Pitcher Value Retention

(OK, Injury Rates)

            I got into a brief, non-confrontational twitter exchange a few weeks about injury rates for pitchers.   Essentially, somebody was trying to talk about why injury rates for pitchers have exploded upward, and I said "Well, before you talk about why that has happened, shouldn’t you first establish that it has in fact happened?  How do I know it is true that more pitchers are getting hurt now?"   The reply to this was "It’s obvious," followed by the usual chorus of Twitter chirpers suggesting that if I hadn’t already observed this, I wasn’t paying attention.  

            OK, well. . .the theory that injury rates by pitchers have increased substantially is not a ridiculous theory or a silly thing to say, it is just that I need proof.   There are lots of things that everybody knows and everybody agrees on that are just not true.  In the modern world this happens because the media will pick something up and talk about and talk about it and talk about it until "everybody" believes it is 50 times bigger than it really is; not "everybody", of course, just us gullible people.   Before the modern media the same thing happened, because kings or churches or the extremely rich would pick up some idea and push it and push it and push it until prairie dogs were thought to be underground hippopotamus.  

            So anyway, how do we figure out if this is true or an illusion?  It’s hard, because there are so few constants over time.   Pitchers used to pitch 300 innings in a season and win 25 games in a good year.  Now they pitch 180 innings and win. . . I’m sorry, do they still keep won-lost records?   How many does a good pitcher win in a season anymore?   You can’t measure injuries by trips to the Disabled List, because the use of the Disabled List has changed so much.   You can check this out; there were times in the 1950s when a player would be out for two months and the team would not even bother to put him on the Disabled List; they’d just play with 24 men for a couple of months.   Also, they don’t call it a Disabled List anymore; I forgot that. 

            Anyway, the use of the injury list as an excuse to shuttle players around is now so common that you can’t study underlying injuries by studying the list of who is placed on the list.   So how do we study this?

I. Method

After worrying about that for a couple of days, I hit on what I thought would be the perfect way to study the issue.  That happens a lot in research; your ideas for new research seem to be perfect in your mind, and then you try to develop them and you realize that you didn’t think about this and you didn’t think about that and. . .oh, here’s this hard case where it doesn’t work.   This method didn’t turn out to be perfect, because they never are, but it did turn out be pretty good, and I was able to carry it through to the finish line and I got some pretty convincing results, so I thought I would report on those here, in case the way that I approached the problem would turn out to be useful to some other researcher sometime in the future. 

Here’s what I did, or what I tried to do.  I’ll start by explaining the method as I originally intended it to be, before some of the problems cropped up and I had to modify it. 

The underlying assumption is that value or production by a pitcher is so closely tied to good health that value retention may be used as a stand-in for health.   Yes, SOME changes in pitcher’s performance level—a few—are NOT tied to injuries, but result from aging or a loss of confidence.   But two things:

a)     When a pitcher is injured, he loses value 100% of the time, and

b)     When a pitcher loses value and in particular when he loses value suddenly, 90% of the time or more that’s the result of an injury.  

You can measure changes in injury levels, then, by measuring value retention.  My original concept was to identify all seasons by pitchers in baseball history by levels of contribution, consistent over time.   In other words, let’s divide all seasons by pitchers in baseball history into 10 equal groups, with the best seasons being "10s" and the weakest seasons being in Group 1.  Equal numbers each season.   I was saying "OK, pitchers stats from 2016 are a whole lot different from 1916, but there are the best pitchers in baseball in 2016, and there were the best pitchers in baseball in 1916.  Let’s look at the "Tens" from 1916 and see how many of them are still Tens in 1917, 1918, 1919, 1920, etc.   Then let’s do the same thing for 2016. . .not just 1916 and 2016, of course, but also 1926, 1936, 1946, 1956, etc., and not just 1956, but 1951, 1952, 1953, 1954, etc.  EVERY year.   The number of 20-game winners changes, the numbers of strikeouts change, the ERAs change, everything changes and there is a point after which some of the best pitchers in baseball are relievers, but if you create a constant frame of reference over time, then you can measure things within that frame of reference.   A standard approach in sabermetrics. 

For illustration, Walter Johnson’s "levels" by season are

            He’s a "10" every year through his peak.  Looking FORWARD from 1910, his first year as a 10, by this approach he retains 100% of his value for five years, retains 100% of his value for nine, finally has an off year in 1920 (because of an inury), so his ten-year value retention is 96%.   Obviously that’s unusual; not too many pitchers are tens every year.   Greg Maddux looking forward from 1991 had a 100% value retention for eleven years. 

            Even when Maddux wasn’t good anymore, he was still pretty good, compared to your average pitcher on the street.  So we have wiped out the "detail differences" between Johnson and Maddux, and represented them both as pitchers who were AMONG the best in baseball every year for a long time.  As opposed to the great-but-not-quite-as-great Stephen Strasburg, who had some top-level seasons but was not as consistent and, because of his injuries, was not able to sustain it for the same length of time:

            Strasburg was a top-level pitcher for the first time in 2012, and retained 92% of his value over a five-year period (46 divided by 50), and 67% over a ten-year period.  Which is still pretty good.

            One of the things that we do in sabermetrics is to reduce an array of different statistics into one statistic representing one season—WAR, or Win Shares, or whatever we are going to call this, the 10-Level scale or whatever.   When you compress multiple statistics into one, you lose a lot of things.   Your ONE statistic doesn’t have the complexity or nuance of the long statistical line from which it is derived.  For that reason, this might not be a completely appropriate way to evaluate the pitcher as a Hall of Fame candidate, for example, although it might be a contribution to that discussion as well, but that’s not what THIS number is designed to do.   What we’re measuring here is a player’s value retention over a period of time, measured in a way that is consistent across decades.

            You will notice that a "Level 10" season is a lower standard than, say, a Cy Young season.   Some Level 10 seasons are better than other Level 10 seasons; that’s embedded in the process.   Every Cy Young season in history is a Level-10 season except one; I’ll digress later and explain that.   Every Cy Young season is a Level-10 season, but every Level-10 season is a not a Cy Young season.   Since 1998 there are thirty Level-10 seasons by pitchers every year, but only two Cy Young seasons.  Historically, 99% of Cy Young seasons are also Level-10 seasons, but only 4% of Level-10 seasons are Cy Young seasons. 

            I had to modify my original concept several times in order to make it workable.  

1)     Originally I was going to include every pitcher/season in baseball history in my study.   I wound up cutting the 19^th century pitchers, because the leagues are so inconsistent, with the rules changing and teams folding in mid-season and new leagues forming and collapsing all the time and pitchers who pitch 400 innings in a season for two years and then disappear or move to the outfield and become .340 hitters.   The instability messes with the project, and 19^th century baseball isn’t major league baseball, anyway.

2)     Originally I was going to include EVERY pitcher, but there are lots of "pitchers" in history who just pitched 2 innings, and those guys also interfere with the process without actually contributing anything to the discussion, so I wound up deep-sixing them, too.   I put every pitcher on a 1-to-10 scale, but pitchers who pitched less than 9 innings (27 outs) are represented as zeroes, the same if they did not pitch in that season. 

3)     Originally I was going to rank every pitcher by their Season Scores, and I did, but I had to stop in the middle and re-work the Season Score formula to deal with some issues.  I invented Season Scores about 20 years ago and I modify the formula every three or four years, but I’m not going to stop right now and explain the latest round of modifications. 

4)      Originally I intended to divide all pitchers into ten groups OF EQUAL SIZE, but this proved to be unworkable, and I had to do something a little more complicated.   In 1900, the first year of the study, there were 8 major league teams and there were 57 pitchers who pitched nine or more innings, or seven pitchers per team.  In 1901 there were 16 major league teams and 114 pitchers, which is the same ratio.  In 2021 there were 30 major league teams but 683 pitchers who pitched 9 or more innings, or 23 pitchers per team. 

Major league talent, as we all know, is a pyramid, with many more players at the base than at the apex.   That’s not a statistical construct, that’s a reality.   Dividing them into 10 groups of equal size puts you in conflict with reality.  If you divide the 683 pitchers from 2021 into ten equal groups, you will have 68 Level-10 pitchers.  It just doesn’t match with reality.   It makes the performance bands too wide, too loose.  Guys who very clearly did not have Level-10 seasons wind up on Level 10.  I stayed as close as I could to the original concept, but I had to modify it somewhat to make the fringe-level groups larger than the top-level groups.   I’ll explain the modifications later. 

But having made those changes, then the study worked.  Sometimes studies work; sometimes they don’t work, and you can never measure what you are trying to measure.   This one worked.   In the end, I felt that I was able to do what I was trying to do.  

II. Details and Digressions

I mentioned that there was one pitcher who won the Cy Young Award but did NOT rank as a Level 10 pitcher in this study.   There have been 125 Cy Young Award Winners; 124 of them were classified by the process used for this study as Level 10 contributors to their team. 

I imagine that many of you are thinking that the exception would be Pete Vuckovich, 1982, but actually it isn’t.  Vuckovich was nowhere NEAR the best pitcher in the American League that year, and may well be the worst pitcher ever to win the Award, but he pitched 224 innings and was among the league leaders in ERA, went 18-6.  There are 26 Level-10 pitchers for 1982; he’s pretty comfortably among the Top 26.

The one Cy Young Award winner who did not qualify as a Level-10 contributor in this study was Mark Davis, NL Cy Young Award winner in 1989.   So what happened there?

It’s interesting; I don’t think I had ever looked that closely at it before.  At that time, the #1 stat for Cy Young voters was Wins.  Mike Scott was the only 20-game winner in National League that year, but Scott was not NEARLY as good in 1989 as he had been when he won the Cy Young Award in 1986, so a certain number of voters probably looked at him and said "He didn’t really have a great year this year," and passed him by.  Which was true; he had barely half as many strikeouts in 1989 as in his Cy Young season, and his ERA was up almost a run.   He finished second in the voting. 

Probably the league’s best starting pitcher was Orel Hershiser, who led the league in Innings Pitched and almost led the league in ERA, missed by .03.   He had the highest pitcher’s WAR in the league, a stat which didn’t exist at that time, but he was 15-15.  That disqualified him for a lot of voters, and he finished 4^th in the voting.  So two of the most obvious Cy Young candidates had enough check marks against them to hold them back in the eyes of the voters.

But actually, and oddly enough, there were a LOT of starters that year who had really good seasons.  Jose DeLeon led the league in strikeouts, but he was 16-12 with a 3.05 ERA, not bad but not obvious Cy Young numbers.  Greg Maddux had his first good year, and Joe Magrane (18-9, 2.91), and Rick Reuschel and Mike Bielecki.  They were all very good, but no one of them was great, and none of them was clearly better than Mike Scott or Orel Hershiser, so they didn’t draw any flies.  There were more pitchers having good years in the American League that year than in the National, and the Level-10 system doesn’t care which league you were in, just whether you were one of the best pitchers in baseball.

So in the Cy Young vote, there was no NL starter having a great year, so the voters turned to the league’s best reliever, Davis.  But in the Season Scores method, Davis gets pushed down the list by a bunch of American League pitchers and some NL starters who rank just ahead of him.  26 pitchers that year are Level 10 pitchers; Davis is 28^th on the list.   Not saying it is right or wrong, but that’s the way the process scored it.  

Speaking of relievers. . . the first sort-of reliever to qualify as a Level-10 contributor was Doc Crandall in 1910.  Crandall was a junk-baller with a funky motion, a colorful country boy, very popular with the New York press.  Rarely walked anybody.  He made 18 starts for the New York Giants that year, also made 24 relief appearances.  He finished the season 17-4 with a 2.56 ERA, finished 21 of his games in relief and would be credited with 5 Saves by modern rules.   He also hit .341 with 4 triples that year, played second base and the outfield a lot but not that year; that season he just played 1 inning at shortstop. 

In my data I list pitchers as "Starters", "Relievers" or "Mixed".   Crandall would be "mixed", and he was the first "mixed" pitcher to qualify as a Level-10 contributor, meaning one of the best pitchers in baseball.  In the next eight years there were three more mixed-use pitchers who qualified as a Level-10 contributors, but the first RELIEVER who made the list was Firpo Marberry in 1926.   Marberry is, of course, a landmark reliever; I suppose most of you know something about him.   He wasn’t really any better in 1926 than he had been in 1924 or 1925, but he was just as good, and for some reason he makes the list in 1926.   Wilcy Moore with the 1927 Yankees was 19-7 and led the American League in ERA; some sources list him at 20-7, which would make him the only "reliever" to win 20 games, but he made 12 starts that year and six of his wins were as a starting pitcher, so I have him listed as a "Mixed Role" pitcher, whereas Marberry had made only 5 starts, so I have him listed as a reliever.  

After Marberry the next reliever to make the Level-10 list, the list of the best pitchers in baseball, was Joe Page, who cracked the list in 1947 and 1949, then Jim Konstanty, who of course was the NL MVP in 1950, as a reliever.   In the 1950s there were 8 relievers who made the Level-10 list.  In the 1960s there were 14.  In the 1970s there were 30; in the 1980s, 37.  In the 1990s there were 31, in the 2000s, 49.  In the decade 2000-2009 16% of the Level-10 pitchers were relievers. 

In the decade 2010-2019 there were only 26, or 9%.  At first I wondered if that was some kind of glitch, but I concluded that there wasn’t; it reflects an underlying truth.   Star relievers are just not as big a part of the game now as they were in the era of Mariano and Hoffman.  Bullpens have changed.   The era of the Bullpen by Committee has arrived.   It’s not that teams don’t look for or rely on Closers, but the thing now is DEEP bullpens.   Teams like to have 7, 8 good relievers if they can get them, but usually there is some OTHER reliever out there who pitches more innings than the closer and has a better ERA.   Craig Kimbrel was the Dodger Closer last year, but Evan Phlllips pitched more innings with a better ERA.  Taylor Rogers easily led the Padres in Saves (2022), but no less than five other relievers pitched more innings with a better ERA.  Tampa Bay had 44 saves, but no closer, nobody with more than 8 Saves.  Ryan Pressly led the Astros in Saves, but Rafael Montero and Bryan Abreu both pitched more innings with better ERAs.   That’s at least as much the rule now as the exception.  Josh Hader was the Milwaukee Closer and easily led the team in Saves, but they traded him away at the trade deadline because everybody else in the bullpen had a better ERA and a lower WHIP.  

The last reliever to win a Cy Young Award was 20 years ago.  Eric Gagne, 2003. 

Since 1900 there have been 2,642 pitchers designated by this study as Level 10 contributors.  786 of them were also twenty-game winners, so 20-game winners comprise 30% of the Level-10 pitchers lists.   In that time there have been 833 twenty-game winners in the major leagues, so 94% of 20-game winners are also Level-10 pitchers, and 30% of Level-10 pitchers are also 20-game winners.  

In that time (1900 to 2022) there are 1,694 pitchers who pitched at least 200 innings with an ERA better than 3.00.  1,103 of those are listed as Level-10 pitchers, so 65% of pitchers who pitched 200 or more innings with an ERA below 3.00 are also listed as Level-10 pitchers, and 42% of Level-10 pitchers pitched 200 or more innings with an ERA below 3.00. 

The highest (worst) ERA for a Level-10 pitcher was 4.81, by Mike Mussina in 1996.  That’s not a great ERA, but the league ERA that year was 4.99.  Mussina led the league in starts, was fourth in innings pitched, fifth in strikeouts, had a great strikeout to walk ratio and was 19-11 for a team that was only 6 games over .500 without him.  So he made the list.   There are 51 pitchers in history who were Level-10 contributors with ERAs over 4.00, whereas there are 258 pitchers on the lists who had ERAs under 2.00.  

Closing the loop on a detail now, before we get to the conclusions.   I originally intended to have an equal number of pitchers on each of the ten levels, but, as I explained earlier, that didn’t work.  So what I did is, I made the Number of Level-10 pitchers each season (including 2020) equal to the number of major league teams.   In 1960, when there was 16 major league teams, there are 16 Level-10 pitchers, not one from each team, but 16 total.  The Braves have three (Spahn, Burdette and Buhl) and the Cardinals have three (Lindy McDaniel, Ernie Broglio and Larry Jackson), while six teams don’t have any.   In 1970, when there were 24 major league teams, there are 24 Level-10 pitchers; in 1980, 26 and 26, and from 1998 to the present 30 and 30.  

Then I increased very gradually the number of pitchers at each level, so that there are more Level-1 pitchers than anything else, but almost as many "2s" as "1s", etc.  

The number of pitchers per team, over time.   In 1900 and 1901, as I mentioned, there were 7.1 major league pitchers per team pitching 9 innings or more during the season (57 over 8 in 1900, and 114 over 16 in 1901.)  This number increased rapidly from 1901 to 1911, reaching new highs almost every year and reaching a peak of 11.4 in 1911.  

And there it stopped.  That number—the number of pitchers per team during the season—increased almost not at all from 1911 until the early 1950s.  Why did it stop increasing?

Roster rules.   In 1900 the rosters were not highly organized.  Teams would carry 16 players, 18, 20, whatever they felt like they needed.  In the 1890s and a few times in the first decade of the 20^th century, there were occasions when a team was short of players, and they would find a local player in the stands, put a uniform on him, and he would play in that day’s game, having paid his way into the game.  That only happened a few times and I don’t think any of those guys wound up pitching 9 innings in a season, but I’m just illustrating the conditions.  

But teams were carrying more and more players every year, and this was recognized as a problem.  About 1912, the leagues agreed on roster limits.  I’m not sure of the exact rules; there were different rules for different parts of the season and I’m not EXACTLY sure when the rules were implemented, but some time in there.   That’s not really what I am writing about.  

In 1946 there were 13.8 major league pitchers per team, but that was a one-off, as the players returning from World War II (1941-1945) blended gradually back onto the rosters, moving out the wartime replacement players.   Basically, the number of pitchers per team stayed in the range of 9 to 11.5 per team from 1911 to 1951. 

So if there is one pitcher per team (on average) who is level 10, and the total number of pitchers is 9 to 11.5 per team, then we’re actually staying very close to the original vision of an equal number of players at least level.  You have 10 pitchers per team; that’s one Level-10, one Level-9, etc.   The original plan.

After 1951, however, the number of pitchers per team began to increase gradually, as teams began to use the Disabled List more and to become more aggressive about calling new players out of the minor leagues.  In 1955 the number of major league pitchers per team hit 13 for the first time (except 1946).  But it stayed in the range of 12.0 to 13.4, with minor deviations, from 1956 to 1984.   In 1984 it was 13.4  per team, really not all that much different than it had been in 1911. 

And then, like so many things in baseball, the pace of change began to accelerate.  The number of pitchers used by a team in a season began to creep up.  And creep.  And creep.  And creep.   It hit 14 for the first time in 1985, hit 15 in 1990, hit 16 in 1995, 17 in 1997.   In the sacred and blessed 2004 season teams used 18 pitchers per team.   In 2009 it hit 19.   In 2015 it hit 20.   In 2019 it hit 21.  In 2021 it hit 22, 22.8.  683 pitchers on 20 teams, not counting the pitchers who pitched less than 9 innings; there were another 50, 60 of those, I think. 

Just my opinion, but anonymity in athletes is not a good thing from the fan’s perspective. Having so many players on a team that the fans don’t know who they are isn’t really a good thing, I don’t think. But in that era, nobody was looking out for the best interests of baseball.   Things were (and still are) changing at a tremendous speed, more strikeouts, more home runs, fewer singles, fewer stolen bases, longer games, then longer, then longer, more relievers, more pitching changes, then more, then more.  Nobody was pushing back against it; everybody was just doing their own thing.  Nobody had the big picture in mind.  As long as we’re all making money, it’s all good, man.  

We were heading toward a future in which teams used 30 pitchers per season, then 40, then 50, calling them up and dispatching them on a daily basis.  Whether that is changing now or whether we are still barreling forward in that direction, I guess we’ll see. 

.

III. CONCLUSIONS

            The belief that the number of pitcher injuries is increasing seems to be clearly true.   The rates of pitcher value retention have decreased steadily in recent decades, and have declined sharply in the last few years.   This is fully consistent with the wide-spread belief that pitcher injuries are increasing. 

            Dealing first with the Level-10 pitchers, the rate of value retention by Level-10 pitchers rose slowly but steadily from 1900 to 2000.   From 1900 to 1909 Level-10 pitchers retained on average 63% of their base year value over the next five years, and 42% over ten years. 

            These numbers dropped in the 1910s, because the Federal League screws with the data in ways that you can’t avoid, and they dropped in the 1940s because of World War II.  But otherwise they didn’t change very much and they generally trended higher.  In the 1950s the averages were 66% for five years, and 48% for ten years.  In the 1970s the averages were 66 and 49. 

            In the 1990s these averages (for Level 10 contributors) went UP to 68 and 52, which appears to be probably a statistical fluke, but could perhaps be because of steroid use.  Steroids were widely used by pitchers, and helped them to recover more quickly from injuries.   But most of the data in this study did NOT show increases in the 1990s.

            In the 2000s, these numbers dropped to 63% and 45%, which are actually exactly the same as the long-term historical averages, but a little lower than the norms from the last decades of the 20^th century.   But in the 2010-2019 era, the averages dropped to 58% value retention over five years, and 42% over ten years.   And they appear to have dropped more than that in the last few years.  It would appear that, when complete data for that decade can be figured, the lost value from pitcher injuries will be at a historic high-water mark. 

            Looking at pitchers at lower performance levels, the conclusion is the same except more so.   Among pitchers at Performance Level 9 in the base season, that group had a five-year value retention rate of 62% in the 1980s, a ten-year value retention rate of 42%. 

            In the 1990s, that dropped to 60 and 42.

            In the 2000s, it dropped to 57 and 39.

            In the 2010s, it was 53% and 35%. 

            It is too early to draw any conclusion about THIS decade, of course, but the early returns do not look encouraging.   We’ve lost Vermont, and we’re in trouble in West Virginia. 

            Looking at Level-8 pitchers.  These are still good pitchers; that’s like the #3 pitcher on your staff.   In the 1980s, the average five-year and ten-year value retention percentages were 63% and 46%.  In the 1990s they dropped to 61 and 44.  In the 2000s, they were down to 57 and 39.  And in the 2010s, these percentages--stable or increasing over time through the 1980s--these percentages dropped to 48% and 31%.  Not just historic lows, but historic lows by a wide margin, except for the anomalous World War II era, when most of the top pitchers left the game for military service. 

            Looking at Level-7 pitchers, pitchers a little better than average in their base years.  In that group, the five-year and ten-year value retention percentages for the 1960s were 67% and 44%.  In the 1970s, the Tom Seaver/Jim Palmer era, they were 63 and 44.  In the 1980s, the Fernando Valenzuela/Dave Stieb/Dave Stewart era, they were 61 and 42.   In the 1990s, the Roger Clemens/Greg Maddux era, they dropped to 57 and 39.  In the 2000s, the era of Pedro and Randy and Johan Santana, they dropped to 53 and 34.  And in the Clayton Kershaw/Justin Verlander era, 2010 to 2019, they dropped to 51% and 34%.     

            Do you want to see the numbers for Level 6 pitchers, and Level 5?  I’ll give you a hint:  they’re dropping.  Fast.   In the 1960s, Level 6 pitchers (in the base year) retained 66% of their value over the next five years, and 45% over the next ten years.   These averages have dropped in every decade since then, reaching 52% and 33% in the last decade, for which of course the data is incomplete, but the study is not structured in such a way that completing the data is likely to increase the averages.  

            When you do a study like this, you are compelled to try to think of every way that you could be misreading the data, every other possible explanation for the changes in the data.  Could these results be influenced by the 2020 pandemic, when some pitchers sat out the season?  Well, sure.  But the numbers for recent years are dropping faster than they dropped in the 1940s.  During World War II a large number of pitchers missed three full seasons in mid-career.   In 2020 a smaller number of pitchers missed one season.   It doesn’t really seem like that would explain the data.   The percentages are dropping because not as many pitchers are having full careers.   I don’t really think there is any other explanation for it.

            Let us ponder the anomaly for a moment.  Athletic training is better than ever.   Medical care is better than ever.

            We do Tommy John surgeries on 22-year-olds now, so that they don’t have to have that done when they are 27.     

            In the 1970s there were 40 pitchers who pitched 300 innings in a season, not 40 different pitchers but you know what I mean.  The last pitcher to do that was Steve Carlton in 1980.  We "protect" pitchers’ arms now, we "monitor" their "workloads", we count their pitches.   We schedule their workouts.   This is all done to keep them healthy.

            But does it?  

            The presumptive causes of this change are (1) excessive workloads at a young age, at an amateur level, and (2) max-effort deliveries.   In the 2002 to 2010 era, when I spent a lot of time with scouts, they would say about certain pitchers that "he has a max-effort delivery", meaning that he is throwing as hard as his body would allow him to throw.  Hell, everybody has a max-effort delivery now. 

            There may also be economic factors in play here—economic considerations from the player, and from the team. A player who has banked 40 or 50 million by the time he is 32 years old may not be in a frame of mind to fight through the injuries that pitchers often have in their early 30s.  I’m thinking of you, Josh Beckett, and you, David Price; maybe, possibly, conceivably you could have gotten to the other end of that struggle, I don’t know.  

            From the team’s perspective, maybe it is cheaper, more cost-effective to run a young pitcher out there for two years and let him blow his arm out than it is to sign him to a longer-range contract and make a commitment to him.  I don’t know; maybe it is.  Maybe those things are contributing to the changes in the game. 

            It is beyond my pay grade to try to solve this problem; I am merely documenting that it is really happening.   Careers really are getting shorter, not only in terms of innings but in terms of years.  Fewer top-level pitchers are able to stay at the top; fewer mid-level pitchers are able to stay in the middle.

            It is beyond my ken to sort all of this out, but I will make a few suggestions for whatever they are worth.   It is my observation that pitchers now throw much harder then they did a few years ago and I think are clearly better at hitting their spots, hitting the corners, but that they are not better and probably not as good at changing speeds, changing locations and varying the batter’s look at the pitch.   They’re not as good because, when they are coming through the minors, they aren’t encouraged or allowed to pitch 9 innings at a time, or even 7.   Here are some things that you might think about doing. 

(1)   When you are drafting amateur pitchers, give more weight to their amateur workload.   Maybe don’t draft the guy who throws 98 and has a great slider or sweeper or whatever.   Maybe let him sit in the draft, and draft the guy who throws 95 but ISN’T as well developed, because he hasn’t pitched 1,000 innings of amateur baseball. 

(2)   Maybe encourage your minor league pitchers NOT to go to those camps where they teach you how to turn your 88 MPH slider into a 92 MPH slider.  Maybe tell them that’s NOT what you want them to do, because there’s an injury speed bump in the middle of that road. 

(3)  Maybe don’t draft that kid who looks great, but who had Tommy John surgery when he was 18 years old.   Maybe focus on the kids who HAVEN’T needed to do that.   And maybe, when a pitcher is 22 years old and his doctor says that he needs to have Tommy John surgery now so that he won’t have to do that in mid-career, maybe you don’t do that. 

(4)  When you have a young pitcher in the minor leagues, maybe you make him pitch 7 innings a game or 9 innings a game, maybe you let him throw 120 or 130 pitches a game, because maybe if you do that he will have to back off the max-effort delivery and mix in some 80 MPH junk along with his fantastic slider, just to learn how to stay on the mound and throw 120 pitches.   Maybe the things that he will have to learn by doing that are the same things that will keep him healthy when he is in the majors and has a major league workload.

(5)  Maybe you don’t absolutely need to build a bullpen around five pitchers who all throw 98 to 100.  Maybe you’re not looking for Jonathan Loaisiga and Yacksel Rios for your bullpen; maybe you’re looking for Darren O’Day and Jake Odorizzi and Tug McGraw and Trevor Hoffman. 

I’m not denying that throwing hard gets hitters out; I know that it does.   But there are changeups.  There are screwballs, or used to be.  Maybe the fastball isn’t the best choice for everybody all the time. 

(6)   Maybe you don’t need for pitchers to use the exact same delivery on every pitch.   Pitching coaches are obsessed with "repeating your delivery"; I don’t know if everybody even knows that, but they are.   Maybe that isn’t always the best thing.   Maybe sometimes it is better to vary your arm angle, vary your release point, vary your arm speed.  Yes, that will interfere with your ability to hit your spots.  But maybe trying to throw every pitch from the exact same point, over time, creates injuries?  I don’t know; I’m just thinking aloud here.

(7)  Maybe, when a pitcher gets to the majors, maybe you DON’T take him out after about 80, 90 pitches, maybe you don’t take him out when the third trip through the order starts.  If he is a reliever, maybe you don’t tell him to throw as hard as he can for as long as he can.  Maybe you tell him that you hope he can finish the game.  Maybe it is better to let him figure out how to pitch through things, make him work through his trouble.  

(8)  And maybe the union can help us out here?  It is part of the traditional role of a union to try to protect their workers from injury.  Maybe there are rules that we haven’t thought of yet that can help us do that? 

Early Wynn said that when he had a sore arm, he would go to the bullpen and throw and throw and throw until he couldn’t throw anymore.  He said that avoided any major injuries.  That was a long, long time ago, and it was a different game. 

Tommy John said that rest won’t make a pitcher stronger; WORK will make him stronger.  That was a long time ago. 

Tommy John and Early Wynn were actually teammates for a few weeks on the 1963 Cleveland Indians.  That was a long time ago, and a different game.   But between them, these one-time teammates pitched in the majors for 50 years, from 1939 to 1989.  Maybe they knew something that we have forgotten. 

Thanks for reading.  Injuries to pitchers are increasing.  That’s a problem.  There is a solution.  Maybe you know what it is.