On Wednesday afternoon, I briefly thought AJ Hinch had lost his mind. I really don’t know how to explain it. Runners are around the corner, in the opening game of the Wildcard Series between the Tigers and the Guardians, Cleveland won 6-1 – Detroit beats best batsman Riley Greene, its best batsman Riley Greene, with a chance to break a 1-1 lead. The guardian went to the bullpen and brought the left-handed Tim Herrin. Herrin, a 6-foot-6 curveball expert, believes Greene is a tough game. He has been fighting the same batsman throughout his career. But then, Hinchi made an amazing call to the bench. He pulled Greene with Jahmai Jones and pinched it – now I’m writing this here.
Jones has one key thing for him here: Like Inigo Montoya, he is not left-handed. He was also much better than the right in his short major league career and his minor league career. Green, on the other hand, is a poor batsman on the left and right sides. So you can at least see where Hinch’s decision comes from. I want to give this shocking decision to give it a full consideration before laughing it out – after all, what if it was the right call? So let’s do all the math to understand Hinch’s abandonment and what he gets.
To model the pitcher’s fight results, I first made predictions for these two players, who are granular in considering specific results. I also calculated the row split for each player by taking the observed career split and splitting their careers based on sample size and restoring it to the league average. I put these two projections (the batsman and the pitcher) into the modified Log5 formula and used it to predict the likelihood of each possible outcome of the board appearance. Then, when Green’s position appears in the lineup, I apply these results to the game state.
This is a lot of explanations in a paragraph, so I think an example is a preamble. Suppose Jones-fighting confrontation has 25% of the time, deep fly time is 25%, strikeout 25% of the time and walk 25% of the time. Of course, these things are almost unreasonable, but this is just an example. A single means the runner leads first and second (at least) and 2-1 with a 73.4% chance of winning. Deep sacrifice flies? That would bring the Tigers to 66.6%. Strike-up? 50.1%. Walk? 65.7%. Average of these four probabilities, the Tigers won the game with a 64% chance. Of course, there are more than four possible results, but this process is how I turn the result into the probability of victory.
Let’s start with the Greene-Herrin confrontation. Greene casts 100 WRC+ batsmen on the left and right as a strikeout issue, but it’s a lot of power. However, when they face left-handed pitching, there are few left-wing events with 100 wrc+ (this is the average overall league, all showdowns). That’s because Green is a very good baseball player. Herrin, on the other hand, predicts left-handed better than right-wingers. But he is slightly worse against left-handed batsman than the average left-handed pitcher, because his projection is a bit worse than the average left-handed pitcher in the first place. Putting them together, Greene predicted the following outcomes for Herrin (the possibility of Fly and off of the Off of the category RBI and Double Plays):
Riley Greene and Tim Herrin, the possibility of outcome
| result | possibility | Probability to win |
|---|---|---|
| Single | 11.7% | 75.4% |
| Double | 4.6% | 81.0% |
| three times | 0.2% | 85.7% |
| home run | 3.1% | 88.8% |
| Walking/HBP | 13.7% | 65.7% |
| Strike out | 32.3% | 50.1% |
| Fly out | 15.5% | 58.4% |
| ground | 18.9% | 51.5% |
To sum up all of this, while the Tigers are expected to win at least 59.4% of my math game, while Green is on the plate. That was the big lead, the fusion of small clues and tie, and enough multi-running clues to make up for the strikeout. In the slash term, it is .250/.332/.431, which is a very respectable behavior. It’s not bad for someone like Greene with strikeouts. But Herrin won’t hit a lot of batsmen and does go a ton, so you can see how this approach draws a conclusion.
Switching to Jones, by contrast, means attacking Herring with a weaker batsman, but the situation is more favorable. Jones has been slightly above average in his career overall, and he has done so well with left-handed people that he has used it as a true left-handed killer even if only a small percentage of Major League Baseball names. Herrin was not very shaking about the right. My same model thinks that Jones projected the absolute Wallop Herrin to the tune of the .299/.428/.472 hit line. Herring is a bit wild, Jones can take a walk, and these trends tend to increase when the batsman has a rank advantage. Here is the grid of the results of the Jones-herrin protein:
Jahmai Jones vs. Tim Herrin, Possibility of Results
| result | possibility | Probability to win |
|---|---|---|
| Single | 13.10% | 75.4% |
| Double | 5.10% | 81.0% |
| three times | 0.70% | 85.7% |
| home run | 1.9% | 88.8% |
| Walking/HBP | 22.0% | 65.7% |
| Strike out | 28.0% | 50.1% |
| Fly out | 16.0% | 58.4% |
| ground | 13.1% | 51.5% |
Adding all of that, the Tigers expect Jones to win 60.9% of the time when they hit the ball. This is a huge advantage over Greene’s numbers. This even seems a little low given the ridiculous slash line, but many of Jones’ positive scores are walking, which is the least impact of reaching the base here. But either way, check it out: Hinch’s decision is proven by mathematics. He inserted a pinch hit because that pinched nail was expected to have a better chance of winning than the guy who had hit before. To be sure, adding 1.5 percentage points to the winning probability in a pinch hit is a meaningful gain.
Well, if we deal with the appearance of this dish in isolation, it will be proven by mathematics. But this is an article by Ben Clemens, a post-season decision-making article by Fangraphs. We won’t leave anything here in isolation. The first consideration is: fines. Historically, the Knocker’s batsman in the starting lineup had about 24 points lower than the same batsman. Presenting it in our format will eliminate half the advantage the Tigers get. Not bad, but not very good either.
My next consideration: Can Tiger save Jones a better place? After all, he is an improvement here, but he is an improvement compared to the original highest starting point. Green is not a weak batsman either, and even in a bad showdown, when there are so many other lefties in the Detroit roster to replace, batting doesn’t excite me for him. Parker Meadows, Kerry Carpenter, Zach McKinstry: In a close match, the Guardian could have attacked those weak links with three other lefties in the bullpen. Of course, this is a high leverage position, but it is unlikely that the game will lack later (forecast level: 1,000). In my estimate, this is a sign against the use of Jones.
A bigger problem: In order, this position will often hit the ball again. By hitting the situation, or extending the situation with a knockout, this position in this command may appear again regardless of the result. Indeed, the Tigers had to replace Jones with defensiveness first and then pinch the replacement with Justyn-Henry Malloy. I might rather have Greene bats there, you know? When you pinch a good batsman, you will give up your future chances against that better batsman, so you’d better determine your strengths.
I do think it’s the highest plate appearance and might be in trouble with Jones. In the tie game, third place and less than two runners with knockout? That’s almost the same. But if Green didn’t open the game, why not wait for the batsman and pinch Wencel Pérez and stand on the deck? I’d rather use the best batsman than get the best platoon duel, and the Jones-Green pairing is significantly better than the Jones Perez duo in my estimate. Greene’s tendency to hit doubles may have played a role in Hinch’s mind, but Jones’s speed has almost continued, and it’s not particularly prone to GIDP before this year.
Hinch essentially said that he had different maths at the post-match press conference. “We know what the back end of the game is,” he said when referring to the Cleveland bullpen, “that’s why we are enterprising.” I put his point here. Essentially, he threw away most of his long-term math and just tried to play against the worst relief he could have seen in the game and play in a big boot position. After math, I can understand why Hinch made the decision. He has a very valuable pinch chip and he decided to play it when he had the first chance. Is it a frustrating thing to happen to be Green’s location? certainly. But he obviously would do the same with other lefties.
I think, at the end, I’m surprised to say, I don’t hate the decision to pin Green. I won’t do it. I’m not willing to throw away these long-term game numbers because even though the Guardian Bullpen is very good, I think it can be beaten and I want to do that with my best batsman. But the expected gains are surely there. There is no doubt that Hinch’s move increases the chances of tigers getting good results in the appearance of the board.
Of course they didn’t. Jones beat. It doesn’t matter. After chewing at Detroit’s Best Relief, the Guardian blows out the game with five runs in Game 8. In the end, nothing was left. But I still find this decision fascinating. It looks very wrong on the surface. It turned out that after I ran the numbers, it was a very good decision. Many other considerations related to the loss of chance of a future bat bring it back to the wrong scope, all of which are in the opposite team that washed away any evidence of any evidence with a lot of running. Playoff baseball is enjoyable.
