Edge vs. consensus across every market we track.
Today’s Games
Compare the odds every sportsbook offers for the same game, and see how they moved.
More
All lists every sport with prices: the NFL, NBA, NHL and Premier League with Scoutics forecasts, and on Pro and Insider every other league and tournament as a price comparison. Pick a sport to see one league or tournament.
We don’t detect your location. Pick your state to see only the apps you can use there.
Checking today’s prices…
One moment.
Models used in Games
Forecasters estimate outcomes. Selection rules compare those probabilities with prices. A tracked forecast is not a proven profitable strategy.
Versions identify model or rule revisions; refit dates are separate. The current selection rules use Darius 3.0 probabilities.
Forecasters and market reference
| Name and role | Method | Coverage | Evidence | Record |
|---|---|---|---|---|
DariusForecaster3.0Method identityVersion-tagged record | Forecasts match outcomes from prior team performance. | EPL · NFL · NBA · NHL | Saved forecasts | View record |
BahramMarket referenceReferenceMethod identityMarket reference | Sportsbook consensus with the margin removed. | EPL · NFL · NBA · NHL | Captured market prices | View record |
Selection rules
| Name and role | Method | Coverage | Evidence | Record |
|---|---|---|---|---|
MitraSelection ruleCurrent rulesMethod identitySeparate from the older EPL Arta/Mitra research | Chooses value at a reference book, then finds the best eligible price. Uses Darius 3.0 probabilities. | EPL · NFL · NBA · NHL | Current decision rules | View record |
Mitra SelectSelection ruleCurrent rulesMethod identitySeparate from the older EPL Arta/Mitra research | Compares estimated value across eligible books and outcomes. Uses Darius 3.0 probabilities. | EPL · NFL · NBA · NHL | Current decision rules | View record |
Arta · one bookSelection ruleCurrent rulesMethod identitySeparate from the older EPL Arta/Mitra research | Uses the reference book's price for the selected outcome. Uses Darius 3.0 probabilities. | EPL · NFL · NBA · NHL | Current decision rules | View record |
Saved forecasts by version
Finished regular-season games, evaluated 30 minutes before kickoff. Lower log loss and Brier are better.
Sign in to inspect results by model version.
Sign in to view resultsCorrect means the highest-probability outcome happened. It does not measure selection profit or a future win rate. Each version can cover different games; these scores are not a matched ranking. A handful of games, such as 3 correct out of 9, is too few to judge a version: a few results either way change the score a lot.
What the numbers mean
Our forecast is a statistical model of past results and form (no simulation, no betting prices), saved before kickoff and scored on the forecast record. Market is the sportsbooks’ consensus chance with their margin removed. Priced above the other books is expected profit: the market’s fair chance times the best price, minus the stake. It never uses our forecast. Scoutics picks apply selection rules to our forecast and a captured price. How the models work. The chance needed to break even at a price is a separate number.
- Priced above the other books
- A price comparison, not a prediction. At a legal app, one side’s price returns +2% or more of the stake on average at the fair chance from the other books (each book’s margin removed, that app’s own feed left out). Under +2% is ordinary price noise. A long-run average, not what one bet returns. On paper the rule is losing so far (see the paper record below), so treat it as a price check, not a pick.
- Better price elsewhere
- A price comparison: another legal app pays at least 3% more than your sportsbook on a side. It says nothing about whether the bet is good.
- Model and market disagree
- Our forecast (the model, saved before kickoff) differs from the books’ fair chance by 8 points or more. The market may know something our model doesn’t; our model has not beaten the books on the forecast record, so this is a reason to look, not a bet.
- No clear edge
- None of the above. Passing is a perfectly good decision.
- Insufficient reliable data
- Prices older than 6 hours, too few independent books to compare, no price at your sportsbook, or no legal app in your state. We explain which.
How much to trust each call
- Price checkAnother app pays more for the same bet. It doesn’t tell you who will win.
- Losing on paper so farThe price is above the other books, but the rule has lost money on paper so far. A price check, not a pick.
- Our model onlyOur model disagrees with the sportsbooks. It hasn’t beaten them yet (record).
- Limited evidenceNHL model calls: our NHL forecasts currently show limited predictive evidence.
- Not enough dataOld prices or too few apps, so we don’t judge it.
Create your free Scoutics account today
Start free with core-sport price comparisons. Upgrade when you want more simulations and research tools. Plans pay for software access, never stakes.
Go Deeper With Paid Plans
Difference From the Market’s Fair Chance
Every market tab, plus each price’s difference from the market’s fair chance: how much more a price pays than the other sportsbooks’ fair chance implies (their chance with the margin removed × the decimal price − 1), shown with the time each price was captured. +2% means a long-run average about 2% above break-even at that chance; it is an estimate, not a promise of profit. See our forecast record.
Pro
Every Other Sport We Cover
Tennis, other soccer leagues, baseball, MMA and more: every app’s price and a market estimate. There is no Scoutics forecast for these.
Pro
Research, Not Promises
Price-gap research highlights games where sportsbooks priced far apart, each price with the time it was captured. It’s research to explore, not verified profit.
Insider
Get a Second Opinion Before Game Time
Compare every sportsbook’s price, simulate matchups and check our public forecast record. Start free and upgrade when you need more. Every number is a model estimate, never a guarantee or betting advice.