Article: Adbert Alzolay and (Hopefully, Eventually) a Better Way to Measure Relievers' Rest

[Matthew Trueblood]

We talk too much about relievers' talent and skill sets, which we have a hard time pinning down, anyway. We talk too little, by contrast, about relievers' usage, and the fact that their rest (or lack thereof) drives much of the performance we see.

Image courtesy of © Joe Nicholson-USA TODAY Sports

 

Although they had a winning road trip, the Cubs had their hearts broken twice in nine games. Adbert Alzolay was called upon to hold thin, late leads on Apr. 9 in San Diego, and again this Monday in Arizona, and each time, he gave up a game-changing home run. Alzolay is off to a very uneven start in 2024, and many fans are already wondering whether he can truly be trusted as the team's closer. Those concerns aren't unfounded, but they're not framed quite right.

There's no lack of fortitude, intensity, or talent with Alzolay. He doesn't throw as hard as most of the game's elite closers, but his fastball sits 95 and noses above 97 frequently, and his slider and cutter can be devastating when he's right. He had a 21.4 K%-BB% last year, which is plenty impressive, especially for a guy who has two fastball looks and can (generally) avoid giving up the long ball.

It's only a cruel taunt to say he 'should' be able to avoid that, though, because Alzolay has had much worse trouble with homers than you want from a closer. In addition to the telling blows each of the last two Mondays, he gave up the team's one-run lead by letting Travis Jankowski take him deep on Opening Day. Last year, he only gave up five homers all season, but three of them came after the All-Star break--a stretch during which, thanks to a forearm strain that sidelined him for much of September, he only made 25 appearances. In there, too, was the would-be home run robbed by Mike Tauchman in St. Louis in late July.

Crucially, at his best, Alzolay really can stop opponents from hitting the ball over the fence, and he's a strike-throwing machine with sufficient bat-missing ability to be a dominant reliever. The question is how often he can be at his best, and to answer that question, we have to go a bit further than simple splits.

Look up a player's splits on Baseball-Reference, and you can find out how he's pitched on various numbers of days of rest. That's about as far as easily queried public data takes us, though, and it's not nearly far enough. It doesn't just matter whether a pitcher worked the previous day, or two days ago, or three. It matters, at the very least, whether he's appeared on two of the last three days. It probably matters whether he's appeared on three of the last five, or four of the last seven. It also matters how many pitches a hurler threw in those games. Counting the days between throws off a game mound is grossly insufficient to tell us whether a reliever is fresh and operating at full capacity on a given day.

I took one small step toward changing that. For each of Alzolay's 67 appearances since the start of 2023, I calculated a Comprehensive Rest Adjustment. I kept this very simple; I'm quite sure we'll work our way to something more robust in time. For now, though, the formula goes: (Pitches thrown over previous three days) * (Number of appearances over previous three days). Then, I added 15 to the score if Alzolay had pitched the previous day, and was therefore working back-to-back.

This leads to plenty of outings where the CRA is zero; I think that's fine. If a pitcher hasn't worked in the previous three days (assuming it was anything short of a traditional start), he should be ready and fully rested for that day's game. Pitching in multiple contests in a three-day span is taxing, by contrast, and that should be reflected in any effort to capture the effects of reliever workload on reliever performance. We know, too, that working on back-to-back days is especially impactful.

Once I had a CRA for each game, I divided them into three categories:

• Level I: Any appearance with a CRA of 15 or lower
• Level II: Any appearance with a CRA between 16 and 30
• Level III: Any appearance with a CRA of 31 or higher

The idea here was just to make sure that any game pitched on a second consecutive day would be distinct from one with multiple days of rest, but that it wouldn't automatically vault to Level III unless the previous day's appearance was a long one, or Alzolay had also pitched another contest within the three-day window. Unintentionally, though, this also distributed his outings about the way I'd have wanted them to be, if I'd been trying to create an artificially smooth breakdown. He had exactly 26 Level I and 26 Level II appearances, and 15 Level III ones.

I didn't even check his ERAs or WHIPs for those games. Those numbers are too noisy for my liking, in samples of this size, and it's not results I wanted to know about, really. Instead, I pulled Alzolay's average fastball velocity (sinkers and four-seamers), his slider usage, the percentage of his pitches thrown inside the strike zone, opposing batters' chase rate, and their whiff rate on swings. In a nod toward results, though, I did also pull his strikeout and walk numbers.

Here's what I found.

Category	FB Vel.	Slider %	Zone %	Chase %	Whiff %	K %	BB %
Level I	95.4	45.7	50.9	31.8	29.5	25.7	7.1
Level II	95	41.8	49.7	32.5	26.3	28	4.2
Level III	95.6	55.9	56.6	29.4	26.9	23.5	3.9

I had various and (in some cases) contradictory hypotheses for this exploration, so the results didn't shock me, but there are surprises here. Some relievers get very wild when they've recently been overworked, but Alzolay actually throws more strikes. That might not shock you, and I bet the drastically higher slider usage doesn't, either. It was the slider on which Alzolay got burned against Ketel Marte Monday, and the slider on which Nick Martini took him deep in a crushing blown save last Sept. 1. Both of those games were Level III appearances.

That problem isn't exclusive to Level III days, of course. It was a two-strike slider with too much of the plate on which Jankowski erased a Cubs lead on Opening Day, and obviously, that was a Level I game. When Fernando Tatis Jr. pummeled a slider to give the Padres the lead last week, it was in a Level II appearance, although one just shy of the barrier between that and Level III. Still, Level III comes with some special dangers, as we would guess. They just aren't the particular dangers we might guess.

Alzolay doesn't lose any juice on the fastball when working on Level III days. On the contrary, he throws even harder. Perhaps trying to protect his tired arm in a different way, though, he gets so slider-heavy that hitters start looking for it and don't chase or whiff as often as they otherwise would. Anecdotally, it has also often seemed like he was trying too hard to economize on Level III days, refusing to stretch or nibble at the zone out of fear of a long outing. More sliders in the zone necessarily means a greater risk of home runs and other hard contact, and naturally, it also means fewer strikeouts.

That version of Alzolay is still an effective pitcher. He's just a different one than when he's more rested. The Cubs could still stand to find a closer who misses a few more bats or is a bit less susceptible to the long ball, but Alzolay is a superb relief pitcher. What the team needs, really, is more protection for him--protection from heavy usage, and from situations in which his tendencies on days when he's recently been heavily used are most damaging. That's why Julian Merryweather was such a painful loss for the bullpen.

This analysis doesn't capture the rising injury risk from one level to the next. It doesn't capture a lot of nuances of usage that teams build into their decisions about when to have a pitcher down and when to limit them to a shorter outing. It's just meant to illuminate Alzolay's problems a bit, by reminding us all that the difference between one reliever and another can often be smaller than that between one individual at full freshness and the same guy running on fumes.

Alzolay and his catchers might need to adjust their approach on Level III days. That's a tough thing to do, though, because again, some of it might be about shielding himself from injury, and you don't want him careening from undue carefulness to recklessness. Nor is it easy to remember and execute a change in plans when tired, and a high-leverage reliever working for a third day in four or on the second of two consecutive days is mentally tired, as well as physically so.

The most important takeaway here might be that Alzolay is clearly at his best in Level II outings. He gets fewer whiffs in those than in Level I appearances, but still racks up more strikeouts. He pounds the zone, but not to such an extent as to expose himself to too much risk of hard contact. I haven't yet run the numbers for others, but I suspect most relievers are best when their CRA is in a Level II range. Teams really might want to consider a relief rotation. In the meantime, the Cubs should avoid letting Alzolay get rusty, even (or especially) if it sometimes means letting others collect saves in games when Alzolay would otherwise be at Level III. He's not someone who wears down badly in such contests, but his margin for error thins out on him.

 

View full article

[Guest234]

Are you looking for a retainer from Tommy Hottovy?

[Transmogrified Tiger]

I wonder if the pitch mix is influenced by repeat outings against the same team too.  Probably next to impossible to control for in Alzolay's sample size, but the thesis might be that he gets more slider heavy when he sees hitters multiple times in a series since his fastball isn't the wipeout pitch that other elite relievers possess.

[Guest234]

How does yesterday's data affect this analysis?