This is derived from a Twitter conversation about Norris.
Norris was worth 2.9 rWAR last year. He pitched 144 innings and he pitched pretty well, especially when he was not asked to go deep into games. After the 75th batter of each outing, Norris allowed hitters a .977 OPS. To put it mildly, that is not sustainable. To give an idea how rough that is, the year Miggy won the Triple Crown, his OPS was .999 -- just a touch higher than Norris allowed these hitters.
The other problem that Norris has is his struggle to maintain his mechanics when he is ahead in the count. Have you ever noticed that he seemed to struggle when he got ahead of hitters? Can you picture him flying open and throwing a high, arm-side fastball on an 0-2 or 1-2 count?
Part 1 -- Norris Struggled to Put Hitters Away
When Norris got 2 strikes on a batter, and if it wasn't a full count, he pitched very carefully. This means that he was unhittable, but highly workable.
When Norris got ahead of a hitter 0-2, hitters had only a .426 OPS if the at-bat ended 0-2. If the hitter survived the 0-2 count, they had a .325 on-base-percentage going forward in the at-bat.
He got ahead of 140 hitters 0-2 -- 86 of them got to 1-2.
Norris faced 173 hitters 1-2. Again, if the at-bat ended 1-2, Norris did great -- hitters whose at-bat ended 1-2 had a .424 OPS. But if the hitter survived the 1-2 count (and got to 2-2), hitters had a .385 OBP and hit 5 homers.
Virtually none of this damage occurred at 2-2. If the at-bat ended 2-2, Norris gave up a .401 OPS -- this is below Josh Harrison territory here.
Unfortunately, hitters often got to a full count against Norris, and if they did, Norris was in trouble. 61 times a hitter escaped 2-2 and got to a full count, and that's where the trouble hit. With a full count (those 61 and 19 times a hitter started 3-1), Norris allowed hitters to hit .321/.525/.554. Yes, a .525 on-base percentage. Worse still, 80 hitters worked full counts on Norris, tied for the second-most common end-count.
Part 2 -- Norris really struggled after throwing Strike 1
Jim Price has told us all repeatedly that strike one is the most important pitch in baseball, and he's right, except for Daniel Norris. Norris' tendency to let batters get away from him was so strong that he actually had significantly worse results if he threw strike one (.801 OPS) than ball one (.718 OPS).
I wanted to see how rare Norris' struggles were when ahead in the count. Last year, 38 starters in the AL pitched 144 or more innings (Norris had 144.1). Of them, only 2 pitchers had worse results when the first pitch to the hitter was a strike -- Daniel Norris and Brett Anderson. Based on OPS allowed, Norris was .083 worse when he threw strike 1, Anderson was .049 worse. So Norris' struggles were pretty extreme.
Of the 38 pitchers, at least 5 were .300 OPS allowed better when they threw strike one -- among them were Tiger teammates Spencer Turnbull and Matt Boyd. This was not a Tiger issue -- it was a Daniel Norris issue.
Another 15 of the 38 pitchers were at least .200 OPS better when they threw strike one. As you might expect, most pitchers have a pretty big gap.
On the other hand, by OPS allowed after throwing strike 1, Daniel Norris was the worst pitcher in the league with an .807 OPS allowed, with only Yusei Kikuchi (.793), Rick Porcello (.792) and Ivan Nova (.780) close.
These numbers will probably bounce back this year, but if we are to see Daniel Norris reach the potential he shows after throwing ball 1, he should adopt the same approach after throwing strike 1.
Part 3 -- Norris' Problems Were Not Team-Wide Issues.
The 2 Tiger starters (Boyd and Turnbull) who threw more innings than Norris did not struggle like he did to put hitters away.
If Boyd threw strike one, hitters had a .606 OPS -- 200 points better than Norris. But Norris was almost 200 points better if the first pitch was a ball (.906 for Boyd, .718 for Norris). If you reached a full count against Boyd, you hit .213/.432/.413 -- less power and less average than Norris. If you ever got down 0-2 against Boyd, whether you hit that pitch or not, you were toast -- after 0-2, hitters had a .454 OPS against him.
If Turnbull tossed strike one against a hitter, the hitters had a .583 OPS -- even better than Boyd or Norris. But Turnbull he threw ball one, he was in trouble -- hitters had a .953 OPS. If you reached a full count against Turnbull, he had you where he wanted you. Hitters hit .220/.402/.256 against Turnbull with a full count (107 PA). Yeah, they got on base, but they had no power whatsoever. Like Boyd, Norris crushed hitters after 0-2 (.477 OPS).
Part 4 -- Conclusions (as to Norris)
The eyeball test suggested that Norris would lose command and control after throwing an 0-2 waste pitch. The numbers show that he not only struggled to put hitters away after he got ahead in the count, but also that the elevated pitch counts caused by these at-bats was also poison to him. The numbers suggest that when Norris gets ahead of a hitter he tries to throw unhittable pitches 0-2, 1-2 and 2-2 -- and if the hitter tries to hit them, he usually gets out. But 2019 Norris let hitters work him to 3-2 far too often, and then gets killed.
In my view, Norris is often an incredibly talented pitcher who is very difficult to hit. Any pitcher who allows only a .718 OPS after ball one is elite (Verlander allowed .728 last year and .743 career). But Norris' struggles after he gets ahead of hitters are unsustainable.
If Norris is to improve, it will start with his approach when he is ahead in the count -- and he has to be more aggressive and not as careful as he's been. What he may lose giving up some 0-2 hits will be rewarded by avoiding those full counts.
[Note -- the idea behind this post was inspired by a conversation with Robert James (@confusion_reign on Twitter)