The concept
Betting on value,
not on teams.
Most people bet on who they think will win. We don't care who wins. We care whether the price is wrong. That single shift is the entire difference between gambling and investing.
What “expected value” actually means
Every price implies a probability. Odds of 2.00 imply the bookmaker thinks it's a 50% shot. If the true chance is actually 55%, then a bet at 2.00 is underpriced, and over hundreds of bets like it, the maths is unavoidable:
A +4% EV bet returns, on average, 4 cents on every dollar staked, win or lose on the day.
You will lose plenty of individual bets. That's expected. The edge shows up the same way a casino's does: not on any one hand, but across thousands of them.
Why mispriced odds exist at all
Bookmakers aren't omniscient. Lines are set fast, across tens of thousands of markets, and they move on money, not just on truth. A backup player is ruled out; a niche league gets little attention; a local book lags the rest of the market by minutes. Each of those is a crack, a moment where the offered price drifts away from the real probability. There are more of these cracks than you'd think; the hard part is finding them fast and consistently.
How we find them
We continuously compare the prices on offer against the sharpest markets in the world, the books and exchanges that professionals trust as the closest thing to a true price. When a line is meaningfully more generous than that consensus, it's flagged. We don't chase narratives or “locks.” We surface the small, repeatable pricing errors and let volume and discipline do the rest.
(We keep the exact sources, models and filters proprietary: that edge is the product.)
Discipline is the strategy
An edge means nothing if one bad week wipes you out. Every pick comes with a recommended stake sized to your bankroll using fractional Kelly, big enough to grow, small enough to survive variance. Flat, patient, unemotional. The boring approach is the profitable one.
We show our work
Anyone can claim a winning record. Every pick we make is logged the moment it's found and graded honestly against the real result: wins, losses and all. The track record on the home page is that raw ledger, not a highlight reel.
How we measure the edge
The closing line is the sharpest price a market ever offers. By kickoff the professionals have bet it into shape. Taking a longer price than the close is called closing line value, and it is the standard evidence that an edge is real rather than a run of luck. It also shows up long before profit does, which matters when a few hundred bets is far too small a sample to judge returns on.
Every settled pick that captured a closing price is measured against an independent book's de-vigged fair in the 30 minutes before kickoff, never against our own pricing, which would be marking our own homework. Markets with fewer than 20 settled picks are left out of the per-market table entirely, because their intervals are too wide to say anything either way.
One caveat we would rather state than bury. Player props settled before Jul 24 were closed against the same book that priced them, which measures a book against itself. The figure quoted on the home page is restricted to the picks settled since, where the closing reference is genuinely independent.
What we removed, and why
The published curve covers only the markets we still send. When a segment shows no edge against the closing line we stop sending it, and we remove its whole history from the record, not just the days after we dropped it. Leaving the good years in and cutting the bad ones is how track records get flattering.
Excluded so far: soccer 1X2, soccer Unders, rugby league, tennis, baseball pitcher strikeouts (23 August 2026), and, as of 15 September 2026, soccer below the second division, plus women's soccer. Exhibition fixtures and hand-placed bets are excluded too. All of it is still priced, settled and tracked in the background; it is just not the product. Restated 8 August, 23 August and 15 September 2026.
The September restatement improved the curve too, so here is that reasoning. We measured how much of a pick's claimed edge survives to the closing line, and found it depends on the league, not just the market. Across 595 settled soccer picks in 207 leagues, third division and below gave back 9.7 points of claimed edge by kickoff (interval −11.6 to −7.8) and women's soccer 9.1, against 3.3 for major top flights: intervals that do not overlap. Thin markets move against you because the price you took was never really the price. Removing those leagues took 34 settled picks and −4.0 units out of the published record. As with August, the decision was made on the drift measurement rather than on the losses: the losses are what a claimed edge that was not real looks like afterwards.
The August restatement improved the curve, so here is the reasoning. Pitcher strikeouts were 52% of everything we published and carried a flat return of −8.6%, so removing them lifts the line by about 20 units. We are not asking you to take that on trust: the decision was made on closing line value, not on results. Measured against a genuinely independent book it came to +0.25%, with an interval from −0.88% to +1.38%: no demonstrable edge. Closing line value is the criterion precisely because, unlike a return, it cannot be chosen after the fact to suit the answer.
Ready to bet on value?
Up to one pick a day is free. $19.99/mo gets you every +EV pick in real time, with the staking already worked out, delivered to Discord the moment it's found, usually a few hours before the event so there is time to get the price on.
Volume follows the sporting calendar rather than a schedule: picks arrive in bursts around the fixtures we cover, and between seasons there are genuinely quiet stretches. The scanners keep running throughout. They simply find fewer prices worth taking, and we would rather send nothing than send filler.