Nationals Baseball: Analysis Still in Action

Thursday, December 03, 2015

Analysis Still in Action

One of the knocks on Max Scherzer entering last year was that he threw a lot of pitches. That happens if you strikeout a lot of people (which Max does), throw a lot of innings (which Max does), and aren't particularly known for precise control (which Max wasn't until last year).

That last point is important because there was some hope during last season that Max wouldn't end up throwing a lot pitches because he was giving up a lot fewer free passes. While he did successfully lower his P/PA to a career low (3.73 after usually being over 4.00) there were enough innings, HRs and various other things to keep his pitches up near the top of the majors. That marks his 6th season in a row of being such.

Last season I tried to analyze the issue by looking at those pitchers who threw a lot of pitches in multiple seasons (10000 in 3 years). You can check out what I did here. The end result was both grim overall and grim specifically.

It was grim overall because pitchers just break down. Up until age 31 it seemed, pitchers were ok. Don't kid yourself, they weren't paragons of health at any age. Pitching was still a wasteland of broken arms and broken dreams, but at least with the group I was looking at (guys who pitched a lot from 26-29 and pitched well) ages 30 and 31 were coin flips. Didn't really matter if you had a lot of work or not, you had a decent chance of putting up a full good season*.  After that though the chances crashed and burned. Age 33 was particularly chilling with only 2 pitchers out of like 25 having a full good year.

It was grim specifically because when I compared those that threw a lot of pitches 26-29 vs those that didn't (but still pitched the equivalent of full seasons during those years) the ones that threw a lot of pitches did fare worse**. Not a lot worse but given that things were so grim you are looking for any hope you can find. With those that did not throw a lot of pitches, there were the occasional bright points. Arms that seemed to last through 35. Buerhle, Hudson, Lee. There were none of these for the group Scherzer was part of.

But this was admittedly with a very limited sample. Not only do few pitchers stay healthy enough to be part of this comparison but the comparison can only go back to 2002 because we only have pitch count data from 2000 on. So I was looking forward to seeing what another year of data would tell us. What did it?

Out of the guys who threw a lot in age 26-29 the bad outweighed the good. Verlander (ages 26-31 falling under "3rd season, 10000 pitches in 3"; 32 this year) couldn't throw a full season. Lincecum (26-29; 31) was bad and not full. Sabathia (27-31; 34) was full but not good. Shields** (29-33; 33) was the same. Weaver (27-29; 32) broke down. The exceptions were Scherzer (28-30; 30) of course, Jon Lester (29-31; 31) and Dan Haren (26-31, 34), who after flirting with above average years finally put one up

It was pretty much the same in the other group. Ubaldo Jimenez*** (26-27; 31). Greinke (26; 31) and Cole Hamels all had good years and Matt Cain (23-27; 30)**, Doug Fister, Anibal Sanchez, Gavin Floyd and Scott Baker all remained injured with various effectiveness.

Twist my arm and I might give a slight edge to the "didn't pitch alot group" because the bad pitching was from people that were injured staying injured as opposed to getting bad (like Weaver did) and frankly a caliber of pitcher a step lower than in Scherzer's group, but that's only if you twist. It's not really that big of a difference between the two groups, honestly. And its pretty apparent that we just don't have enough data points, especially in the Scherzer group. In this day and age using a guy like Max is used is an anomaly. Add that to walks going down and I'm not sure we'll get another data point any time soon. The other group added 7 pitchers this year (Price, Felix, Gallardo, ZNN, Gio, Cueto and Wei-Yin Chen) that we can use in comparison next year. Max's group added nobody.****

The takeaway from the analysis previously was "it seems like if you don't pitch a lot you have a chance to be good into and maybe past mid 30s. If you do pitch a lot it seems like you don't". Nothing here really changes anything. To do so we'd need a pitcher throwing a lot then be doing well right through his mid 30s. Dan Haren's good year was after 3 years of mediocre pitching, not to mention a reduced number of pitches (and now he's promptly retired so we really can say nothing). Shields could have been notable but right now looks like another log on the fire. Scherzer and Lester are still too young.

What this limited data has suggests is that Scherzer and Lester will breakdown, probably in the next 2 seasons, while one or maaaaybe two of Jimenez, Greinke, Hamels might do well for the next 4+ years. That's about as strong as I can go. It'll take at least 2 more years (for Lester) to really begin to refute this assumption.

*"full, good season" defined as 28+ starts, 100+ ERA+
** I was very interested in Sheilds because he was an outlier. Most guys throw a ton of pitches while younger, like starting at 24/25. Shields didn't really get going until 27 and was doing something no one else had been doing - throwing a bunch of pitches at an advanced age and doing well. 2015 though ended that. I'll again be interested this year to see if he bounces back though.
***some of this group pitched a lot either before that full 26-29 range or too limited time to fit into the "alot in 26-29"). The years they did pitch a lot are reflected here.
**** and no one will be added next year. The one guy with an outside chance was Lance Lynn and... well there you go. 


DezoPenguin said...

Not a lot of comments here...basically with Max we're all in the "hope we get several years before he breaks or degrades," "hope he bounces back after he breaks," with a side helping of "please just be Randy Johnson and make Rizzo look like a genius." Shields was weird last year, though; while he had at least one previous bad year (2010) for the Rays, the concept of a solid #2-grade starter suddenly becoming a dinger machine in San Diego of all places (though even so, he was still worth 1.1 fWAR and 1.9 rWAR (bWAR? What the heck are we calling Baseball-Reference, anyway?) and over 200 innings eaten, so even so he was more of an "overpaid #3-4" than actually "bad").

In other Nationals news, while I can understand non-tendering Stammen if they have any concerns about him coming back from injury, getting *rid* of otherwise-good bullpen arms seems like a bad move. And speaking of people who are objectively bad at baseball, what blackmail pictures does Tyler Moore have on Rizzo, anyway?

(Considering that my second-favorite team is the Blue Jays, I'm having an unhappy offseason.)

Harper said...

Dezo - someone has to be the next Randy, right? One of the things that always gets me about historical comparisons is when someone starts with Nolan Ryan. The guy was a freak who should be never used in comparison.

I used ERA+ not WAR. I wanted "good" I think WAR - because IP becomes a big part of it - can confuse that with "useful". Which is in itself not bad, just not what I was looking for.

Moore will pay for peanuts. The Nats like those guys

G Cracka X said...

Harper, what will be your measuring stick for determining if the Nats ended up being better off, 2016 and beyond, with Max or whether they should have kept ZNN by offering 'fair market' rather than 'low end of fair'? Is it just counting up the WAR that each guy produces for 2016 through the rest of his career? Or is there another way to resolve the 'Max vs. ZNN' debate going forward? What if Max puts up a great 2 more seasons, and doesn't do much beyond that, but the Nats win big in the playoffs one of those 2 years as a result? Could that possibly counteract ZNN potentially accumulating more overall WAR if each season from now on out is only 'good' (100-110 ERA+) at best?

blovy8 said...

Even if Max isn't worth it, weren't you the one saying that's the kind of thing you have to do to compete? The tightrope a franchise walks is figuring out the viable extension candidates vs. the underachieving, fragile guys with good agents. It's a difficult thing to be done sabermetrically or outside the full scouting report etc., as by the time you have enough data to be confident, the guy is WAY too expensive to get any sort of deal, and then why not just bid like everyone else with a QA-supplied pick in hand?

My feeling has always been that if we are going to assume steroids as a default position in recent outlier performance, we ought to look very hard at Nolan Ryan in the recent past as well. Who is like that guy?

Johnson probably isn't immune to that either, but he also was really friggin tall, and I guess my assumuption would be that the mechanics of the big guys are even more important than their velocity and age. Fister doesn't throw hard, is tall, and seems to also be very reliant on mechanics.

The Kitten won't cost them much if they cut him in March. It's certainly no worse a baseball move than signing Reed Johnson, and Heisey et al. Hey, Houston is looking for a Cris Carter replacement!

Harper said...

GCX - There's no one way. Anything that leads to a WS with Max being a big part I think you'd just have to accept as an end of the argument. Not that it would prove Max over ZNN definitely made them better off, perhaps ZNN does pitch better that WS season, more like "who cares". Assuming the Nats don't make a big playoff push (or you just like to do these things) then some sort of statistical comparison through the next 5 years would be fair. But nothing is going to be definitive unless you can play out alternate universes.

Basically I'm saying this is a fun "barstool" type topic more than anything.

blovy8 - It is the kind of thing the Nats have to do to compete though I guess I'd say - keeping ZNN and using that money to boost the offense in 2016-2017 would have made more sense, or using that money for in season acquisitions if necessary. Of course that's easy to say in hindsight. I LIKE the Max here for all in 2015. But I fear that wasn't the plan. The plan was Max n Lito replacing ZNN n Stras and I don't like that plan. I don't think it'll ever really line up like they want where as ZNN n Lito or Stras n Lito (if you MUST only have one expensive pitcher) might have.

I think you don't have to link all outlier performance to steroids especially prior to the late 80s unless you have reason to. Given that Ryan already did ridiculous things by then I think you just have to write him off as a once in a lifetime arm. You are going to have exceptions.

While I like to make a clean break from Moore, there really isn't harm done signing him, no.

John C. said...

I do agree with your underlying/understated point that it's hard to draw any conclusions at all because the data set is so small and the boundaries are so arbitrary (it would be interesting to see the results if you'd used FIP or FIP- instead of ERA+, or if you had moved the pitch count boundaries a bit). But the fact that there are obvious exceptions and the numbers are small increase the chance that there isn't much signal to the noise. It's not just Randy Johnson is a freak (he is, but in a lot more ways than pitch count/durability) - what about Mike Mussina? He pitched over 200 innings for nine seasons in a row, peaking with 237 2/3 IP his last year in Baltimore (Angelos obviously decided "the hell with it, he'll be someone else's problem anyway) at age 31. That was the 6th of the nine straight seasons 200+ inning seasons, and Mussina kept right on keeping on. His ERA+ of 143 his first season in NY, right after the 237 2/3 IP season, was one of his best, and he varied between good (109 ERA+) and very good (130 ERA+) the two seasons after that. And he was effective right up until his retirement at the age of 39. I'm not sure where his pitch counts were in that time range; it would be interesting to check. Given that Mussina is the comp that I've seen postulated for Scherzer (power pitcher with multiple weapons and good control that enables him to adjust as his raw stuff deteriorates), that's encouraging although no more definitive than anything else. In Scherzer's "similarity scores" on B-R, there's nothing really to judge from either. Looking at "most similar through age 30," the list of ten includes Halladay, Peavy, Matt Morris, Lackey, Bunning, Millwood, Lester, Lincecum, Appier and Haren. Conclusion ... um, I hope he's more Halladay and Bunning, less Lincecum and Haren?

As for the last line of "[I]t'll take at least 2 more years (for Lester) to really begin to refute this assumption, I think that's getting ahead of ourselves a bit. Given the limited data, I would say that, whatever happens with Scherzer and Lester, the jury will still essentially be out because even if things go well, your hypothesis could be correct (and they're just outliers) or incorrect. So a more balanced presentation would be that the data "suggests" that Scherzer "may" be more likely to breakdown.

And yes, I think you're unduly dismissive of the counterargument concern about Zimmermann and TJ surgery - but that's another story :)

SM said...

The weight of your analysis is sufficient to push most of us--John C. excepted--waist deep into the Slough of Despond. I do wonder, though, how significantly--if at all--your conclusions might differ if pitch count data were available beginning from, oh, say, 1970.

Being of a pessimistic bent--but in a good way!--I suspect your assessment of Scherzer's future wouldn't be much different. You mention those crazy optimists who point to Nolan Ryan's career and note correctly that he was an anomaly. (Not a "freak"--Robin Ventura was the last person to call Ryan a freak.)

There are always exceptions, of course. Tom Seaver, for example, had a pitch count limit in NY of 135--Ryan's was 150!--and had a long career. Randy Johnson, however, another beacon of hope, didn't start piling up innings until he was 26. But at age 32, he sat out most of the season with back issues. Similarly, Warren Spahn was 26 when he began carrying a heavy workload. (He had spent 3 years in the army as a grenade-throwing instructor or something.)

Steve Carlton pitched 193 innings at age 22, and over 200 innings for 16 of the next 17 seasons (190 IP in the '81 strike year), leading the league in IP at the ages of 35, 37 and 38. Now HE was a freak.

You could scour through the careers of 100 pitchers--foolish me, yes I did--with the longest careers and find one anomaly after another that doesn't apply to Max: the dead-ball era, the grandfathered spitball rule, the Roger Clemens magic vitamin regimen, Gaylord Perry's K-Y-Jelly pitch, etc. No hard statistical data here, obviously, just historical and anecdotal evidence. So I can't tell if any of this offers anything worthwhile to the discourse.

But I can say this: After the age of 32, the chances of Scherzer pitching like SCHERZER is possible, but it's not probable.

John C. said...

SM, I very much enjoyed your turn of phrase. In my travels about the internet, I've found that most fans spend most of their time waist deep in the slough of despair. All of the moves of [their favorite team] are foolish, doomed, or otherwise to be seen in the worst possible light. Moves of [other teams] are accorded much more leeway for optimism - or at least not DOOOOOM. I don't know whether it's something about fandom generally, or particularly on the internet.

FWIW, no one is saying that Scherzer is going to be SCHERZER next year, much less 4-5 years out. Hell, Bryce isn't guaranteed to be BRYCE going forward, and Bryce is only 23. I'm simply unconvinced that the model that Harper has used sheds much light on what a reasonable expectation is for Scherzer going forward beyond the general expectation that pitchers decline as they go through their 30's. As Harper himself acknowledges, the sample size is too small to really draw conclusions (of course, he then proceeds to draw conclusions). As we note every April when someone gets off to a hot or cold start to the season, the fact that April data is the only seasonal data we have doesn't mean that it actually provides anything useful for projecting the rest of the season.

I tend towards the meadow of optimism rather than the slough of despair, but I'm realistic about it. I've seen too many baseball seasons to be taken in by the idea that I/we really have all that much idea of what is going to happen. But it sure beats reading the front page of the paper these days.

SM said...

My apologies (I'm Canadian, naturally), but let me approach this another way now that I'm in the Lair of the White Worm. (I don't know if this will buttress either side of what is essentially a speculative argument.)

Of the 20 players who have had the longest major league careers--24 years or longer--exactly 10 have been pitchers. And the earliest year those 10 pitchers retired is 1983 (Jim Kaat). In other words, the pitchers with the longest careers are all from the modern era.

3 were power-pitching starters: Ryan, Carlton and Clemens.

2 were primarily knuckleballers: Phil Niekro and Charlie Hough.

The other 5? Dennis Eckersley split his career between starting and closing; Jesse Orosco was a closer, a set-up man, a LOOGY and a mop-up man; Kaat a finesse-pitching starter who gave up more than a hit per IP; Jamie Moyer, like Kaat a lefty starter (but not as crafty) who in 2004 gave up 44 home runs. (The longevity of his career remains a mystery.)

The other pitcher was Tommy John who, after his eponymous surgery at age 32, hung on for another 14 years. He was more of a finesse pitcher after the surgery, true, but still holds the record for most years played after Tommy John surgery. (And so he should.)

Attempting to glean from any of this--including the limited sample size of hard data--a reasoned projection of Scherzer's future performance is a tangle of converging possibilities.

Modern era pitchers have longer careers than those of earlier eras: Max earns a point.

Gloomy pitch-count data since 2000: Max loses a point

Small sample size: No points awarded.

And so it goes all the way down the line: Right-hander vs. left-hander; power vs. finesse; starter vs. reliever; PEDs vs. clean, etc., etc.

So, with all due respect: There is no way this Scherzer vs. JZimm debate can be conclusively resolved until the end of the 2016 season at the very earliest (if even then).

As much as I admire the intelligence and liveliness of the debate, might you now turn to another topic/issue/subject?

SM said...

@John C.

Only after posting did I read your response.

You are correct, of course. (I used SCHERZER as shorthand for "I-hope-he-doesn't-regress," rather than Cy Young-dominant. It was misleading.)

My subsequent post--I think--clarified what I should have said more baldly: There is no categorically definitive conclusion.

By the way, I'm having a "Meadow Of Optimism" fridge magnet made. It's as sweet a phrase I've read in some time.

Donald said...

I don't know Mike Maddox's approach, but it seems there's a strong argument for dominant pitchers like Strasburg and Scherzer to stop trying to be work horses. I don't know why that is, but they both seem to think it's a huge badge of honor to complete games or to work into the late innings. Maybe it would make sense to start dialing them back and tell them that 7 innings max is the most that's needed. There's so much analysis about pitch counts and innings limits for young pitchers and TJ recoverers, but not so much about guys entering their 32 or 33 age season. Maybe that's next. Of course, can you imagine the furor if the Nats were to shut down Scherzer in September over an artificially imposed innings limit?

Harper said...

SM/JC - I was hoping saying the data "suggests" added enough uncertainty to the "conclusion".

I'd say the most reasonable take-away is that when you hit your mid 30s chances of a full good year dramatically decreases (for those that started a bunch of games from age 26-29). This is kind of an across the board thing. A less reasonable take-away is that maybe by controlling pitches you can increase those limited chances. There isn't going to be anything scientifically significant from this data. I do though think that Lester going into age 33/34 with good seasons would be meaningful. Expecting someone to do something that hasn't been done before (which is what we're hoping with Scherzer - at least compared to these dozen contemporaries) is different that expecting someone to do something that has. It's meaningful in that way.

Moose is an interesting case. He did show up in pitch counts for 2000-2002 but not 2001-2003. I think he'd show up a handful of times if we did some estimation but not continuously. Being a control pitcher with not a ton of K's helps him. But with Moose something shows up that I've noted as an potential important longevity marker - the strike. Having 2 "light" years in your career has to have a positive impact.

Anonymous said...

Strasburg, a "work horse"? You've got to be kidding me dude. He's the antithesis of a work horse. He rarely pitches deep into games. In the equivalent of four full seasons worth of starts he has fewer than six innings per start, and just one complete game.

I agree fully with Harper about the Scherzer-Giolito plan though. In case anyone has forgotten, Giolito had Tommy John Surgery over three years ago. So if you believe in the "second elbow limited shelf life" theiry like the Nationals apprar to, that means the clock is already moving full speed ahead on him.