Friday, May 9, 2014

Nolan Arenado is not, and never will be, Joe DiMaggio...

...which isn't to say that what he's doing isn't pretty darn cool.

Just in case you have no idea who Nolan Arenado is, he's the third baseman for the Colorado Rockies of the National League, and he's currently enjoying a 28-game hitting streak.  The record, set by the legendary Joe DiMaggio in 1941, is 56 games, and it considered by many to be the most unbreakable record in all of sports.  By the simplest measure, Arenado is half of the way there.

Nolan Arenado's BA last year was .267. The odds of a .267 hitter getting at least one hit in a single game are 70.7%, assuming an average of 3.94 at-bats (Arenado's average for the season). The odds of this 28-game hitting streak are that number to the 28th power, which comes out to about .0061%, or a one in 16,408 chance. The odds of a 56 game hitting streak are that 70.7% to the 56th power, which gives you a .00000037% or one in 269,238,850 chance. Baseball plays a whole lot of games, so one stretch of 28 games for every 16,408 such stretches is not an unimaginable event. One stretch of 56 games in the next two- to three-hundred million stretches of 56 games means that a player like Nolan Arenado is going to have a 56 game hitting streak about the same time that humans are extinct and dinosaurs rule the earth again.

If you take Arenado's streak-influenced career batting average of .280, which one might or might not argue is more reflective of his true abilities, this current streak looks like a more likely event. Assuming a .280 hitter, this streak had a .013% or one in 7624 chance of occurring. The odds of a 56 game hitting streak would be .00000172% or one in 58,119,593. That's closer by almost one order of magnitude, but still ridiculously unlikely.

Arenado's batting average this season is .324. If that's reflective of his true abilities (which seems pretty unlikely, as Hall-of-Famer DiMaggio's career batting average was .325 in an era when more players were over .300), then the odds of him having a 56-game hitting streak are .000144% or one in 692,712. That still won't happen in this lifetime. This current streak, on the other hand, would have had a .12% or one in 832 chance of happening, which actually starts to look pretty likely.

That's why you see a hitter have a streak of 20-30 games every few years, why Arenado is not really halfway to matching DiMaggio's famous streak, and why Joe DiMaggio's record won't fall unless baseball is played for the next several thousand years.


(Credits: BaseballReference.com statistics for Nolan Arenado, Google calculator, and the Society Heading Egregiously Ridiculous Baseball Research Today [SHERBRT])                                                                

Tuesday, April 22, 2014

The Secret Wisdom of the Two-Man Pitching Rotation

It was suggested to me by a Pirates fan on an internet message board that the team would be better off if Gerrit Cole started every other game for the Pirates.  He suggests that a pitcher who started so many games could not fail.  In fact, there is no precedent for such a move.  No pitcher has ever started eighty games, and few have started more than seventy.

Can a pitcher who starts at least seventy games fail?  I suppose that depends upon your definition of failure.  Bobby Matthews started 70 games for the New York Mutuals of the National Association in 1875, and finished 69 of them. He had an ERA+ of 90.  ERA+ works like IQ.  100 is average.  90 is close to a standard deviation below average.  That's okay for an innings-eating fifth starter, but while pitching close to half of a team's innings is an impressive feat, if half of a team's innings are pitched that poorly, it will make it hard to win games.

Every other pitcher who started 70 games or more has been a rousing success, by any definition.  Pud Galvin started 75 games for the Buffalo Bisons in 1883, finishing 72, and managed an ERA+ of 117, well above average, while leading the league in innings pitched.  He followed that effort with and even more-impressive 1884 campaign of 72 starts, 71 complete games, and a 155 ERA+ with the Bisons.  Galvin shares the record for games started with one Will White, who pitched 75 complete games for the Cincinnati Reds in 1879.  He predictably led the league in innings, and posted an ERA+ of 120.

Others to start 70+ games in a season are as follows:
Jim McCormick (1880 Cleveland Blues):  74 starts, 72 complete games, 126 ERA+
Charles "Old Hoss" Radbourn (1884 Providence Grays) 73 starts, 73 complete games, 205 ERA+
Guy Hecker (1884 Louisville Blues):  73 starts, 72 complete games, 170 ERA+
John Clarkson (1889 Boston Beaneaters):  72 starts, 68 complete games, league-leading 150 ERA+
John Clarkson (1885 Chicago White Stockings): 70 starts, 68 complete games, 163 ERA+
Bill Hutchinson (1892 Chicago Colts) 70 starts, 67 complete games, 113 ERA+


Galvin and Clarkson, the only two men ever to accrue two seasons with 70-plus starts, are Hall of Famers, as is the Old Hoss.  Perhaps there is something to this two-man rotation thing.


(Credit:  BaseballReference.com pitching statistics)