and how long the odds are. The arithmetic is the edge divided by the odds minus one. For our 6% edge at 2.65 that is 0.06 divided by 1.65, which is 3.6% of your bankroll. On a $1,000 bank, full Kelly says $36.

Most people do not bet full Kelly, and for a good reason. Full Kelly is correct only if your estimate of the edge is correct. Yours is an estimate built on other people's prices, so it is a bit wrong, and being a bit wrong at full Kelly produces swings that are genuinely unpleasant. Half Kelly gives up a small amount of long run growth for a much smoother ride, and quarter Kelly is smoother again. Half is the common choice.

Set Min +EV edge in Settings while you are there. It is the point below which a bet is not worth the effort or the account wear. Two percent is a reasonable line, because an edge under that is easily inside the error of your own fair price estimate.

Promos are expected value too

A bonus bet converted at 68% is not a different kind of thing. It is a bet with a very large edge and almost no variance, which is why promo work is the sensible place to start. You are not trying to out predict anyone, you are collecting a known margin the bookmaker has handed you as a marketing cost.

It is also why the conversion percentage on a scan result and the edge percentage on a value bet can be read the same way. Both are answering how much of your money you expect to keep, or gain, on average.

Proving it to yourself

Expected value is a claim about the long run, so the only way to know whether yours is real is to record every bet and compare. The Tracker keeps the fair odds you had at the time in its own column next to the price you actually took, so months later you can look back and see whether the bets you called +EV actually behaved like it.

Your Dashboard plots realised profit over time from the same records. A curve that wanders up and down but trends up is exactly what a small edge looks like. A curve that only goes down is telling you the model is wrong, and that is worth knowing early.