Sharpstieri

Model check

Is the edge real, and can you get the prices? Every number shows how many calls it rests on.

How this works

Is the edge real, and can you get it? Every number shows how many calls it rests on; anything under 30 graded calls is marked “not enough yet.” Worth a look every Monday.

All timeLast 7 days

Data health

✓ Everything is updating normally.

Hourly health check · Wed 2:50 AM ET

Second price feed agrees on FanDuel 24 of 24, Caesars 27 of 27, BetRivers 26 of 26, theScore Bet 27 of 27 prices. Serious problems are also sent to your phone, once a day each.

Football prices checked 1 min ago · baseball full pull 8 min ago · splits 11 min ago · 4,570,752 odds-data credits left.

Are the edges real? 1190 graded

Each call's predicted edge vs how it actually did against Pinnacle's closing price. If calls predicted at +4% only beat the close by +2%, the app is overestimating, and bet sizes are too big.

Predicted edgeCallsPredictedActual vs close
+1% to +2%334+1.6%+1.0%
+2% to +3%468+2.4%+1.4%
+3% to +5%301+3.7%+1.4%
+5% or more87+7.1%+4.7%
All1190+2.9%+1.5%

The app overestimates its edges: calls do noticeably worse against the close than predicted. Bet sizes are too big for the real edge; shrink them until this recovers. (actual is 53% of predicted so far)

Can you get the prices?

From your “Got it / It was gone” taps on each Take it now (on the game page and the Track record) and the bets you log.

Calls1414
You got it63
It was gone34
Didn't see it in time491
Saw it, passed0
Not marked826
Got it, of the ones marked65% (97 marked)

Sign in on My Bets to compare your bets with the called prices.

How each alert played out 361 graded

Checked automatically from the saved prices: was the called price still up 5 minutes later, and an hour later (or at kickoff) was it still better than Pinnacle's price? Gone fast = a flicker or a race nobody wins. Pinnacle caught up = the edge was Pinnacle lagging, not your book. Held = still worth +0.5% or more. Your own taps show next to each; they catch what the feed can't (a price that was never really there).

HeldShrankPinnacle caught upGone fastToo soon
1722051118273

Pinnacle was behind (it moved to erase the edge within 15 minutes of the alert): 20 in the last 7 days. Circa already had the right price on 5 (the Circa check stops these now); both sharp books were slow on 4 (Under 47.5 (-105) at theScore Bet; Saints moneyline -148 at FanDuel; Oklahoma +5.5 (-108) at Bally Bet; Georgia Southern -9.5 (-112) at Caesars); 1 had no Circa price (before Circa was added: football Sat 6 PM, baseball Sun; or Circa didn't post the game); 10 were in a sport Circa isn't pulled for. Circa is your second sharp price (since Sep 26): when it has a newer price than Pinnacle that says the bet is worse than fair, the alert doesn't go out. You only get a phone warning if both sharp books were slow on 4+ alerts in a week.

Player props

Every prop flagged at +3% or more against Pinnacle, graded against Pinnacle's closing price for the same player, prop and line once the game starts. Props are alerted at a quarter unit until this shows they beat the close.

FlaggedGradedBeat the closeAverage vs the close
30000–

NFL opener watch

When one of your books opens next week's NFL line at least 2% better than Circa's price (Circa posts first), it's alerted at a quarter unit, labeled “Opener · unproven”. Each one is graded against Pinnacle's closing price once the game starts. If these beat the close over a few weeks, they get normal sizing.

FlaggedGradedBeat the closeAverage vs the close
20013173+0.6%

Quiet arb log

Every pair of prices at two of your books that would have locked in a profit (the same bet, opposite sides), logged from Sep 26 to Dec 31 to see whether arbitrage is worth it. Research only: no alerts. Arbs are also what gets accounts limited fastest.

Arbs found2771
Typical size (median)0.75%
Typical time they lasted (median)5 min
Locked-in profit if you'd put $1,000 into every one$40385.62

Book pairs: BetRivers + FanDuel (363), Bally Bet + FanDuel (362), Bally Bet + theScore Bet (322), BetRivers + theScore Bet (317), BetMGM + BetRivers (225), Bally Bet + BetMGM (214), BetRivers + Caesars (213), Bally Bet + Caesars (213), FanDuel + theScore Bet (117), Bally Bet + BetRivers (100), Caesars + FanDuel (81), BetMGM + FanDuel (71), Caesars + theScore Bet (71), BetMGM + theScore Bet (68), BetMGM + Caesars (34)

Which signals work?

Every signal a call rested on, graded on its own (a call counts under each of its signals). Keep an eye on the ones that reach 30 graded calls.

SignalCallsBeat the closeAvg vs close
Price only1071526 of 884+1.4%
Priced from Pinnacle's own line956477 of 755+1.6%
Thin market610281 of 457+1.5%
Pinnacle moved their way372207 of 319+1.9%
Outlier price275144 of 230+2.2%
Early in the week20198 of 173+1.1%
Exchange confirmed195107 of 176+1.8%
Reverse line move175111 of 158+2.6%
Fast-moving price17290 of 154+0.3%
Steam move15490 of 132+2.3%
Big money at DraftKings10561 of 91+2.4%
Big money at Circa8751 of 82+0.6%
College mismatch156 of 11+0.4% *
circa_confirmed95 of 8−0.8% *

* fewer than 30 graded calls: not enough to judge yet.

By sport

SportCallsBeat the closeAvg vs close
NFL12377 of 119+2.0%
College football597352 of 564+1.9%
NHL11064 of 99+2.1%
basketball_nba298127 of 244+0.4%
basketball_ncaab22265 of 101+1.3%
MLB6444 of 63+1.3%

Which book lags?

For each of your books: how long its called prices usually lasted, how often it pulled or locked them, and how often you found them gone. The book whose prices last longest is the one to check first when an alert comes in.

BookCallsUsually lastedPulled / lockedYou found it goneAvg vs close
Bally Bet3977 min2% of 39075% of 24+3.2%
BetRivers3638 min2% of 36121% of 24+1.3%
theScore Bet2679 min5% of 2620% of 9+0.2%
FanDuel2126 min4% of 21124% of 21+0.3%
Caesars1127 min7% of 10838% of 13+1.2%
BetMGM636 min5% of 6217% of 6+2.0%

See every call on the Track record. Scheduled deeper looks: the baseball replay on Fri Oct 2; on Mon Oct 5, the first full week of calibration and signals, plus a pre-set test of the signal weights (if big Pinnacle moves of 2+ points keep predicting the close on new games, their weight goes from 1 to 2).