Tuesday's 8-3 defeat in Washington was possibly lost in the bottom of the 5th inning. The Nationals' 3-4-5 hitters were due up for their third time, Lance Lynn was sitting at 80 pitches, and the Cardinals were dealing with a 4-2 deficit. A leadoff six-pitch walk to Bryce Harper was immediately followed by a two-run shot to right field from Daniel Murphy. As it would turn out, the Nationals entered that inning with enough runs to win the game, but by the time the 5th came to a close, their win probability jumped from 76 percent to 92.
Should Lynn have pitched that entire inning for the Cardinals? (Note: I was out of pocket for the game, I didn’t watch or listen to it, so I have no idea if this was an issue that was even debated, it was just something that caught my eye when scanning the box score.) In hindsight, no. Obviously. Harper and Murphy are both lefties and Lynn has pretty profound career splits. Left-handed hitters have a .775 career OPS versus Lynn but that number plummets to .629 for righties. And, as mentioned, the Nationals were already batting through the order for the third time and surely a casual glance at Lynn's line against lefties their third time through the order will resemble something horrible, or at least considerably worse than his career .775 OPS allowed versus lefties, right?
Well, actually no. When Lynn is facing the order for the third time, lefties only have a career .678 OPS against him. What's pretty ugly are the run prevention numbers though: 43 earned runs allowed in these situations, aided by the fact that his walk rate jumps to 13.6 percent and his strikeout rate falls to 14.3 percent. In contrast, Lynn has career a 22.8 and 8.7 percent strikeout and walk rates, respectively, and against lefties, no matter what time through the order, the numbers are 18.2 and 12.3 percent, which is bad but not quite as unsightly as above.
Splits aside, take a look at Lynn's career numbers versus lineups their third time through the order, or, more specifically, look at Lynn's last two full seasons and what becomes clear is that Lynn, naturally, is usually at his strongest right out of the gate, but he doesn't fall off a cliff his third time through as is typical of the average starter.
From 2014 and 2015, here are some of Lynn's stats his first, second, and third time through the order as compared to the rest of the NL:
Because of the way the final graph is formatted, Lynn’s divergence from the rest of the field his third time through the order looks a bit more extreme than it actually is but the theme remains pretty much the same: Unlike the league on average, Lynn was typically more effective as the game went on. The same holds true when looking at fielding independent pitching:
And, as noted above, even though Lynn’s K-BB rate has suffered versus lefties as games wear on, take a look at the overall picture from the last two seasons:
Pretty interesting, right? As expected, the NL average gets worse each time, but Lynn consistently breaks from the pack the third time through. This is not a case of a large sample divergence for each time through the order, or pitchers only seeing the third time through the order when they’re pitching well. Combining 2014 and 2015, Lynn only pitched about 30 more innings the first time through the order versus the third. Across the board, starting pitchers typically last three times through the order more often than not. And, as illustrated above, most of the NL was not in line with Lynn. He excels here.
Where Lynn does not excel is against lefties, and that’s basically true no matter what inning it is and how many pitches he’s thrown. In his first outing this season, a similar situation to this past Tuesday arose when Mike Matheny let Lynn start the 6th inning with 86 pitches under his belt and left-handed hitters Anthony Rizzo and Ben Zobrist due up. Rizzo doubled and moved to third on Zobrist’s subsequent groundout. Lynn was then removed from the game and Rizzo would soon score an important run.
To circle back to the original question, should Lynn have been allowed to pitch to Daniel Murphy in the 5th inning on Tuesday? If forced to answer that question in real time instead of with hindsight I’d have to say, well, I don’t know, but probably? Given his recent history of pitching well late in games it was very defensible (the Rizzo/Zobrist situation might be different), especially since his pitch count was still pretty low and five innings is about the expected minimum for a starting pitcher this early in the season.
2014 and 2015 were a long time and one Tommy John surgery ago. We don’t know what we have yet with Lance Lynn in 2017. Whether he can continue to break from the norm and pitch well his third time through the order will be something to keep an eye on. And in the meantime, maybe VEB’s own Joe Schwarz can come up with a good formula to help him become a bit more effective against lefties, too.