Detroit Sports Reporter

• Tigers
• Lions
• Pistons
• Red Wings

2026-08-09 · Detroit Tigers · Analysis

The Tigers are 0-6 against Cleveland with the worst offensive matchup in baseball. I went looking for the reason and found out there probably isn't one

Cleveland comes to Comerica on Tuesday and the Tigers have not beaten them all year. Not once. Six games, six losses, and the losses were not the fun kind where you get blown out and forget about it by Thursday. Four of the six were decided by one or two runs.

This is the one thing about Detroit's season that the regression argument does not touch. I wrote last week that Detroit's close-game record is about 61 percent as meaningful as it looks, which covers a lot of ground, and a reader correctly pointed out that it does not cover this. You can tell me that 26-44 in close games will regress. Fine. Tell me about 0-6 against the one team sitting directly between Detroit and a wild card, with seven more games left against them.

So I went and looked at all six games, and then at every other team-opponent matchup in baseball, and the answer is genuinely not what I expected when I started.

First, how bad it actually is

Every completed Detroit-Cleveland game in 2026, matched by the league's own game id. The June 14 game was postponed and is on the September 4 schedule as a makeup, which is why it is not a seventh loss here.

Eleven runs in six games. That is 1.83 a game from a team that scores 4.58 for the season. Detroit struck out in 30.6 percent of its at-bats in these six games against a season rate of 25.6 percent, and left 92 men on base.

Cleveland scored 24, and 8 of those came in the first game. Take out the opener and it is Cleveland 16, Detroit 9 across five games. Nobody was hitting. Detroit was hitting less.

It is not the bullpen, and I want to be clear about that because it is the easy answer

The Tigers' bullpen is the story of this season. Twenty-two saves and 25 blown ones. It is the reason the record does not match the run differential, and if you had asked me to guess why Detroit is 0-6 against Cleveland before I pulled the game logs, that is what I would have guessed.

It is wrong.

Detroit's relievers threw 24 innings in this series with a 3.38 ERA, and four of those nine earned runs came in one third of an inning from Brant Hurter in the May 18 blowout. Take that one appearance out and the bullpen ran a 2.25 across the rest of it. That is not a unit that lost six games.

Here is the number that settles it: in these six games, Detroit led at the end of an inning from the sixth onward exactly once. You cannot blow a lead you never had. The bullpen was not handing these back. The offense never gave them anything to hand back.

What did happen is that Cleveland's relievers threw 17 innings and gave up one earned run. Cade Smith appeared in four of the six. Tanner Bibee started two of them and went 15 innings with three earned runs, and he is the probable starter on Tuesday.

So is this a real thing about the matchup?

Here is where I expected to write a paragraph about how Detroit's lineup has some specific problem with Cleveland's staff, and instead I have to write this.

Minus 2.75 runs per game is a huge number. Out of every team-opponent pair in baseball with at least six games played, 294 of them, Detroit against Cleveland is the single worst offensive matchup in the sport. Rank one of 294.

The problem with that sentence is that somebody has to be rank one of 294. The real question is not whether minus 2.75 is a big number. It is whether minus 2.75 is bigger than the worst number 294 coin flips produce.

So I tested it. Hold every team's actual game-by-game runs scored exactly as they happened, then shuffle which of those games belong to which opponent, keeping the number of games in each matchup the same. Recompute all 294 splits. Write down the worst one. Do that two thousand times, and you have a picture of what the most extreme matchup in baseball looks like when opponent identity means literally nothing.

All 294 team-opponent pairs with six or more completed games, 2026 regular season. Detroit-Cleveland is in the leftmost bin. Detroit-Athletics is in the rightmost.

The result:

• Detroit-Cleveland, observed: minus 2.75 runs per game.
• The worst pair in a shuffled season, median outcome: minus 2.94.
• Shuffles that produced a worst matchup at least as extreme as Detroit's: 1,473 of 2,000, or 73.7 percent.

The most alarming number in the Tigers' season is milder than what pure randomness typically hands you. In a league where nobody has any matchup problems with anybody, the worst-looking pair is usually worse than this one.

The part that convinced me

I could have stopped there and some of you would reasonably say the simulation is a trick. Here is the thing that is not a trick.

Detroit is rank 1 of 294 in the worst direction. Detroit is also rank 292 of 294 in the best direction, scoring 8.00 runs a game against the Athletics, plus 3.42 over their season rate, in a 6-0 sweep. And while doing it, Detroit held those same Athletics to 1.83 runs a game, which is the fifth most extreme offensive suppression in baseball.

Same team. Same season. Both tails of the distribution, plus a third entry near the end. If Detroit having the worst offensive matchup in baseball proves something is broken about how they match up with Cleveland, then Detroit having nearly the best proves they have solved the Athletics forever, and Tigers fans who watched this team for four months know exactly how much that is worth.

Two more things in the same direction. Seven team-opponent pairs in baseball are winless at six or more games, including the Mets at 0-7 against the Cubs and the Padres at 0-6 against Philadelphia. If you work out how many sweeps you would expect from team strength alone, using each club's season win rate, the answer is 5.0. Seven observed against five expected is not a phenomenon. And Cleveland is not some offense-eating machine: the second-worst matchup in baseball is Cleveland scoring 1.33 a game against Tampa Bay.

What I would still watch on Tuesday

None of the above means 0-6 was fake. Those are six real losses in the standings and Detroit is 57-60, two games back of a wild card, with Cleveland at 58-60 sitting a half game in front of them. The games happened. The argument is only about what they predict.

The one thing that survives every test above is that Tanner Bibee has thrown 15 innings against Detroit this year and given up three earned runs, and he takes the ball Tuesday at 6:40. That is not a statistical artifact of a small sample the way the team-level split is, it is one pitcher who has had a specific plan and executed it twice. It is also, and I have to say this because it is the same trap, fifteen innings. Two starts. If I told you a pitcher was now permanently good against a lineup on the strength of two starts I would be doing exactly the thing this piece just spent a thousand words arguing against.

So: I am not scared of the 0-6. I am mildly scared of Bibee, and mostly I want to see whether a Detroit lineup that struck out in 30.6 percent of its at-bats against this staff can hand its bullpen a lead late. Across six games it has done that once.

The pick for Tuesday's game goes up before first pitch, as always, on the record page. This piece is not it. This piece is me finding out that the scariest thing on Detroit's schedule is a coin that came up tails six times.

Method and sources: all game data from the MLB Stats API, regular season only, matched by gamePk and filtered on abstractGameState rather than the detailed status string, which reports real completed games as "Game Over" and "Completed Early." Series detail from scripts/det_cle_series.py, the distribution and permutation test from scripts/opponent_splits.py, the chart from scripts/opponent_split_chart.py, all in the repository. The permutation test uses a fixed seed so anyone can reproduce the 73.7 percent exactly.

Not betting advice. Just calls, made in public and kept in public.