Mike Matheny and pulling the starting pitcher too late

I wrote this before last night’s game. Everything holds true, perhaps moreso.

Generally speaking, when you look at starting pitcher stats by inning, they start off somewhat high in the first inning due to facing the opponent’s best hitters. They get much better in the second inning and then see a slow creep up until the sixth inning when they get better and in the seventh inning, the numbers are even better than the sixth. That has not been the case for the Cardinals this season.

It’s possible you are wondering why starting pitching numbers get better in the sixth and seventh inning as that’s when pitchers start to get tired. There is a survivorship bias in effect here. If a pitcher pitches into the sixth inning and into the seventh inning, presumably that pitcher is having a good game. If he were not a very good pitcher generally or not having a good game, he’s likely already out of the game, which is how you get better numbers in the sixth and seventh innings. The chart below shows the MLB numbers for this season.

Here are the numbers for the Cardinals starters this season:

Now we are dealing with a relatively small sample of around 50 innings or so for the first five with the sixth closer to 40 and the seventh around half of that. Even so, the Cardinals starters have gone against the general pattern of pitching better when they are allowed to pitch the sixth and seventh innings.

It’s just by innings either. By times through the order, there is a general worsening of pitcher stats as the game goes on, whether due to familiarity or just pitcher tiredness, the pitchers don’t get as good of results as earlier on. The Cardinals have a pretty massive jump compared to the rest of the league as the chart below shows.

So how much of this is a bullpen problem? Not much in terms of letting inherited runners score. Cardinals relievers have come in with 72 inherited runners, very close to the league average of 77, and they have allowed 27 runners to score, also close to the league average of 24. For what seems to be a pretty poor bullpen by results so far this season, the inherited runner situation is pretty close to league average and the roughly five more runs allowed than expected isn’t going to make a huge difference overall.

Now how much of the issue is a bullpen problem due to a lack of confidence to get outs? We could be on to something there, although Mike Matheny has long had a problem with leaving starters in too long even when he has had a good bullpen so that can’t completely explain things.

I decided to do a bit of a dive into the game situations this season to see if I could come up with anything. I looked at every single game and took a look at when a starter was pulled and how they performed in their final inning of work. If the pitcher came out for another inning, but did not complete it, I only included the partial inning.

Out of the Cardinals 55 games, the starting pitcher had a full, clean inning (no runs), 26 times. In the other 29 outings, either the pitcher gave up runs, didn’t finish the inning, or both. Now to some extent, this might be expected. A pitcher is more likely to get taken out once they encounter trouble. I wondered if looking at individual pitchers might help matters.

Let’s start with Adam Wainwright, the one-time ace who has the trust of the manager, and Carlos Martinez, the team’s new ace, who has the confidence of the manager.

Carlos Martinez

Clean, full final innings: 2

Extra outs without clean innings: 15

Total ER added: 13

ERA: 3.08

ERA without final inning, either full or partial: 1.61

ERA in final inning, either full or partial: 16.71

Adam Wainwright

Clean, full final innings: 5

Extra outs without clean innings: 9

Total ER added: 8

ERA: 3.79

ERA without final inning, either full or partial: 3.02

ERA in final inning, either full or partial: 9.00

It would certainly appear that Matheny lets these two guys go until they get in trouble and could be pushing them to get extra innings and extra outs. While this confidence is nice, it is likely hurting the players’ numbers. Carlos Martinez has given up more than hlaf his runs this season in the final inning he has worked. In six of Martinez’s 11 starts, he has not finished his final inning of work and four of those outings have come in the seventh inning or later.

One other pitcher has really been hurt this season by his final inning, and it reared its ugly head on Saturday as Leake was left in to pitch the seventh inning and pitched to seven batters, the sixth one a grand slam that turned the game. Here are Leake’s number for the year:

Mike Leake

Clean, full final innings: 5

Extra outs without clean innings: 15

Total ER added: 11

ERA: 2.64

ERA without final inning, either full or partial: 1.48

ERA in final inning, either full or partial: 12.38

Leake doesn’t have the same cache with Matheny that Martinez and Wainwright do, but Leake is incredibly deceiving when you look at a pitch count. Since the start of 2015, Leake has a 3.77 ERA and 3.73 FIP in the first five innings. In the sixth inning, the ERA moves up a bit to 3.94 with the FIP up to 4.62. In the seventh inning the ERA jumps way up to 5.10 with a 5.51 FIP. During that time, for the first and second times through the order, Leake has a 3.34 ERA and 3.59 FIP. The third time through the order, it is a 4.69 FIP and 5.16 ERA.

Leake is incredibly efficient so when you look at the standard pitch counts and assume 100 is a magic number in terms of getting tired, you have already missed the boat with Leake. For his career, on pitches 76-100, Leake’s OPS against is .800, much higher than his .736 career figure. Last year, it was .942 compared .756 and this year it has been .832 compared to the season total of .620. There is little reason to have confidence that Mike Leake will pitch well deep into games. He takes the ball every fifth day and can go five innings every time out, but a quicker hook is needed.

There hasn’t been any real issue with Lance Lynn and not much of one with Michael Wacha, who has generally been given a quick hook, although there have been a few exceptions. (I wrote up a quick narrative on every single game and put them all in one post here for those who would like to take a look).

The bullpen has been a problem this season, but there seems little point in constantly carrying eight relievers and running a short bench if those relievers aren’t going to be used. Using relievers to start innings instead of putting them in once the starter has gotten into trouble might be beneficial as well, allowing the reliever to ease into the situation.

The Cardinals are posting unusual numbers for starters as they get later into games this season. Generally, pitchers who get the sixth and seventh innings do well because they are having good games. The Cardinals haven’t seen the same bump because their pitchers are being left in too long, particularly when it comes to Adam Wainwright, Carlos Martinez, and Mike Leake.