Written by Justin Zeth
Published: 05 July 2007
In Part 1 of the Smarter Pitching Staff series, originally titled “Whither Hast Thou Gone, Four Man Rotation?”, we looked at why the five-man rotation makes no sense (in quick capsulation: Why take 8 starts away from Johan Santana and give them to Sidney Ponson?) and then the reasons why the four-man rotation has been dead for 25 years. Now let's expand on that second thought and clarify the discussion, after which we'll talk about whether it's feasible in modern baseball to run with four starters and how it might be done.
What are the causes of pitcher fatigue? In order of importance, they are:
1. Pitch Count.
There is no question that some pitchers have more stamina than others and thus can throw more pitches per game than others. Twenty years of research has concluded, quite decisively, that pitch counts that stray over about 115, for most pitchers, are dangerous, and pitch counts over about 125 are destructive. As I mentioned in my last article, researchers believe that, once a pitcher is over 100 pitches, every 10 additional pitches make the outing twice as fatiguing to his arm. Pitchers very often will demonstrate a lasting loss of effectiveness after a 130+ pitch outing; it's unbelievable, knowing what we now know, that any manager would be irresponsible enough to let a pitcher throw that many pitches at once, and unbelievable that any GM would be irresponsible enough to continue to employ that manager. But the Raccoon Lodge idea that a pitcher proves he's a man by pitching until his arm falls off... well, that dies hard.
2. Pitch Count Per Inning.
We've known for twenty years now that pitch count for inning is very important; specifically, we've known for a long time that after a pitcher passes 15 pitches in an inning, he immediately begins a downward slide of losing effectiveness. That's not because he's losing concentration, kids; it's because going over 15 pitches in an inning exhausts his arm. In particular, managers need to do a better job of noting when a pitcher has a high-stress inning, and pull him earlier when it happens.
3. Length of the Game.
be obvious, but it's possibly the most-overlooked important factor in pitcher fatigue levels. Exerting your muscles pitching, then sitting cold for awhile, then exerting them pitching again, is hard on your body. Throwing a lot of pitches causes fatigue and then exhaustion of the arm, which obviously will, for most human beings, lead to shoulder and elbow injuries. But stretching out those pitches over an unusually long period of time will cause fatigue and then exhaustion of the entire body, and the body will begin to adjust to take some pressure off itself. The result is that mechanics break down as the pitcher's body tires, which, once again, will tend to lead to arm problems.
Pitchers that work slow cause extra fatigue for themselves, and a pitcher sitting cold in the dugout while his team has a big inning will also contribute to his fatigue level; that's a good reason why, if you have a 9-2 lead, pulling your starter after 5 or 6 innings is a good idea. In particular, this is a challenge for teams that build their lineups around plate discipline, taking walks, and working deep into counts, as a smart team should: That strategy is very effective at wearing out opposing pitchers, but it also contributes somewhat to wearing out your own
Pitchers will fatigue faster when it's cold (less than 60 degrees) or hot (more than 90 degrees). They will fatigue faster when it's very humid (a particular challenge for teams like St. Louis and Atlanta that play in humid climates. As you've probably heard by now, they fatigue faster at high altitudes (see Colorado and, to a lesser extent, Arizona). Teams need to account for this as well.
5. Bearing Down.
This effect isn't as empirically proven—or really provable—as those above it, which is why it appears down here on the list, but I do believe it makes a difference. A pitcher doesn't have
to throw with max effort all the time, and in fact, most pitchers don't. You hear the argument a lot that in the old days, pitchers would coast through games and save their best stuff for tight spots. I have little doubt that this was true, though perhaps not to the extent some believe it. Modern pitchers throw with near-max effort on nearly every pitch, and there are two reasons for this, one you probably know and one you probably don't. First, because nearly every hitter in a lineup these days is at least somewhat of a home run threat. Everyone tries to hit home runs now, and pitchers don't ease up even at the bottom of the order for fear of giving up a home run to the other team's weak-sticked catcher. And second, pitchers' agents keep them keenly aware of the importance of their own statistics when it comes time to sign that free agent contract. Yes, if your team's up 8-1, you can just cruise and win 8-6, but your agent will tell you it's better to bear down and win 8-2, keeping that ERA down and strikeout numbers up.
I don't believe this has a huge
impact on pitcher fatigue now, versus in days of yore—mostly because I think the pitchers of yore weren't on cruise control as much as some claim—but I do think it has some impact.
Can you see why fatigue levels in modern baseball are, beyond question, higher than fatigue levels in days past? Modern hitters, as a whole, work deeper into counts than hitters ever have before, driving up pitch counts; increased walks and increased overall offense leads to many more 34-pitch innings; and the same walks and offense levels, combined with players' tendency to stand around wasting time, has made the games significantly longer than they were even a generation ago, to say nothing of baseball in, say, the 1930s, when a typical game ended 8-6 yet took only 2:25 to play.
The answer baseball has come up with to counteract the effects of increased fatigue game-wide has been to switch to a five-man starting rotation. In other words, the people who run baseball teams became convinced that the effect of fatigue and the risk of injury to their best pitchers was so great that it was worthwhile to take 25% of their starts away and give them to the team's fifth-best starter/long reliever type. A typical ace gives up around 3-3.5 runs per game; a typical fifth starter gives up 5-6.5 runs per game. You're talking about knowingly allowing the other team to score between 1 and 3 additional runs per game, 32 times a year, because you think the fatigue problems are so bad that that's your only choice. It's worse than that, really, because fifth starters produce many more Disaster Starts (less than 3 IP, more than 5 runs allowed) and many fewer Quality Starts (6 IP, 3 or fewer runs) than staff aces do, meaning your long and middle relievers, who are presumably even worse than the fifth starter, are pitching a lot of innings that might otherwise have gone to your #1 or #2 pitcher.
You can infer from my tone in that last paragraph that I strongly believe using your fifth-best starter 32 times a year is not a reasonable solution to the problem.
So, what is?
Inform yourself. Equip yourself with information—and then hire a manager who you can equip with that information and trust to manage accordingly—to help you manage your best pitchers' fatigue levels while still getting 40 starts and 250 innings a year out of them.
How do we do it? Below I will lay out the program that I believe would help any team get more innings, and better innings, out of its best pitchers and phase some of its weaker pitchers out entirely. After all this talk, I am not
going to advocate a strict four-man rotation. No one runs strict rotations, and in fact, no one ever really has; modern baseball tends more
toward a strict rotation than baseball in decades past did, for numerous reasons, most importantly the former prevalence of double-headers that necessitated some rotation juggling and managers' tendency to use certain pitchers against certain teams, starting them on short or long rest to get them into what they, the managers, believed were the best matchups for their pitchers. But I am going to advocate:
1. Use only your 4 best starters as often as possible.
It should be brain-dead obvious that you want to use your best pitcher more often than your fifth-best pitcher; it should be brain-dead obvious that you want to use your fourth-best pitcher more often than your fifth-best pitcher. This is Rule Number One, and the goal we are trying to achieve with the rest of the rules to follow.
2. Keep track of your pitcher's fatigue in-game.
That means keeping track of his pitch count, of his pitch count per inning, number of high-stress situations, and the length of the game, also noting what the weather is like today if that's relevant. Hire someone to specifically do that job—an assistant pitching coach, if you will—if that's what it takes to keep your manager informed.
3. Know in advance when to remove your pitcher, and follow through.
It's nowhere near good enough to just say “110 pitches and then he's done,” or “Six innings and that's all,” or even “as soon as the game hits two hours, he's taking a shower.” You need to invest the time and money to know
each of your pitchers individually, so you can make informed decisions as to what he can and can't handle. Then you have to apply the in-game information we have from Rule 2. Now let's get into some particulars...
4. If your starter is getting kicked around, take him out!
If he goes out there for the first and clearly doesn't have it, don't try to wait a few innings and see if he figures it out. If the first and/or second inning is stressful—big innings usually feature pitch counts like 32 for the inning—it can really contribute to strain on the arm the further into the game the pitcher forges. If it's the second inning and he's already thrown 49 pitches and given up 4 hits, 2 walks and 4 runs, get him out of there. This is one of the perks of not using your fifth-best long pitcher as a starter: When one of your starters gets hit, you can get him out quickly and use one of your long men. If the long man struggles, you can leave him in a little longer and write the game off. You don't have to be as cautious with your fifth- or sixth-best long pitcher's arm, because, if I may wax Machiavellian for a moment, his health just isn't nearly as important to your success. Fifth-starter types are extremely fungible.
Remember, we're not running a strict four man. If your #2 starter goes out and starts getting shellacked in the first, replace him with your #5 or #6 starter immediately. Then your #2 guy will be fully rested and ready to go again in two days. A 40 minute, 40 pitch performance doesn't take long to recover from; that's about what a lot of pitchers throw on the side between starts.
5. If you're blowing out your opponent, take your starter out!
8-1 in the sixth inning is not a situation that requires the services of one of your four best pitchers. You can, and should, use bottom-of-the-staff fodder to finish those games. Yes, once in a while your bullpen will blow the big lead, but that will be rare, and you'll get more value out of being able to use your starter again sooner. If you're way ahead, be kind enough to your starter to let him pitch 5 innings and get his W, and then take him out. Hopefully he'll have only thrown 60-70 pitches in those 5 innings, and you can pitch him again three days later instead of the usual four (five in modern baseball). There's no sense wasting your best pitchers on blowouts.
6. Adjust your pitchers' rest schedules according to how much they worked recently.
Sometimes a game is too important to remove your best pitcher when he's hit 100 pitches, or 6-7 innings, or whatever; sometimes you need him a little longer. Never, ever let him ring up 125 pitches in a 4 hour game, but within reason you can stretch him a little. Do it only if you're sure you really need to—perhaps the bullpen's worked hard the last two games and really needs a break—and when you do, give the pitcher in question an extra day's rest. Again, you should limit this as much as possible.
The other side of this coin is just as important. If you pulled your pitcher early from his last start because the game was an early blowout, you can start him sooner! If he pitched 1.2 innings, threw 41 pitches and gave up 4 runs his last time out, one day's rest is usually enough before you can send him out again. If you pulled him after five innings because your hitters plastered the other team's fifth starter (poor suckers), two days' rest will be enough before you can start him again.
7. Encourage your pitchers to work fast.
You're not going to take J. Random 15 Year Veteran and tell him to change the pace at which he works, but you absolutely should teach your younger pitchers to work fast. Explain to them that working fast helps them keep their arms healthy; staying healthy long enough to get that big free-agent contract will be a powerful motivating factor for them. Changing a pitcher's pace will take time with a lot of pitchers, but teams need to invest what it takes.
8. Remove your pitcher as soon as he starts becoming fatigued.
Again, this is going to vary from pitcher to pitcher, but as a general rule, if your starter is pitching well enough, you may safely let him go about 100 pitches or about 7 innings, whichever comes first. If he reaches 100 pitches before the seventh, that almost certainly means he's had some high-pitch innings, and he should be removed sooner. In modern baseball, usually by the seventh the game's gone on long enough that it's time for your pitcher to call it a day, even if his pitch count is low; this, again, will vary from game to game. This is why a manager who knows how to think is essential to smart pitcher usage.
There are a few simple rules that influence when to take the pitcher out: If he has a high stress inning (more than 15 pitches), that should make you inclined to lift him a little sooner. If he has a very high stress inning (more than 22-23 pitches), you should be making a mental note to pull him after the fifth, even if he's only allowed one run, unless he follows that up with some very quick innings (doesn't usually happen). Basically, though, the cardinal rules already given are enough to keep your pitchers well-oiled: If it's a blowout, either way, get him out early; if he hits 100 pitches before the seventh, it means he's thrown too many pitches per inning and it's time to remove him; if he doesn't, let him pitch seven and then remove him. You can bend these rules as team needs dictate, but never, ever, abuse your starters!
The golden rule is: “Pitching doesn't hurt your arm. Pitching when tired
hurts your arm.” Don't let your pitchers pitch when they're tired.
This is a program that I believe any MLB team can implement as quickly as next year. Some pitchers are special and should be treated specially. Pedro Martinez, for instance, probably really can't pitch on three days' rest, now matter how carefully you protect him. He'll need four, sometimes five days between starts if you want him to have any chance of staying healthy. That doesn't mean you have to give everybody
four days' rest. You can run a modified four-and-a-half-man rotation, starting the rest of your pitchers every fourth day when possible and working Martinez in whenever he's rested. With a guy like Roy Halladay, who is both very effective and very healthy—Halladay works fast and keeps his pitch counts down—you can start him on a little more aggressive schedule and perhaps get near 300 innings out of him, if you use him right.
I believe that if you follow this program, you will get from your typical #2 type starter—an Aaron Harang or a Mark Buehrle or a Chris Young—about 40 starts. Some of those starts, say 5 to 10 of them, will be of the disaster variety where you pull him in the second inning because he's getting hit hard, but in most of them you'll get strong work for 6-7 innings. You'll end up with 240-250 innings pitched, and that's not
an excessive total; guys have been pitching more innings than that for many decades. Not only that, but you'll be funneling him to the most important 240-250 innings you can, by avoiding using him in blowouts. A typical team plays about 1450 innings in a season, give or take. If you do this right, you can give 900 of those innings to your four best pitchers—in modern baseball a team's four best starters typically account for about 700-750 innings—while keeping them at least as healthy, and probably healthier, than modern pitcher usage does. When you only have 550 innings to distribute to the rest of your staff, you can give 250 to your long men, 80-90 to your relief ace if you use him properly, 100-150 to your short relievers, and the rest will mostly be mopup innings. What I'm saying is that, if you systematically funnel innings to the correct pitchers, you can work most of the season with as few as 9 pitchers on your staff, certainly no more than 10 or very occasionally 11.
I believe there is an absolutely enormous amount of ground to be gained by an enterprising team that commits itself to optimizing the use of its pitchers. It doesn't make sense for the Yankees or Red Sox or Angels or Mets to do this—they're already on top and so have more to lose than to gain—but for a team like the Blue Jays or the Cardinals or the Twins, they could realize upwards of 8-9 extra wins by getting more and more important innings to their best pitchers, and make the difference between playing into October and going home for the winter. Middle-of-the-pack or on-the-cusp teams have the most to gain by streamlining their pitcher use, and not much at all to lose; the only difference between third place and fifth place, really, is that the fifth-place team gets to draft higher.
In the next few articles in this series—no promises on when they'll come out—I'm going to discuss a more comprehensive, organization-wide commitment to developing and maintaining a smarter pitching staff, and get into bullpen usage and roster management.