Pick No. 2: the luckiest pitcher in baseball starts Sunday, and I am taking him anyway
The call: Tigers beat the Giants Sunday afternoon in San Francisco. Confidence: low. And the guy I am counting on is having the luckiest season of any starting pitcher in baseball.
Both of those things are true at once. That is the whole entry.
Troy Melton's season, at face value
Twelve starts. 7-1. A 1.58 ERA over 74 innings, a 0.91 WHIP, and opposing hitters batting .170 against him. In July he threw 29 and a third innings and allowed three earned runs, a 0.92 ERA for the month. His last two starts, against Baltimore and Seattle, were seven innings of shutout ball each.
He is 25, he has sixteen major league starts in his life, and he has already been the best pitcher on a staff that until two weeks ago employed a two time Cy Young winner.
If you stop reading the numbers there, Sunday is easy.
The number that says stop reading there
Batting average on balls in play. Once a hitter puts a ball in play, and it is not a home run, how often does it fall in? A pitcher influences that. A pitcher does not control it. Over a full career it drifts toward the league rate for almost everyone, which is why an unusual BABIP is the first place to look when an ERA looks too good.
Melton's is .191. The median starter this season is at .287.
| Pitcher | BABIP | Rank | Innings | ERA |
|---|---|---|---|---|
| Troy Melton | .191 | 1 of 141 | 74.0 | 1.58 |
| Yoshinobu Yamamoto | .217 | 2 | 133.2 | 2.76 |
| Trey Yesavage | .218 | 3 | 93.2 | 3.65 |
| Jacob Misiorowski | .226 | 4 | 127.0 | 1.63 |
| Keider Montero | .233 | 5 | 122.1 | 3.38 |
| Eric Lauer | .236 | 6 | 92.0 | 4.89 |
| Logan Webb | .274 | 49 | 125.0 | 3.74 |
141 pitchers with 70 or more innings and 10 or more starts in 2026. Median .287.
First of 141, and it is not close. Second place is Yoshinobu Yamamoto at .217. The 26 points separating first from second are almost as much as the 31 points separating second from tenth, which is Tanner Bibee at .248. Melton is not leading a tight pack, he is standing off by himself. Ninety six points below the median is not a hot stretch. That is a pitcher whose contact is landing in gloves at a rate nobody sustains.
Two other things in his line point the same direction. He strikes out 7.78 per nine, which is fine and ordinary, not the strikeout rate of a man with a 1.58 ERA. And he has already given up eight home runs in 74 innings, which is not the profile of somebody suppressing hard contact. Great pitchers with tiny ERAs usually miss more bats than that.
So: the 1.58 is not what Melton is. Somewhere between that and league average is what Melton is, and the correction can arrive on any given Sunday, including this one.
Why the call is still Detroit
Because the other guy is worse right now, and by more than people expect from the name.
Logan Webb is a three time All Star and the closest thing San Francisco has to an ace. His season line is a perfectly respectable 3.74 over 125 innings. His recent line is not.
| Webb, last six starts | IP | ER | Note |
|---|---|---|---|
| Jul 3, Colorado | 3.0 | 7 | eleven hits |
| Jul 8, Toronto | 7.0 | 5 | |
| Jul 18, Seattle | 6.2 | 3 | |
| Jul 24, LA Angels | 6.0 | 4 | |
| Jul 29, Milwaukee | 6.0 | 2 | |
| Aug 3, Texas | 6.0 | 0 | eight strikeouts |
| Total | 34.2 | 21 | 5.45 ERA |
Across those six, 5.45. Across the whole of July, five starts, 6.59.
There are reasons behind it and they are not encouraging for San Francisco. He went on the injured list May 9 with right knee bursitis, his first real absence since 2021. His strikeout rate has fallen to 20.2 percent from a 9.74 per nine in 2025, and he is allowing a career worst 48.6 percent hard hit rate. Patrick Bailey, who caught him and was regarded as the best pitch framer in baseball, was traded to Cleveland and is a Guardian now. And Webb's entire method is precision on the edges of the strike zone, which is exactly the thing an automated zone takes away from a pitcher who used to earn those calls.
The August 3 start against Texas, six shutout innings with eight strikeouts, was his best in two months. So this is a slump with a possible end date rather than a collapse. It is still a slump.
Now the teams. Detroit has been outplayed by its own record all year and it is still true: 528 runs scored, 451 allowed, which projects to about 66 wins against an actual 56-60. San Francisco is 49-67 and has been outscored by
- Detroit's offense sits tenth in baseball by OPS, the Giants fourteenth. Take the names off and this is the better team, on the road, against a home club that is 27-27 in its own park.
What scares me
The same thing that scared me Saturday. 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. That is the mechanism behind the ten win gap and it has not been fixed by writing about it.
Melton has averaged just over six innings a start, which is more than most, but it still means three innings for the group that has lost this team a season.
And there is the obvious one: I have spent half this piece explaining that my starter's ERA is a mirage. If the regression lands Sunday, I will have called it in print and lost the game anyway. That is the honest shape of this pick.
Detroit also lost Friday night at Oracle, 5-2. This is not a team walking in hot.
Why low, again
Two picks, two lows, and I want to be straight about that instead of letting it look like a habit.
Here is the rule I am using. High means the edge is unusual enough that missing should embarrass me. A better team beating a worse team in one baseball game is not unusual. It is a coin flip with a thumb on it. Baseball is the sport where the best team in the league loses sixty times a year, and a label that reads High on a garden variety good spot is a label that means nothing by October.
The Highs are coming. They will show up when a number is genuinely out of line with what everyone else is looking at, and I will say so in those words. This is not that. This is the better team, a starter I trust less than his ERA, and a bullpen I trust less than either.
Tigers win. Low confidence. Committed before first pitch, timestamped, no takebacks.
Sources: MLB Stats API, pulled 2026-08-08, for all records, run totals, save and blown save counts, pitching lines and game logs. BABIP figures are the league's own advanced pitching splits, filtered to 141 pitchers with 70 or more innings and 10 or more starts; median computed across that population. Webb's six start and July ERAs were computed from his game log rather than quoted from anywhere. His knee bursitis stint, strikeout and hard hit rates, and the automated zone argument are per Yahoo Sports; Patrick Bailey's move to Cleveland was confirmed against the league's own roster data. Chart generated from live data by scripts/babip_chart.py in the repository.
Not betting advice. Just calls, made in public and kept in public.