Throw It Like a Ballplayer

providing baseball commentary and ponderings since April 2010

Punctuated Home Run Equilibrium: Part III

Posted by dannmckeegan on May 27, 2010

From Coming Home Through Expansion: Waves #2 and #3

This is the third in a multi-part series examining the history of home run rates in Major League Baseball. Inspired by my disagreement with J.C. Bradbury’s opinion on the importance of steroids in the rise in home runs over the last 15 years, this research attempts to look beyond statistical expectations.  Rather than providing explanations for changes in home run rate, this series hopes to provide the reader with causal relationships not drawn from correlation, but rather from actual events.  Part One introduced Mr. Bradbury‘s argument, as well as my initial concerns with the position.  Part Two explores the Roaring Twenties and the decade’s end, where we see that a small piece of large puzzle can explain what appears to be widespread instability.

Following a plummeting rate of home runs per game (HR/G) during World War II, the return of the stars and the subsequent evolution of the American way of life led to what appears as a dramatic spike in power at the big league level.  Today’s entry examines what history has to say not from a purely numerical level, but also from a historical examination.

To be perfectly clear, I am not attempting to weave a narrative in retrospect.  Rather, my argument is that the peaks and troughs throughout history are largely evolutionary and incidental, byproducts of other socio-historical factors endemic to the game or experienced by the players.  The ups and downs that appear as data points are functionally independent of one another.  Extremes either high or low have historically tended to correct, if not overcorrect, themselves from year to year.  We will see this a number of times in today’s article.

When the Boys Came Home from War: Wave #2

HR per G chart

J.C. Bradbury's chart of HR/G from his Sabernomics website. As stated in previous entries, I disagree with the inclusion of year-to-year line segments connecting independent data points.

Nothing to see here.  Move along, people.  Seriously, there was no rapid increase in home runs between 1945 and 1950.  Sure, the chart says there was a spike.  But there is a reason that we have a saying in English about “lies, damn lies, and statistics.”  World War II happened and simply rendered four-plus years of data moot.  There is virtually no difference between the home run rate in MLB in 1940 and that of 1947 or 1948.

In 1940, before the U.S. became involved in WWII, there were 1,571 home runs in MLB, for an average of about 98 per team.  1941 was the season in which DiMaggio’s streak was established and Ted Williams hit .406.  But the power dropped to 1,331 home runs.  More pronounced than the Wilson/Hartnett drop is the fall of Cardinals trio Johnny Mize, Enos Slaughter and Terry Moore from 77 homers in 1940 down to 35 the following year.  So of the 240 home run difference, 42 (17.5%) belong to those three Cardinals.

As the rates fall for the next couple years and then jog up a tiny bit for the 1944-45 seasons, we can see that the majority of the drop was due to World War II itself.  Taking elite hitters off of active rosters limited the number of hitters who could hit the ball out of a big league park.  Even if some top-flight aces were also gone, the good hitters without power wouldn’t suddenly develop it.  In 1946, when players returned victorious from the war, they immediately made an impact on the home run rate, but four years of rust, age, and changing competition prevented a simple resumption of prior skill across the board.

But by 1947, that old rate was back.  As we look back, we see that something like a 0.2 HR/G increase occurred between 1921 and 1930.  Then, no net change happened comparing 1930 to 1940.  Fluctuation occurred based on individual performance and larger trends, but nothing permanent was developing.  Then, from 1940 to 1950, the HR/G rate again increased by 0.2 per game.  The jump appears to be so striking because of the WWII dip.  But given the level of performance of some of the games stars in 1940 and 1941 – Williams, DiMaggio, Mize, Hank Greenberg – it’s not hard to see that trend developing more smoothly were it not for the war.  It would have been difficult for the pitchers to readjust to this new level of power.  Not everyone had a rocket like Bob Feller, and home runs came to rule the day.

The Cold War Era and Expansion: Waves 3.1 and 3.2

Between 1950 and about 1980, there are some ups and down in the home run rate, but the net change in home runs per game between those two years is negligible.  The range, however, is something like 0.595 to 0.995.  The addition of expansion teams in 1961-62 and in 1969 added a lot of players to MLB.  And following Bradbury’s application of Eldridge and Gould’s principle, there ought to have been an increase in the standard deviation of talent levels.  This, then, would show itself in statistical changes.  Top individual players would turn in superlative performances against the marginal talents now garnering serious playing time.

The first expansion added the Angels and would-be Rangers to the American League in 1961, with the Mets and would-be Astros joining the NL in 1962.  In the same year that the AL expanded, the Class AA Southern Association disbanded.  The significance of this is that this league, based in the Deep South, had refused to integrate.  While other minor league teams and their leagues had followed MLB’s lead in the late 1940s, the SA refused to do so.  While the color line was broken in 1947 with Jackie Robinson’s debut, the Boston Red Sox’s first African-American player didn’t take the field until 1959, and that club didn’t fully embrace integration until the late 60s.

So integration and expansion came hand in hand.  The dilution of talent and the influx of African-American and Latin American talents largely offset one another.  The short-term drop in offensive production and home run production that occurred during the 1960s is understandable.  With 100 new roster spots to fill in MLB, there were certainly plenty of players from outside MLB’s historical channels ready to step in at a high level, rather than that of a rookie or reserve pushed forward.  Combined with the still-high mound, the enhanced net talent level (of course stressing the universal emphasis on starting pitching as the key to winning) led to a jog down in HR/G.

The final two years before the next expansion (1967-68) saw the most rapid drop, while the further dilution of talent in 1969 was but one factor that turned the tides in favor of hitters.  One other major factor was the decision to drop the apex of the pitching mound from fifteen inches to ten.  The other big factor was the redefinition of the strike zone.  The armpit became the new top of the strike zone, while the top had previously been the top of the shoulders.  The bottom of the zone was pinched from the bottom of the kneecap to the top.

An Aside: Pitchers, Control, and Changing the Game

The pitching mound issue bears extra focus.  I’ll come back to it at a later date as a subject in and of itself, but it was a major issue in 1969.  From the above-linked Sports Illustrated article:

But the lowered mound has many of the game’s theorists puzzled. Put simply, the mound is now only 10 inches higher than home plate. When a pitcher throws—and particularly a straight overhand pitcher—he finds the ground coming up to meet him a lot quicker than it did before, and he is thus off balance when he releases the ball. “The difference I have noticed is that most of the pitchers seem to be throwing high,” says Larry Shepard, the manager of the Pittsburgh Pirates and a close student of pitching. When a pitcher throws high he often throws himself out of a job.

Stan Musial, who this year enters baseball’s Hall of Fame at Cooperstown, leaned against the batting cage last week and looked out at the lowered mound. “It must help the hitters,” he said. “When a man is standing at the plate against a sidearm pitcher he has no trouble picking up the flight of the ball. He should be able to see all the ball. An overhand pitcher firing down off a high mound is another matter.

Don McMahon of the Detroit Tigers has already noticed the difference this spring. “Look at the number of walks that are being given up,” he said. “In some games there have been as many as 13 or 14. Many pitchers are high and wide. You really have to force yourself on the mound to get the ball down. During the season a pitcher who has to start forcing himself in the early innings is going to be awful tired in the late ones.”

So the lowered pitching mound was causing a lot of trouble for pitchers.  So arm and delivery angles, location and control, arm troubles, and other factors were coming into play based on this mound.  Looking at numbers of the aces of the period – Bob Gibson or Mickey Lolich, for example – we see that they were not all that affected by the change.  But the noticeable difference for Gibson between his storied 1968 season and 1969 was his walk rate.  :

Bob Gibson of the Cardinals was largely unchanged in 1969, so long as we consider the freakishly strong 1968 season to be a career year (aided by very, very soft baseballs and a little bit of luck).  However, the noticeable difference is the increased walk rate:

Year Age W L ERA GS CG SHO IP H R HR BB SO WHIP H/9 HR/9 BB/9 SO/9
1967 31 13 7 2.98 24 10 2 175.1 151 62 10 40 147 1.089 7.8 0.5 2.1 7.5
1968 32 22 9 1.12 34 28 13 304.2 198 49 11 62 268 0.853 5.8 0.3 1.8 7.9
1969 33 20 13 2.18 35 28 4 314.0 251 84 12 95 269 1.102 7.2 0.3 2.7 7.7
1970 34 23 7 3.12 34 23 3 294.0 262 111 13 88 274 1.190 8.0 0.4 2.7 8.4
Provided by Baseball-Reference.com: View Original Table
Generated 5/27/2010.

Detroit Tigers lefty Mickey Lolich likewise maintained a similar level of performance across the board with the exception of his control.  He wasn’t able to cut them back down until his amazing 376 inning season of 1971:

Year W L ERA GS CG IP H R HR BB SO WHIP H/9 HR/9 BB/9 SO/9
1967 14 13 3.04 30 11 204.0 165 71 14 56 174 1.083 7.3 0.6 2.5 7.7
1968 17 9 3.19 32 8 220.0 178 84 23 65 197 1.105 7.3 0.9 2.7 8.1
1969 19 11 3.14 36 15 280.2 214 111 22 122 271 1.197 6.9 0.7 3.9 8.7
1970 14 19 3.80 39 13 272.2 272 125 27 109 230 1.397 9.0 0.9 3.6 7.6
1971 25 14 2.92 45 29 376.0 336 133 36 92 308 1.138 8.0 0.9 2.2 7.4
Provided by Baseball-Reference.com: View Original Table
Generated 5/27/2010.

Generally as I’ve looked through the pitching numbers between 1968 and 1969, a trend emerges where the better pitchers appear to have had trouble with the strike zone, while the lesser but established pitchers experienced rises in both walk and home run rates.  This leads me to suspect that a more in depth analysis (note to self: when there’s not much to do in the off season, examine every pitcher’s year-to-year stats from 1967 to 1972) would correlate this specific leap in HR/G with BB/G.  Pitchers who press to throw strikes often get more of the plate than they’d like.  Perhaps this era’s home runs have less to do with diluted pitching talent (when Milt Pappas is a #5, there isn’t a shortage) and more to do with pitchers who are asked to do something (throw off a lower mound to a smaller strike zone) that hasn’t been asked of them before.

This was only able to spike the home run rate for a few years, though.  In fact, by 1971, the home run rate was back below 1966 levels, and the 1976 rate was a trough even lower than pre-expansion.  After another bottom in 1972, the adoption of the DH by the AL once again provided a temporary spike for 1973.  But once again, it appears that the league immediately readjusted and overcompensated over the next few years into a 30-year low in HR/G.

The leap of 0.3 HR/G from 1976 to 1977 is an oddity, certainly, despite the addition of the Mariners and Blue Jays to the majors.  But again, we shouldn’t focus too heavily on the year of the leap itself.  The introduction of the expansion clubs here and previously have correlated to big jumps in HR/G, but the area of interest should really be the few years thereafter.  How do the numbers change from year to year?  So the end of the 70s into the early 80s are interesting: the year-to-year numbers fluctuate some, but average out to similar rates from 1978 to 1983.

MLB grew over a 15-year period, from 16 to 26 teams, and had a large diffusion of talent to those 10 teams.  With 250 new roster spots, let alone the increased size of the minor leagues, there were many, many chances for inferior talents to stick and for previously ignored talents to prove themselves.  And because of the inconsistencies of league size, roster makeup, and a handful of rules and layouts; trends could not really emerge.  The talent level, the workload, the usage of bullpens, and even the evolving nature of pitching staffs in general play a role not only in the quality of pitching in MLB and on a specific roster, but also in the distribution of innings between the best and worst a given team has to offer.

The only real trend is more general, but still important.  As pitchers gain an upper hand, the eventual offensive adjustment is typically an overreaction.  The opposite is true, as well: powerful offenses that dominate pitching eventually find that pitchers have overtaken them.  It is a balancing act that, from year to year, deviates in seemingly inexplicable ways.  Lots of things were involved in the ups and the downs.  As we saw earlier in the week, even a couple players having monster or minuscule years concurrently can skew the numbers just as well as anything else.  The unpredictable long-term effects of expansion are rather unlike previous periods where expectations for growth could be predicted.  That is, using balls in the 1920s that eliminated spitballers and made it easier to see the ball would help hitters.  Likewise, the lowering of the pitching mound was intended to help hitters.  But it may have had the indirect effect of harming pitchers’ arm health, increasing the length of games (more pitches thrown on account of wildness, more pitching changes), and once again forcing a shift in the type of pitchers who will succeed.

Go ahead and read the SI article from spring training 1969 quoted above.  Think about the reactions of the hitters, pitchers and coaches.  Then think about the changes in the 70s and on into the 80s: more heavy bullpen use; fewer and fewer CGs and 300-inning starters; split-fingers, sliders and forkballs largely displacing overhand curves; three-quarter arm angles largely displacing long, slow, overhand deliveries; massive increases in the number of pitching injuries.  The little things they noticed then are the exact things we recognize now.  While not denoted in any box scores, the rule changes easily may have led to greater statistical change than any talent issue could have.

Wrap-Up

Friday will see the conclusion of this piece, with a look through the 1980s and on into the steroid era.  While the biggest official changes to the game took place during one off-season, their effects lingered and took time to sort through.  Pitchers who could adjust to the new angle would succeed, while others would lose their spots.  As we get closer to the steroid era, we’ll continue to see changes evolve naturally with a HR/G rate jogging alongside them.  I hope to provide a broader view of the landscape than Bradbury provides either on his website or in his book, The Baseball Economist.  Claiming that Rafael Palmeiro (3020 hits, 569 HR, .288/.371/.515, .283 BABIP, career 1:1 SO/BB, 759.5 RAR, 75.6 WAR, 3 Gold Gloves) is a nice player compared to a great like Mark McGwire (1626 hits, 583 HR, .263/.394/.588, .255 BABIP, career 1.21:1 SO/BB, 711.1 RAR, 70.8 WAR, 1 Gold Glove) is just a silly opinion to hold, even if it was written after Palmeiro’s positive test for stanozolol and before Big Mac’s teary-eyed confession.  Either way, both found themselves in the same clubhouse with Jose Canseco early in their careers and around the time their power grew.  There’s plenty to look at in the more recent past.  We’ll look at some ice crystals on the iceberg tomorrow.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
%d bloggers like this: