Learn

The Fair Price Audit: a positive EV scanner checklist in three questions

This positive EV scanner checklist is the Fair Price Audit: three questions to ask before you trust any +EV flag, whichever tool shows it. Whose odds set the fair price, how the margin was taken out, and how the flag was tested. Every flag is a comparison against one fair price, so the audit checks that number, not the flag.

Updated October 21, 2026 · 7 min read

A +EV flag looks like an answer, but it is really a comparison. The scanner takes a fair price for an outcome, compares it with the price a book is offering, and flags the gap. If the fair price is wrong, the flag is wrong too. The Fair Price Audit is the short checklist we use for any scanner, ours included.

It is the same three questions from the video, in the same order.

# Question What a good answer looks like
1 Whose odds set the fair price? A named sharp book or a stated market consensus, with the margin removed
2 How was the margin taken out? A named method (power, Shin, additive or normalization) you can see or switch
3 How was the flag tested? Flagged bets graded against the closing price over a large sample, not just a win-loss record

Who this is for

This is for bettors who already use, or are thinking about paying for, a positive EV tool. You don't need to code. You need the tool's own help pages or FAQ, a calculator that removes the margin from a two-way price, and about ten minutes per tool. Our free no-vig calculator and implied probability calculator are enough for every step below.

1. Whose odds set the fair price?

Every scanner needs a reference price. OddsJam, for example, describes its positive EV tool as taking the vig out of the sharpest odds and then flagging books that pay more, on its Positive EV package page. Its tools hub lists the positive EV, arbitrage, middles and low-hold tools that sit on top of that reference.

That is a clear answer, and it is the kind you want. What you are checking:

  1. The reference is named. "Sharp books" or "the market" should come with a list or a rule you can read.
  2. The reference changes by market. The sharpest price for a main line is not always the sharpest price for a player prop.
  3. You can see the fair price itself. A flag with no fair price next to it can't be audited.

Write the answer down in one line, for example: "Fair price = sharp book on main lines, consensus on props, margin removed." If you can't write that line from the tool's own pages, ask its support team before you pay.

2. How was the margin taken out?

Books build a margin into both sides of a market, so the two implied chances add up to more than one hundred percent. A fair price needs that extra removed, and there is more than one way to do it. A 2017 paper in the American Journal of Sports Science by Clarke, Kovalchik and Ingram compared methods and found the power method did better than simple normalization.

Here is the arithmetic for one two-way market, so you can check any tool by hand. The conversion from American odds to a chance is standard (moneyline odds): a minus price is price divided by price plus one hundred, and a plus price is one hundred divided by price plus one hundred.

Step Over Under Note
Offered price (demo) -110 -110 hypothetical two-way market
Implied chance 52.38% 52.38% 110 / 210
Total 104.76% the extra 4.76% is the margin
Fair chance (normalized) 50.00% 50.00% each side divided by 104.76%
Fair odds +100 +100 the number a flag is measured against

With equal prices every method gives the same answer. The methods split apart on lopsided markets and on props, where margins tend to be bigger. In the Bookmaker Margin Panel for 2026, average measured margins were 4.77% on head-to-head markets and 7.93% on player props, so the method matters most exactly where a lot of flags show up.

What you are checking:

  • The tool names its method.
  • You can switch methods, or at least see the fair price under each one.
  • The fair price on a lopsided prop still looks sensible when you redo it by hand with our no-vig calculator.

3. How was the flag tested?

A flag is a claim that a price is good. The test is whether prices like it keep beating the market's final word, the closing price. Kaunitz, Zhong and Kreiner built a strategy on exactly that idea, using the books' own published odds as the benchmark, and they also reported that bookmakers started limiting their accounts.

The reason you don't grade a flag rule by its win-loss record is that a flagged bet near even money still loses about half the time. Over a few hundred bets that noise can bury any real edge, or invent one. The closing price is available for every single bet, which makes it a much faster check.

What you are checking:

  1. Every flagged bet is logged with the price you got and the closing price for the same side.
  2. The rule is fixed before the test. "Bet every flag above 2% edge at the opening price" is a rule. "Bet the ones that felt right" is not.
  3. No look-ahead. A rule that bets at the open can't use information you only learn at the close. Our backtest lab blocks that for you, and a spreadsheet works too if you are strict about it.

A worked audit, step by step

Here is one audit on a demo flag, so you can copy the format.

Item Demo value
Flag Hawks +105 at one book (demo odds)
Reference Sharp book at -110 / -110 (demo odds)
Method Normalization, power method checked as well
Fair price 50.00%, or +100
Edge at +105 about 2.5 cents per dollar staked (hypothetical)
Closing price Hawks +100 (demo)
Result of the check Your price beat the close

The bet itself still wins or loses like a coin flip. The audit only tells you the price was right.

What to watch out for

  • Stale reference prices. If the sharp book has moved and the tool hasn't refreshed, the flag is a ghost.
  • Thin markets. Props and small leagues can have a reference price nobody really bet into.
  • Limits. Books may limit accounts that keep beating the close, as the Kaunitz paper describes. Plan your stakes around that.
  • Too many filters. Every extra filter you test makes a lucky rule more likely. Pick your rule, then test it once.

How to tell it worked

After a month, you should be able to answer three things from your own log: what share of your flagged bets beat the closing price, by how much on average, and whether that held across leagues. If most of your flagged prices beat the close, the fair price is doing its job. If they don't, go back to question one, because the reference is usually where it breaks. You can grade your own bets for free in our closing line value calculator.

Quant+ shows the reference price, the margin method and the fair price next to every line, and it does not place bets. 21+. For informational purposes only. Nothing is guaranteed. If gambling stops being fun, call 1-800-GAMBLER.

Questions people ask

What does a +EV flag actually mean?

It means the price offered is better than the tool's fair price for the same outcome. It is a statement about the price, not a prediction that the bet wins.

Which way of removing the margin is best?

There is no single answer for every market, but a 2017 comparison by Clarke, Kovalchik and Ingram found the power method did better than simple normalization. Ask your tool which one it uses and whether you can switch.

How many bets do I need before I trust a flag rule?

More than most people expect. Results swing a lot over a few hundred bets, which is why the audit grades each bet against the closing price instead of waiting for the win-loss record.

Does Quant+ place bets or promise results?

No. Quant+ is analytics software for adults 21 and over. It does not place bets, hold funds or promise any result. Nothing is guaranteed.

Sources

  1. OddsJam Positive EV Global Subscription Package, OddsJam
  2. Betting Tools Hub, OddsJam
  3. Adjusting bookmaker's odds to allow for overround, American Journal of Sports Science (Clarke, Kovalchik, Ingram)
  4. Moneyline odds, Wikipedia
  5. Bookmaker Margin Panel, Mendeley Data (E. Glush)
  6. Beating the bookies with their own numbers, arXiv (Kaunitz, Zhong, Kreiner)