Analysis in Action

The Nats are going to be hard pressed to "win" this Scherzer deal. Such is how it goes with long term deals. You don't hope to have a guy perform up to his salary each year. Instead you hope to get enough value on the front end so it balances the value you lose on the back end (and you secretly hope real world inflation/baseball salary inflation makes your deal look better).  What we are looking for from Scherzer is really 3-4 very good to great years, plus 2 or so middling ones, followed by one or two bad ones. That would be the most reasonable scenario. 

But what can we really expect from Scherzer? That's hard to say. You don't have access to an infinite number of dimensions to test out Scherzer theories and while I do, I have better things to do with that power. Instead we'll have to look at comparables to Scherzer and kind of guess at how he'll do based on how they did. Which brings us to pitches.

One of the notes I heard about the process that brought Scherzer to the Nats had Boras telling Lerner that despite the high number of innings pitched, Scherzer hasn't worked that hard. He used the number of batters faced to show he really wasn't Top 5 in burden, more like Top 20. Boras of course is spinning a half-truth to favor his client. It's his job. He's right it's not innings that we should be looking at, but its not batters faced either. Both are proxies for what we are really interested in, the number of pitches thrown. That tells you how much work a pitcher has really had and with Scherzer, that number is pretty high.

I started the analysis there, Scherzer is a good pitcher that throws a lot of pitches. What do those people look like in their early to mid 30s? At first I looked at pitchers who threw over 10000 pitches over any 3 season period. The good thing is that this also tends to subset down to good pitchers. You don't get to throw the number of innings necessary to throw that many pitches if you are doing poorly.  What did I find? Not much good, but enough noise that I didn't feel good saying any one thing. So I narrowed the analysis. What I was really interested were those pitchers that threw that many pitches in the periods ending in their 28th and 29th years, like Scherzer has. What did I find now?

For those looking at it the legend is as follows
Black : DNP
Stripes : Started fewer than 28 games
Red : Had an ERA+ of under 100.
"X" or "YYYY" : Ended a 3 yr period with 10000 pitches or more
Heavy Black Borders / Gray : 2014 / Years that haven't happened yet

Anyway what we see is a mess. Almost none of these guys produced a fully useful year (100+ ERA+ with at least 28 starts) after the age of 31. Very few even produced one good pitching year of any number of starts after 31. A few were out of baseball within a couple years. If there is any good news to glean from this chart it's that most of these guys had more use than Scherzer prior to age 28 and the most comparable guy usage wise, Roy Oswalt, while done at 33, would come closest to giving the Nats what they want from Scherzer. Three good years and one good but injured one before being claimed by the baseball gods.

(Why did I stop at 35? Because when I started looking at this I thought anything that happens after 35 really can't be linked to arm abuse. Of course now I'm like should have just went to 36 because that's where Scherzer's contract ends but whatever. Too late now)

There are two big problems though thinking about the analysis. One is of course small sample size. Nothing I can do about that. The second it doesn't help answer the question - Scherzer over Zimmermann. Zimmermann is an extremely efficient pitcher who is not throwing over 10000 pitches. Does that make him a better bet to pitch well into his mid 30s? Or is it just that all pitchers crash and burn and you are rolling the same dice with any contract?

So I changed the conditions a bit for easier subsetting, using 13333 pitches from age 26-29 and an ERA+ of at least 100 during that time (Scherzer is only 117 mind you). That loses Lincecum and Livan from the above and adds Aaron Harang, Jon Lester, James Shields and Javy Vazquez. OH! Another thing to note - these data are only from 2000 on. That's what BRef has for pitches - How does that change the interpretation of the pitching group? It makes it slightly more favorable to have a decent age 32 year, but after that you are pretty much dead.

How did the other group do, the under 13333 pitches from 26-29 and and ERA+ over 100 (112 starts as a guideline - the ones where the color coding starts at 30 never had any 10000 3 year periods)

What we found was mostly the same thing. Ages 30 and 31 were coin flips. Some guys got injured. Some pitched poorly. Others were fine. At age 32 though things changed quickly and for the worse. However this group didn't completely die out like the other group in the 33-35 age range. For the right pitcher you could in fact still do well right into your mid 30s.

What's the conclusion? Well there isn't any. One of the strangest things about this whole process (other than realizing Edwin Jackson is only 31) is James Shields. For nearly all pitchers when they threw a ton of pitches over the course what amounts to 5 years around age 31 something bad happened. They got bad. They got injured. Something. But Shields ended up starting having these pitch burdens at age 29 and has kept them going right on through last season with no noticeable effect on his results. He did start a little later than most, and maybe he'll blow up next year but if he doesn't he'll really be an outlier doing something no other pitcher had done in the past 15 seasons. The point of that tidbit is there are outliers. There always will be outliers. Nothing can be said with 100% certainty espeically not with this kind of comparison work. (though I can break out some small-sample categorical analysis tests if you really want to find out if these groups are statistically significantly different)

Anyway I'm skirting around though what the question really is, what do I think? Well I'd say this. I'm not very hopeful that Scherzer will give the Nats any value past age 32. The above is a work in progress of course as some of these other pitchers Verlander, Lester, Greinke, etc. work into their mid 30s but looking at 50 or so pitchers during this whole analysis made it clear that pitching well in your mid 30s is hard enough given any amount of work. When you give someone a large workload, it becomes nearly impossible.

I do not see any reason though that we can't reasonably hope that if his workload is decreased and he's not already hurt, that he can pull a career like Oswalt did. I think frankly that would be pretty much on target for expectations from this contract.

Looking at all this, I'm not loving the deal no. I think from what he'll do on the mound, it's tough to see Scherzer give value back for the Nats on this deal. BUT that's only part of whether it's a good deal. The other part is whether they can win a series with Scherzer and he's probably got as good a shot as any arm at being healthy for the next 2 years which right now is what we care about.

As for the ZNN vs Scherzer - well this is just another piece of gold on Scrooge McDuck's side of the huge wealth measuring scales ("Scherzer's a better long term deal" people are obviously the Flintheart Glomgolds of the world). I imagine if the data wasn't skewed to get those that pitched a lot until age 29 we'd see that age 29 seasons are like age 30 and 31, maybe a tiny bit better. Meaning that you have a decent chance of getting a good year at that age. Given the real world age difference between ZNN and Max you are almost getting a year younger player for the contract. That favors ZNN giving you more good years right off the bat.  Add to that the fact that the "not heavy workload in their late 20s" group has Hudson, Lee, and Buehrle all successful into their mid 30s while the "heavy" has... well no one right now and you get a feeling that while it isn't likely, that ZNN does have a much better shot of being good through most of the next contract than Scherzer does.(those that don't think that at this point have to be literally draped over "Tommy John" hanging onto it for dear life)

There you go - my thoughts. Collect your own. Trade them with your friends.