Livan Hernandez, old school innings machine.

Roy Halladay, CC Sabathia, and Mark Buehrle all say hi

since 2002:
Halladay: 271 starts, 1961 ip
Sabathia: 289 starts, 1946.2 ip
Buehrle: 298 starts, 1992.2 ip
Livan: 297 starts, 1939.2 ip

Why Are Modern Pitchers So Fragile?

Money.

Next question.

The freaks who can handle high workloads are underutilized.

The basic problem is that to date nobody has found a reliable way of identifying said freaks other than testing the non-freaks to destruction.

And while it's interesting to know at what point a pitcher failed, ... well see #3. Ted Higuera is a good example of why teams are risk adverse. Brewers thought they had a horse, signed him to a big (for the day) long-term deal and got zip for it.

Continuously declined continuously? Ye gods and little fishes...

The basic problem is that to date nobody has found a reliable way of identifying said freaks other than testing the non-freaks to destruction.

Does anyone think it might help if pitchers could be completely truthful when asked if they felt tired or otherwise strained?

The basic problem is that to date nobody has found a reliable way of identifying said freaks other than testing the non-freaks to destruction.

Of course given modern medicine, the downside of "breaking" a pitcher is much, much lower.

It probably makes more sense today to find out who's durable then it did back in "the day".

Hmm, a study on pitchers that assumes one 'inning" is the same as any other "inning", despite large changes in offensive environments from era to era. This year in the NL the average runs scored per game is 4.35, and average OBP is .324. In 1968 it was 3.0 and .300. Might pitchers have thrown a bit fewer pitches to get through each inning during the 1960s?

What happened in the 1960s and 1970s that caused an increase in pitcher workloads?

I suspect one factor was the increasing predictability of the schedule. Teams moved to sunnier climates, laid down artificial turf, put up domes, traveled by jetliner. Fewer rainouts, fewer doubleheaders, fewer travel days, more possibility of moving toward an ideal situation where four starters would each start 40 games.

Then that came to seem not such a great idea by the 1980s, but it was the ideal in the 1960s and early 1970s, at least. Rotations evolved from the early-mid-20th century pattern where the starters had a workload 1>2>3>4>5 (rotating by days rather than games, with frequent irregularities) to one where four guys would have a more balanced workload (the 1966 Dodgers, the 1971 Orioles as type cases). And in those highly-balanced 4-man rotations, the top guy would get 41 or 42 starts, which would have been very unusual a few years earlier as well as a few years later.

That's an overgeneralization; some staffs like Johnny Sain's early-70s White Sox were doing something else altogether, stepping up the load tremendously on just a couple of guys. So it's just one factor, in some cases.

Didn't RTFA, but if they are using innings as a measure I won't bother. I know pitch counts are not available for the 60's outside of some Dodger games, but the least you could do is use batters faced.

Pitches per batter has almost certainly increased, and outside of Angel lineups that end with Wood, Aybar, and Mathis, or pitchers hitting in the NL, pitchers can't relax for any portion of the lineup. Another thing is that throwing 120 pitches in 6 innings might be tougher on the arm than throwing 130 pitches in 9 innings.

Are we even sure that pitchers are any more fragile than before, or are we just remembering the Carlton/Ryan/Sutton/Perry survivors and forgetting about guys like Larry Dierker, Dean Chance, or Gary Nolan who didn't make it so far? Strasburg's 14 strikeout debut did not break a record, that is still held by Karl Spooner, who lasted all of 16 games, 488 batters faced.

Of course given modern medicine, the downside of "breaking" a pitcher is much, much lower.

Similarly, it may enable diagnoses that cause pitchers to be rested or used differently for reasons that were undetectable at an earlier time.

5,000 or more batters faced, 2005-2010

Does that table include post-season innings? I would guess that the three round playoff system puts the innings leader at the end of the season at at least 260.

Does anyone think it might help if pitchers could be completely truthful when asked if they felt tired or otherwise strained?

I doubt it. Pitching is a physically destructive act, and if we stopped every pitcher from throwing when they were tired or strained, they'd never pitch full seasons.

I love seeing how economical Halladay is. 1 of only 2 players on that list to throw less than 19 thousand pitches (and he did it to 235 more batters than Blanton in 136 more innings).

The number of complete games has continuously declined continuously since the late-1970s.

With two small blips (1880s, 1970s), the number of complete games has been falling since the 1870s.

Might pitchers have thrown a bit fewer pitches to get through each inning during the 1960s?

Clearly.

At least for the Dodgers the average pitches per start stayed broadly the same from the late 40s until now. 99 pitches gave more innings in the 60s than in the late 90s. It's just not huge. What's changed over the years is that back in the day when a pitcher had nothing they got him out quickly and if he was on he'd stay in the game. So Koufax was going 120+ pitchers fairly frequently.

EDIT: (Hopefully) for clarity

#11 I'm constantly amazed that there seems to be so little effort to identify issues with a pitcher before he has to go under the knife.

#10 I recall an old RSB post where I listed all of the pitchers who had arm problems in 1969. The notion that they don't make them like they used to didn't make a great deal of sense to me.

Anybody remember Steve Barber from Ball Four? I can't think of anybody like him today. I'm thinking specifically of the chronic pain issues, not the blew his arm out and trying to hang on.

#11 I'm constantly amazed that there seems to be so little effort to identify issues with a pitcher before he has to go under the knife.

Didn't Moneyball mention the A's trying this? With their injury record since then, you have to question if there's any benefit at all.

Pitchers are sissies now.

(In 1948, that is.)

Didn't Moneyball mention the A's trying this? With their injury record since then, you have to question if there's any benefit at all.

There's definitely a difference between the pitching injuries in the Peterson "pre-hab" administration and the Young administration. Zito in particular took the "flexibility to the max" concept to heart, making yoga the centerpiece of his workout routine. He's been the most durable pitcher among any recent A's pitcher. Mulder, on the other hand, did a lot more with weights. Harden was another "weights" guy. Hudson was a good athlete and fairly durable but had oblique problems. Of the current guys, Anderson and Cahill seem fat and worked hard at a young age. I'd bet on Gonzalez having the longest career of the three.

Pitchers are sissies now.

And of course Kid Nichols is right. During his peak, the 1890s, pitchers completed about 85% of their games. By the 1940s, it was down by about half. Of course, in the 1880s, pitchers completed 94% of their games, so I'm sure there is a quote out there somewhere about how soft 1890s pitchers were.

Of the current guys, Anderson and Cahill seem fat and worked hard at a young age. I'd bet on Gonzalez having the longest career of the three.

If the topic is longevity/injury avoidance always bet on the fat pitcher.

I doubt it. Pitching is a physically destructive act, and if we stopped every pitcher from throwing when they were tired or strained, they'd never pitch full seasons.

You mean as opposed to now, when they hardly ever pitch full seasons?

Money.

Next question

If money is the issue, why wouldn't teams want to find out which pitchers give them more bang for the buck by handling Halladay-esque workloads?

I'm constantly amazed that there seems to be so little effort to identify issues with a pitcher before he has to go under the knife.

The A's kept Ynoa out of games for nearly two years for this very reason, and he still needed Tommy John surgery after pitching in a handful of games.

Most obviously, the rise of the five-man rotation gave pitchers fewer games to cover. On top of this, teams began to rely more on relievers within games pitched than in the past, going with fresh pitchers in late innings rather than asking starters to pace themselves.

Pitches per batter has almost certainly increased, and outside of Angel lineups that end with Wood, Aybar, and Mathis, or pitchers hitting in the NL, pitchers can't relax for any portion of the lineup. Another thing is that throwing 120 pitches in 6 innings might be tougher on the arm than throwing 130 pitches in 9 innings.

Cool. This gives me yet another reason to despise the DH.

#20 What I'm thinking about are things like staff kinestheologists and regular MRIs. IE trying to actively monitor the current state of a pitcher's arm.

Does that table include post-season innings?

Yes, a point that's at least worth someone collating that data.

So if you RTFA, there are plenty of holes. First, not everyone who actually had 300+ IP in the 60s-70s upturn is listed--it's probably a formatting issue, since the guys missing (Stottlemyre, Tiant, Wood) are all at the end of the alphabet. Second, there's little if any extra work represented here than what was put out in the SABR 40 paper, which was itself not remotely earth-shattering in scope.

The big issue, completely unaddressed here, is looking at what impacts, short and long-term, came about from the way those thirty pitchers were used. Since we're missing some data for those years that we have now, we can only go so far, but here it is. In the season following their 300+ IP (if there are multiples, as is the case for 17 pitchers on the list, the first 300+ year is used), these guys averaged 275 IP the next year (not counting Blue and Singer, who missed parts of the following years for reasons not related to their arms). Overall, this group recorded 57% of their wins and 54% of their WAR in the years after their first (only) 300+ IP.

That's an interesting comp between wins and WAR, especially given that it's possible to go negative with WAR.

The averages for these 30 pitchers in their careers: 204 wins, 52.1 WAR, 115 wins after 300 IP season, 28.3 WARA (WAR After 300+ IP). The only top pitchers (60+ WAR careers) who got hammered by 300+ usage were Bunning and Tiant, and their usage could be said to have hastened their decline in effectiveness. But this happened at age 34, so it was a case of late overwork.

The casualties here were almost universally the lesser talents (Jones, Osteen, Goltz) or guys worked too hard late (Bunning, Tiant) or guys who'd already shown some vulnerability to heavy workloads but got a second dose (Singer). Of the young (-25) pitchers who got 300+ IP, there are a lot of HOF-types: Blyleven, Palmer, Jenkins, Drysdale, Marichal. The non-HOFers: McLain, Singer, Dierker, Blue.

Now if more than half the young guys to whom you gave a heavy workload wind up in the HOF, what can we really say about all this? Fluke? Small sample size? Sure. But it doesn't seem to support the hysteria that's existing regarding this issue--or the practices that have been implemented, either.

In 1919, the pitcher in the NL with the 10th highest IP had 248. Looking at every ten years, that peaked in 1969 with 284 (a fluke: the AL counterpart was only 250). By 1979, it was already down to 237. In 2009, it was down to 214. The regularizing of the rotation and the addition of the bullpen has taken about 20-25 IP off these guys over the past thirty years.

In 1919, the pitcher in the NL with the 10th highest IP had 248.

You may not want to use 1919 in this comparison, since that was a shortened season.

The basic problem is that to date nobody has found a reliable way of identifying said freaks other than testing the non-freaks to destruction.

Exactly. And experimenting with destruction in the effort to find one freak - let alone three, which is very likely what you'd need to sustain a return to the pre-1980 ways - is very likely to be counter-productive.

What's gone unremarked upon is that the typical relief outing has *also* gotten shorter as we have progressed from the mid-70s on. I think it boils down to one basic tenet: It's easier to find pitchers who can give you a good inning or two every other day than it is to find pitchers who can give you seven or eight good innings every fourth or fifth day.

-- MWE

NL IP for the #10 man, every 10 years beginning in 1920:

280.2 1920
237.1 1930
233.0 1940
249.1 1950
238.2 1960
258.2 1970
229.1 1980
223.2 1990
217.2 2000
208.2 2010 (so far, will probably wind up in the mid-210s)

-- MWE

If the topic is longevity/injury avoidance always bet on the fat pitcher.

QFT

What MWE said. It's not at all clear that, even if you've got a number of guys who can go 300 innings, that it's optimal even in the short term to have them do so.

Also, while I've never looked at it systematically, I tend to suspect that (basically) pitchers have so many innings in their arm and it's not clear that it's optimal long-term to burn those innings up in fewer seasons. That is Maddux, Clemens, Johnson, Glavine and Moyer all topped 4,000 innings pitching in 4.5-5 man rotations ... that's 5 of the 16 pitchers to do so in the expansion era. (OK, let's call it 17 as poor Dennis Martinez is sitting on 3999.2!) Most of the others spent about half their careers in such rotations -- e.g. Nolan Ryan never made more than 35 starts after age 30.

Now it is true that you see heaps and heaps of those guys starting their careers in the 60s. But in addition to the above, Mussina, Smoltz, Wells, Schilling, Brown, Rogers, Finley, Wakefield, and Pettitte all topped 3000 innings and started their careers after 1985. So that's 15 (of 47) who spent basically their entire career in 5-man rotations. And Livan will likely join them next year.

So while the 5-man rotation and heavy bullpen usage have clearly lowered the IP for starters in-game and in-season, it's not clear to me that they've yet limited IP for a career. Even if they have led to a reduction in expected career innings, it wouldn't surprise me if they've led to an increase in expected career seasons. And given that by the end of the 6th you're usually either in a game that's pretty much been decided or a close game where you want your most effective set of pitchers over the next 3 innings, it's not clear to me there's usually much advantage to having your starter go more than 6-7 innings (as long as you're willing to live with only 13 position players).

And given that by the end of the 6th you're usually either in a game that's pretty much been decided or a close game where you want your most effective set of pitchers over the next 3 innings, it's not clear to me there's usually much advantage to having your starter go more than 6-7 innings (as long as you're willing to live with only 13 position players).

I think the big losses are:

1) Can't effectively platoon.

2) Your good RPs waste too many IP finishing games good SPs could finish themselves, and are less available to be used optimally. i.e. you use 2-3 of your best RPs to finish a game where your #1 or #2 starter is up by 2-4 runs after 6-7 IP.

If that SP went 8 or 9 instead of 6 or 7, you'd be able to more often use your best RPs in the 6th or 7th innings of close games with lesser SPs. You never see a closer or top setup man enter in the 6th or 7th of a tie game, or even with a one-run lead, while it used to be fairly common.

You never see a closer or top setup man enter in the 6th or 7th of a tie game, or even with a one-run lead, while it used to be fairly common.

I'm wondering why there aren't even a few 2-3 IP relievers -- former starters who can get through the lineup once. You hardly ever see a 100 IP reliever anymore. So much for the next Rollie Fingers or Goose Gossage.

I'm wondering why there aren't even a few 2-3 IP relievers -- former starters who can get through the lineup once. You hardly ever see a 100 IP reliever anymore. So much for the next Rollie Fingers or Goose Gossage.

Probably because the CG has become an oddity. It's much easier to use a good reliever in 2-3 inning stints when you have a starter or three who can delive an 8-9 innings start with regularity. As it is now, manager have to go into every game thinking that they'll have to go to the bullpen and leave enough gas in the relievers' tanks to pitch the next day. If a team carries a long man these day, it's because he's not very good, and his only role is to take one for the team when a starter messes the bed.

I'm constantly amazed that there seems to be so little effort to identify issues with a pitcher before he has to go under the knife.

The A's kept Ynoa out of games for nearly two years for this very reason, and he still needed Tommy John surgery after pitching in a handful of games.

Yeah, this never works. Rudy Seanez had it right. Throw as hard as you can for as long as you can until something breaks, then get surgery and start all over again.

You hardly ever see a 100 IP reliever anymore.

You never saw them that often anyway, especially not guys who could do it year after year. 1977 seems to be the modern record with 25 pitchers doing it but I think 1984 is the only other year over 20 (24) since 1970. So even in the 70s and 80s, it was never the case that every team had one of these guys, usually it was about 2/3 of them.

Still it does surprise me that teams don't stick one of their AAA journeyman guys out in the pen to soak up a lot of low leverage innings. Sure, might turn the guy into the next Scott Proctor but then Proctor has made over $2 M which is, what, about twice what you'd make pitching 20 years in Iowa City.

Still it does surprise me that teams don't stick one of their AAA journeyman guys out in the pen to soak up a lot of low leverage innings

You never want it to look like you're not trying to win... it's not good for fan morale and probably not that good for team morale.

Also, I think teams UNDERUSE low leverage opportunities to find spots for young players (both hitters and pitchers). It's a good situation to see if a young guy CAN get all the way through a lineup by throwing 2-3 low leverage innings, which means he could be a potential starter. Major league playing time experience is extremely finite and thus very valuable... I don't think wasting it by just biding time is a good use of that time. Although I would agree that using your best pitcher because he "wants the complete game" even though the score is 7-0 is probably also not a good use of that time.