We found a pricing error in our own screen this morning, and it changed the card.
our data platform's board shows the best available price across every book it tracks — and on a large
number of rows that best price is a DFS operator, which we do not use and cannot access. a DFS operator
prices its legs at a flat house number that implies roughly 58%, so every row where a DFS operator was the
best book was showing us an implied probability around 58% no matter what the real sportsbooks were
charging. On Gabriel Moreno Over 0.5 Hits the actual best book price was −205
— a 67.2% implied probability, not 58%. That is a nine-point error in the same direction on
every affected row, and it was inflating Edge Gap on roughly two-thirds of the board.
Every price on this card is the best usable sportsbook price, read off the expanded book list,
with a DFS operator excluded. Where a hidden book might have a better number than the ones our data platform renders,
our figure is the conservative one — which understates Edge Gap rather than flattering it.
What the board actually looked like once it was priced properly
Fifteen games, and the screen returned 159 propositions at the base filter (odds from
−160 up, one hit-rate window at 80% or better). Re-priced against real
books, the ranking changed materially. Two of the four legs we would have published on the uncorrected
numbers fell out of the top ten. Six legs cleared. We are publishing six.
Five of the six sit in different games. The two that share a game — Bassitt and Ramos, both at
Camden Yards — are worth naming as a pair, because they are not the concentration risk they look
like. Ramos bats against Bassitt. If Bassitt is sharp, he collects strikeouts (bad for one leg) and
Ramos stays quiet (good for the other). If Bassitt gets hit, the reverse. They are negatively
correlated, which is closer to a hedge than a double-down. We would rather say that out loud than
have someone notice it and assume we did not.
The matchup layer removed more than it approved
Only one of the six is a Confirms. That is not a stylistic choice; it is what the slate gave us.
Here is what the gate threw out, in order of the edge we walked away from.
Ben Rortvedt Under 0.5 Runs — Edge Gap +31.1, dropped. Blended 92.0
against a −156 price, which would have been our second-biggest edge of
the morning. He is batting ninth for the Dodgers at Coors Field, in the only game on the board
with a total of 11.5. We have dropped Coors Unders three days running and we are not going to start
making exceptions for the ones with the biggest numbers.
Moisés Ballesteros Under 1.5 H+R+RBI — Edge Gap +23.5, dropped. He is a
left-handed bat and Houston starts Ethan Pecko, who has an 0.00 ERA because he has not thrown an
inning. A hitter with the platoon advantage against a starter with no major-league sample is not a
matchup we can grade in either direction. There is no number here, so there is no bet.
Mickey Moniak Under 1.5 Bases (+21.4) and Teoscar Hernández Under 0.5 RBIs
(+21.3) — dropped, same reason as Rortvedt. Coors, 11.5, and both hitters occupy lineup slots
that see four plate appearances.
Spencer Horwitz Under 0.5 Runs — Edge Gap +19.9, dropped. He leads off
for Pittsburgh against Jackson Jobe and a 6.23 ERA. The most plate appearances in the lineup
against the worst starter on the board is the wrong side of an Under-runs prop no matter what the
last ten games say. His head-to-head number is 33%, which agrees.
Nick Loftin Under 1.5 H+R+RBI (+19.5) and Sal Stewart Under 0.5 Runs (+16.8)
— dropped. Both are right-handed bats holding the platoon advantage against left-handers with
ERAs of 6.17 and 5.07. Same shape of mistake, twice.
The Coors drops sit awkwardly against something else we found today, and we are
not going to hide it. We dropped three Coors Unders on park grounds. Meanwhile 94% of the money
in that game is on the game Under 11.5 — the market's own read is that even 11.5 is too high a
number for tonight. Those two positions are in tension. Our park rule is crude: it treats "Coors" as a
blanket disqualifier for any Under rather than asking whether the specific total already prices the
park. We are leaving the rule in place for today because changing a filter mid-morning to admit picks
we already rejected is how methods rot — but this is now on the list to test properly, and the
game-lines post carries the Under 11.5 as a published lean.
The argument against our own number one
Chris Bassitt Under 4.5 Strikeouts is our largest edge and our most contestable
pick. Here is the case against it, which is stronger than we would like.
The Yankees just struck out fourteen times. Last night at Camden Yards, against Shane Baz and two
relievers, New York struck out 14 times in nine innings. Tonight they run out a lineup with
seven left-handed bats against a right-hander. A strikeout-prone lineup is precisely the input
that should move a pitcher-strikeout Under, and it moves it against us. That is why we have this leg
as Neutral and not Confirms, despite Bassitt's own profile being about as supportive as a profile gets.
The line has been moving both ways. our data platform's line history shows this number opening at 3.5 at
11:07 PM last night and then bouncing 3.5 → 4.5 → 3.5 → 4.5 in twenty-one minutes. A number
that unstable overnight is a number the market has not settled on, and a settled market is exactly what
Edge Gap assumes when it treats the price as a probability.
The price is telling us something. −100 on a proposition that has
hit 91% of the time this season is not a book being lazy. Bassitt's season strikeout rate is
5.84 per nine over 61.2 innings — about twelve starts, which is a real but not enormous sample
— and his 10.95 hits per nine and 1.62 WHIP mean he is on the mound with traffic constantly.
Our read is that traffic shortens his outing before it lengthens his strikeout total; the market's read
is evidently closer to a coin flip. One of us is wrong and it is not obvious which.
What we are actually leaning on: his last ten starts show an average of 2.7 strikeouts and a median
of 2.5, with a single our data platform of seven. Nine of ten cleared. That is the whole case, and nine of ten is
nine of ten — not a probability.
The rest of the card, briefly
Heliot Ramos Under 1.5 H+R+RBI at +110. We
published this same leg yesterday at −104 and it hit — he went
0-for-3 with a walk. That is the uncomfortable part: the 100% last-ten figure driving today's blended
rate is partly a result we produced ourselves, and his season number is 59%. The plus-money price is
what makes it the second-biggest edge on the board; the gap between a 59% season and a 100% last ten is
what should keep anyone honest about it. Right-handed bat, right-handed starter, batting sixth.
Jake Bauers Over 0.5 Bases — the only Confirms on the card. Left-handed
bat batting third for Milwaukee against right-hander Logan Gilbert. Platoon advantage plus a top-of-order
slot. The caveat is that Gilbert is the best pitcher any of our hitters face tonight at 3.28, and that
Bauers homered in last night's 22–0 game — a night nobody should be extrapolating from.
Justin Crawford Over 0.5 Bases and Andrés Giménez Over 0.5 Bases.
Both left-handed bats with the platoon edge, both facing genuinely good right-handers — Sandy
Alcantara at 3.43 and Drew Rasmussen at 2.78. Both bat low in the order. We have them Neutral because
the platoon advantage and the pitcher quality cancel, and we are publishing them on the Edge Gap alone.
Jacob Young Over 0.5 H+R+RBI. He missed this yesterday in the same series
— 0-for-3 in a game Washington lost 5–0 without scoring. He bats ninth again. What is
different is the arm: Cal Quantrill threw seven shutout innings yesterday; today it is Kumar Rocker at
4.50. Publishing a leg that just lost, one day later, is either discipline or stubbornness and we will
know in about ten hours.
What we did not publish
Pete Alonso Over 0.5 Hits re-priced to a +10.2 edge once a DFS operator came out of the
calculation, which put it outside the card. Bryson Stott Over 0.5 Bases (+15.3) and
Gabriel Moreno Over 0.5 Hits (+15.6) both cleared the gate on merit and were cut for
diversification and price respectively — Stott because we already had a Philadelphia leg with a
larger edge, Moreno because −205 is a lot to lay into Payton Tolle, who
is carrying a 1.05 WHIP and 10.02 strikeouts per nine. One note on Stott worth keeping: his
Over 0.5 Bases and Over 0.5 Hits are the same event — you cannot record a total base
without a hit — and the Bases version was priced at −200 against
−212 for Hits. Identical outcome, twelve cents apart.
Prices are the best non-a DFS operator
sportsbook price on our premium data platform, captured 8:05–8:25 AM ET. Hit-rate windows and line-movement
history via our data platform. Confirmed
lineups, starting pitchers, handedness and weather via
our lineup service daily lineups.
Yesterday's box scores via CBS Sports.
Implied probability and Edge Gap are computed from the odds by our build, never typed by hand.