Announcement

Collapse
No announcement yet.

Pitch counts

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Pitch counts

    When did people start talking about pitch counts? I don't remember them from when I was a kid.

    Is there really any credible evidence that the wear on a pitcher depends primarily on the number of pitches he throws? As opposed to, say batters faced, time played in a game, pitches thrown per inning, etc.? And what's the rationale for not counting warmup throws too - that he doesn't give it all he's got?

    Seems to me that a guy who's pitched 10 pitches in the 1st and 40 in the second would be more worn out than if he'd pitched 25 pitches in each inning, no?

    Any solid research on this?
    http://biblemetrics.blogspot.com
    The blog for Israel Baseball League analysis

  • #2
    Well, batters faced, pitches per inning, time and so on would simply be an approximation for total pitches. No real to count warm up pitches since warm up pitches I'm betting are pretty constant nor do I think the drop off is of cliff like dimensions. It isn't like a pitcher has smooth sailing through 99 pitches and then as soon as the 100th pitch is thrown he falls off a cliff.

    I would assume a variance in pitch amounts during innings would lead to higher run scoring and a quicker exit. Afterall throwing high amounts of pitches in an inning means either you are letting a lot of guys on or that you are wild.

    Comment


    • #3
      Originally posted by Ubiquitous View Post
      Well, batters faced, pitches per inning, time and so on would simply be an approximation for total pitches.
      Maybe. But what I'm asking is which is the best approximation for pitcher wear and tear. What if a pitcher throws more pitches per inning, but has more rest between innings? (Or the reverse - fewer pitches, but shorter breaks between them?)

      Is there any research which indicates that pitch counts on their own really matter?


      No real to count warm up pitches since warm up pitches I'm betting are pretty constant nor do I think the drop off is of cliff like dimensions.
      Warmup pitches aren't constant - they're per inning. The more innings, the more warmup pitches relative to real pitches. Do those warmup pitches add to strain on the arm or the reverse? What about throws to first base?


      It isn't like a pitcher has smooth sailing through 99 pitches and then as soon as the 100th pitch is thrown he falls off a cliff.
      Clearly. But I'm less interested in assessing pitcher when a pitcher's losing his stuff and should come out - I would think he, the catcher and the coach should be able to tell regardless of how many pitches he's thrown. I'm thinking in terms of the impact on his effectiveness in subsequent outings.

      I would assume a variance in pitch amounts during innings would lead to higher run scoring and a quicker exit. Afterall throwing high amounts of pitches in an inning means either you are letting a lot of guys on or that you are wild.
      Or that they're fouling them off and working deep into the count. Or he let a couple runs score in the first inning, then pitched four eventless innings. Is that a reason to take him out?
      http://biblemetrics.blogspot.com
      The blog for Israel Baseball League analysis

      Comment


      • #4
        I believe the first team to track pitch counts was the Dodgers during Branch Rickey's tenure. This probably wasn't to predict fatigue or prevent injury, but simply part of Dodger statistician Allan Roth's desire to have a complete record of each game.

        In the mid or late 1960s the Athletics farm system had a rule limiting their young hurlers to no more than 100 pitches. I don't know if they were the first to do this or were simply copying another organization. This rule did not, as far as I know, carry over to the major league Athletics.

        The use of pitch counts in handling a pitching staff probably did not occur until the late 1980s. This was when STATS Inc began marketing their data to newspapers and broadcasters. Now that everyone could know how many pitches were thrown managers began to be second-guessed about this issue. At the same time salaries were spiraling upward, multi-year contracts had become the norm, and teams began to take a greater interest in protecting their pitchers. Before free agency 1-year contracts were the norm, and if a player suffered a career-ending injury the team was on the hook only for the remainder of that season. With multi-year contracts teams now had a financial interest in keeping players healthy for the length of that contract. Pitch counts became part of that, the theory being you could reduce the chance of injury by limiting the number of pitches thrown.

        As for whether or not pitch counts actually prevent injuries, I'm not sure that there is any definitive proof. When pitchers were expected to throw complete games they paced themselved differently. In the 1960s and 1970s there were many pitchers throwing 250+ innings, and a fair number throwing over 300 IP. Some of those pitchers got hurt, but many also had very successful and long careers. Today nobody throws 250 innings anymore, but strikeout rates are at a historically high level. Is it better to go all-out for 100 pitches or pace yourself for 130 or 140? Today we see pitchers throwing well beyond 40 years old, and that was extremely rare before 1990.

        One advocate of pitch counts was Craig R. Wright, co-author (with Tom House) of The Diamond Appraised. Published in 1989 when Wright was working for the Texas Rangers, one chapter of this book was titled "Learning How to Live to Pitch Another Day". Wright noted that pitchers who were worked hard before age 25 almost never had long careers. I'm sure Wright was not the first to notice this, but he did do a systematic examination of the issue. He advocated limiting the workloads of young pitchers, and pitch counts were a part of that. I think it is important to note that the current de facto pitch limits used in MLB are even more conservative than Wright's recommendations. Wright was primarily interested in preventing injuries to young arms, and MLB teams have applied pitch counts universally to pitchers of all ages.

        Comment


        • #5
          Thanks for the background. Still curious about the research, though.
          http://biblemetrics.blogspot.com
          The blog for Israel Baseball League analysis

          Comment


          • #6
            Evidence gathered since James refuted the whole Pitcher Abuse Points system seems to suggest that it's not the number of pitches you throw...it's the number of pitches in stressful situations requiring max effort and concentration that are key.

            No pitcher throws his absolute hardest on every pitch except closers maybe. The harder you have to throw to get the results you need...the worse it is on your arm...and note carefully I'm not talking about velocity...I'm talking about hardness of effort relative to your ability. This is why the most injury prone pitchers aren't the guys who throw 98 mph...they're the guys who throw 88. Those guys are usually not talented enough to throw that hard but they go max effort more to try to get the velocity.

            Comment


            • #7
              Originally posted by SABR Matt View Post
              Evidence gathered since James refuted the whole Pitcher Abuse Points system seems to suggest that it's not the number of pitches you throw...it's the number of pitches in stressful situations requiring max effort and concentration that are key.

              No pitcher throws his absolute hardest on every pitch except closers maybe. The harder you have to throw to get the results you need...the worse it is on your arm...and note carefully I'm not talking about velocity...I'm talking about hardness of effort relative to your ability.
              So is the pitcher drop off through a game due to fatigue, or is it due to batters simply hitting a pitcher better the more times they see him?

              I will also raise another question I've asked before. Are pitcher runs allowed normally distributed, or bimodal (good games and bad games). It would seem that with pitching, success breeds success in that if you don't let guys on base, you don't have as many stressful situations.

              Comment


              • #8
                Brett...the correct to the second question is NEITHER.

                Pitcher runs allowed follow a Weibull distribution:

                http://en.wikipedia.org/wiki/Weibull_distribution

                Allowing a run is analogous to failing and failure rates and all real major league baseball distributions, from batting average to game ERA follow the Weibull distribution. It means that their ERA is actually the product of many starts where they wree better than their ERA and a few starts where they were worse but it's not bimodal.

                The pitcher-drop-off is a combination of both slowly increasing fatigue (because as the game goes on, the pitcher faces an increasing number of stressful situations) and batter familiarity and timing.

                Comment


                • #9
                  Speaking of high pitch counts... what about what Dusty Baker's doing to Volquez and Cueto in Cincy right now? These guys are regularly getting up around 120, just like Prior and Wood did in Chicago in 2003.

                  http://www.sportsix.blogspot.com

                  Comment


                  • #10
                    I went to baseball-refernce.com to check some of the old Dodgers whom we have pitch counts for.

                    Don Drysdale starting pitching 200+ innings a year at age 20. He would consistently get to 120 or 130 pitches, not much more than today's starters, but in a higher percentage of games and mostly on 3 days rest.

                    However, 2-4 times a year he would have games of 160-175 pitches. It did appear, in casual observation, that these games did have a detrimental impact. The first was late in his first season, and he struggled mightily the first half of the next year. Most other times the ineffectiveness seemed to last the next few starts.

                    Not considering injury risks for the moment, why waste the next several starts to get a few more innings in the current game, instead of bringing in the bullpen?

                    I personally believe that the injury risk comes mostly from when a pitcher changes his mechanics when pitching tired. A coaching staff can try to build arm strength so that a pitcher doesn't become tired as quickly, and also teach the pitcher how to be more aware of his mechanics so that they don't change when pitching tired (per Mike Marshall). But the coaches must also have some system of knowing each game when each pitcher is tired, just not adhering to some arbitrary number per inning or per game.
                    Baseball Prospectus articles
                    FanGraphs articles
                    MVN Statistically Speaking articles
                    Seam Heads articles

                    Comment


                    • #11
                      If you look at the logs in detail, you will see that his pitches per start is the same as it was in the 1990s pitchers. Same for Koufax.

                      While they went 120+ pitches alot more than pitchers today, they also went 90- pitches far more than pitchers today. The variance was simply far far higher.

                      I first read about it here:
                      http://www.baseballprospectus.com/ar...?articleid=383

                      A related article is here:
                      http://www.insidethebook.com/ee/inde...e_old_dodgers/
                      Author of THE BOOK -- Playing The Percentages In Baseball

                      Comment


                      • #12
                        Tom...I have long been of the considered opinion that the reason pitchers went deeper into games back in the 50s/60s was because the game allowed it to happen because the umpires called a gigantic strikezone and the hitting was bad and there were fewer pitches per plate appearance and fewer plate appearances per 27 outs. It would seem your data on pitch counts confirms this.

                        Comment


                        • #13
                          Well, they went deeper alot more, but they also were pulled very quick alot more to. (See the BP article.) Overall, Koufax and Drysdale ended up with 105 pitches per start, more or less.

                          PA per IP was fairly constant. If I take 1964 ML: 29137.2 IP, 122997 PA
                          http://www.baseball-reference.com/pi...=TOT&year=1964
                          That gives me 38.0 PA per 9IP.

                          That's just 1 PA less than what we are used to today.

                          Back then, they were probably throwing 3.5 or so pitches per PA (total of 133), while today it's 3.75 for 39 PA (146).

                          So, my guess is that if you expect to throw, on average, 133 pitches per 9 innings, that's say 120 - 146 pitches per 9 innings. If you are at 110 pitches, you are probably close enough to smell the complete game in many cases, so they let them go for it. Not to mention that if you are the home pitcher, and you have the lead, you are only pitching 8 innings, meaning again, you can smell the complete game even earlier.

                          Furthermore, with managers having a propensity for pulling starters after only 60 or 70 pitches, that means that they'll give them the longer leash in subsequent starts.

                          Like I said, the average number of pitches thrown per start is exactly the same in the 60s and the 90s. It's the variance that is widely different. That, and that they would pitch on 3 days rest, not 4.
                          Author of THE BOOK -- Playing The Percentages In Baseball

                          Comment


                          • #14
                            Uhmmm.. correction... away pitcher, and trailing.
                            Author of THE BOOK -- Playing The Percentages In Baseball

                            Comment


                            • #15
                              I've looked at some of the links but haven't gotten to the stats yet.

                              However, off the top of my head I would disagree with Matt on the strike zone. In the 50's, there were more BB than SO, so I doubt that would come from a big strike zone.

                              The strike zone was enlarged from 1963-1968, and that definitely shows in the league walk and strikeout rates.

                              From my quick review of the game logs, the early exits seemed to be from ineffectiveness. There did not seem to be the situation of the starter is pitching well, but has reached 100 pitches, so let's get him out of there. If the starter was still getting guys out, he was left in until he might be pinch hit for. That was usually good for 120-140 pitches.

                              What seemed to be the injury risk to me was the 2-4 times a season the starter would go up to 170 or so pitches, instead of going to the pen.
                              Baseball Prospectus articles
                              FanGraphs articles
                              MVN Statistically Speaking articles
                              Seam Heads articles

                              Comment

                              Ad Widget

                              Collapse
                              Working...
                              X