Positive EV betting mistakes, and how to fix them

The costliest positive EV mistakes are errors in the fair price, not bad luck: a benchmark that has not updated, margin left in the fair side, a one-sided or thin market treated as if it were sharp. Each one makes a losing bet look like a winning one. After those come the sizing and behaviour mistakes: overbetting, chasing, and running all your volume through one book until it limits you. Each is below with its arithmetic and its fix.

1. Trusting a stale fair price

An edge exists when a book's price differs from the fair price. It also appears when the fair price is out of date. If injury news breaks and a recreational book moves first, your benchmark still shows the old number, and the book that has already moved looks like a bargain. The edge is on the wrong side of the news.

2. Leaving the vig in the fair side

The fair price has to be devigged. Using the sharp book's raw implied probability for your side overstates the edge; using one minus the other side's implied probability understates it. Only removing the margin from both, as in the devig guide, gets it right.

3. Treating a one-sided market as fair

Some markets have only one side: anytime touchdown scorer, first goalscorer, many futures. With no opposite price you cannot measure the margin in that player's price, and these markets commonly carry more margin than a main two-way line. Assume too small a margin and a losing bet looks positive.

4. Trusting thin prop markets

Player props are where recreational books misprice most, and also where the benchmark is weakest. Even sharp books take small limits on props, so their price carries less information, and a prop line can differ between books (24.5 points at one, 25.5 at another), which is a different bet, not an edge. Fix: compare only identical lines, prefer props where several sharp sources agree, and stake them smaller.

5. Betting into one book's limits

Recreational books watch who wins. Putting most of your +EV volume through one book, always on the price it is slowest to move, is the fastest route to a limited account, and a limited account stops producing edge at all. Fix: spread volume across every book you hold, mix in ordinary bets and promotions, and move size to exchanges, where another user takes the other side and there is no house position to protect. The limits guide has the full approach.

6. Overbetting

Full Kelly on an edge that is really half what you think is double Kelly, and double Kelly grows nothing. Because every edge here is an estimate, the fix is structural: stake a quarter Kelly or less, recalculated from the current bankroll, with a hard cap per bet. The Kelly guide works the numbers.

7. Chasing and quitting

Raising stakes after losses does not change the edge; it only raises the variance, at the point you can least afford it. Quitting after a bad month reacts to noise that is completely normal for this method (see variance). Fix: size from the bankroll, not the last result, and judge the method by closing line value over at least 100 bets.

8. Chasing huge edges

A 12% edge on a main market is almost never a sharp read. It is a stale line, a wrong line, or a misread market, and books void obvious errors under their house rules and restrict accounts that take them repeatedly. The bettors who last live on 1–4% edges at volume. Fix: read a very large edge as a question to check, not a bet to max.

How the board guards against these

The Positive EV board removes rows as prices move, lets you cap quote age so stale numbers drop out, devigs every outcome before comparing, labels fair prices built without a reference-grade source as lower confidence, badges likely mispricings so you decide deliberately, and pre-fills a confidence-scaled fractional-Kelly stake. It cannot stop you overbetting or chasing; the walkthrough covers those pitfalls too.