Sunday, June 15, 2014

DIVERGENCE IN WPCT FOR COMPLETE GAMES...

We keep looking "under the hood" of that fading entity known as the complete game, which may be tantamount to looking a gift horse in the mouth...and we don't even have the excuse of being descended from a long line of dentists. As with so much of what happens these days, however, these things just can't be helped...

Would you be surprised to discover that the aggregate winning percentage in complete games since 1964 is below .800? The exact total: 21253 wins, 5604 losses, .791 WPCT. Probably not too surprising, given the former plentiful nature of CGs, though you may have expected complete games to produce more wins than that.

Turns out, though, that--back in the day--they did. Prior to 1973, pitchers who threw CGs had a significantly higher WPCT. The exact total from 1964-72: 6764 wins, 1092 losses, .861 WPCT.

From that point, however--as the chart at right depicts--CG WPCT took a hit. From '73 on, the aggregate CG WPCT dropped about a hundred points. The exact total from 1973 to now: 14489 wins, 4512 losses, .763 WPCT.

The historically astute among you out there (and who's to say you're not? You're certainly "out there" if you're reading this...) may have locked onto the specifics vis-a-vis the divergence in this data.

And you may well remember that 1973 was a watershed year in MLB history (and not just because it was the final season for Ron Swoboda, who was sent packing to the broadcast booth at the tender age of 29--and, according to many of you, not a moment too soon).

No, 1973 was the year in which the designated hitter (aka the "dreaded DH") was introduced to the game, creating some significant differences between the AL (who adopted the rule) and the NL (which didn't).

And as it turns out, one of these significant differences which evolved (or devolved, depending upon your perspective...) has to do with CG WPCT.

As the chart at left shows, the aggregate drop in CG WPCT that happens in 1973 would appear to be directly related to the implementation of the DH rule. AL CG WPCT drops below .700 in '73, while CG WPCT in the NL remains at the pre-DH level (somewhere above .850).

What's most interesting about this shift/schism is how it has persisted over time, even as the number of CGs went into freefall. Somewhere in the 90s, however, the pattern began to change...perhaps a new generation of AL managers began to veer away from what had previously been a knee-jerk reaction to let starting pitchers who were performing well in games where they were losing stay in the game all the way to the end.

However, as you can see in the chart, that change has remained mostly intermittent over the past twenty years, and while the gap has closed a bit between AL and NL, that's as much due to decline in the NL CG WPCT in recent years (a drop from .872 in 1973-93 to .844 from 1994 to the present) as an uptick in the AL (a .685 CG WPCT from 1973-93; .708 from 1994 till now).

Of course, one thing that has never happened from '73 on is the AL having a higher WPCT than the NL for starters who throw CGs. Until 2014, of course. So far, at least, AL pitchers are 19-2 in their CGs (.905 WPCT), while NL pitchers are 20-5 (.800).

Of course, it's a subset of what has become an increasingly small sample size, and it takes something like that to create the possibility of such a single anomaly out of 40+ years of data. As we always do, we'll keep an eye on this one as 2014 unfolds...stay tuned.