Graded: Pick No. 2, Tigers at Giants. Correct in ten innings, and wrong about both pitchers
The call, committed before first pitch: Tigers win. Confidence Low.
The result: Detroit 3, San Francisco 1, ten innings. gamePk 823190, confirmed Final against the MLB Stats API.
Record: 2-0.
Now the part worth reading, which is that almost none of the reasoning survived the game.
Melton did not regress
Half of Sunday's entry was an argument that Troy Melton's 1.58 ERA was built on a .191 batting average on balls in play, the lowest of 141 qualifying starters in baseball, and that the correction could land on any given Sunday including that one.
It did not land. Six innings, four hits, one walk, five strikeouts, no earned runs, 92 pitches. His ERA went down, to 1.46 over 80 innings.
The one honest thing to pull out of the line: balls put in play against him Sunday fell in at .250, sixteen balls in play and four hits. That is well above his season rate and still under the .287 league median for a starter. So the day was a small step toward the thing I predicted and nowhere near the correction I described. Sixteen balls in play settles nothing, which is exactly what the original piece said about small samples and is worth applying to the sample that flattered me too.
Webb was excellent
The other leg of the pick was that Logan Webb had a 5.45 ERA across his last six starts and a 6.59 in July, and that this was the version of him Detroit would meet.
He threw eight innings and gave up one run on four hits and a walk, with twelve ground outs. Whatever the last six starts said about him did not show up. San Francisco lost that game anyway, which is its own kind of brutal.
So the pick was right on the teams and wrong on both starters. Detroit is a better club than San Francisco by run differential, by offense, by roughly every measure the entry cited, and that is the leg that carried it. The pitching matchup, which is what the piece actually spent its words on, went to the Giants and did not matter.
The bullpen did exactly what I said it would
This is the part I want on the record, because it is the first time the stated fear actually arrived.
From Sunday's entry: "Detroit has 22 saves and 25 blown saves in 47 opportunities. Forty seven percent. The team ERA is 3.55, which is good, and the bullpen keeps handing back leads anyway."
Melton left in the seventh with a 1-0 lead, a runner on first and nobody out. Kyle Finnegan came in. A force out, then a Drew Gilbert single, and the game was tied. That is Detroit's 26th blown save in 49 chances, 47 percent, right on the season rate. Eight of the 26 belong to Finnegan.
What is different is what happened next. Drew Sommers, Kenley Jansen and Tyler Holton threw three and a third innings without allowing a base runner between them, striking out seven. Then Hao-Yu Lee singled in the tenth to score Riley Greene, and Max Clark forced in a second run, and Holton struck out the side in the bottom half.
Detroit blew the lead and won anyway. That is not the shape of the ten-win gap; it is the opposite of it.
One number that moved
Detroit is now 27-44 in games decided by three runs or fewer, and 31-16 in games decided by four or more. Sunday was a two-run game, so it goes in the bucket this team has spent all season losing.
The close-game piece from earlier this week argued that Detroit's .371 in those games regresses toward about .442 over the rest of the schedule, worth roughly two wins. One game proves nothing about that. It is one game in the right direction, and there are 44 left to see whether the rest follow.
The Tigers sit 58-60, three and a half back of the White Sox in the Central and one game out of the last wild card with Cleveland arriving at Comerica Tuesday.
Sources: MLB Stats API, pulled 2026-08-09, for the final, the box score, the play by play, both pitching lines, team save and blown save totals, the close game splits computed from Detroit's own game log, and the standings.
Not betting advice. Just calls, made in public and kept in public.
Got a Detroit number you want looked at? Email projectunmuted@proton.me and it gets looked at properly. A stat somebody quoted that smells wrong, a thing you have always assumed about one of these 4 teams and have never seen checked. Every question that arrives is listed with its answer, including the ones where the answer is no.