"Circumstantial evidence is a very tricky thing," answered Holmes thoughtfully. "It may seem to point very straight to one thing, but if you shift your own point of view a little, you may find it pointing in an equally uncompromising manner to something entirely different."
—Sir Arthur Conan aka "Denny" Doyle in "The Adventures of Sherlock Holmes"

This morning my sleepy train ride to work was interrupted by, of all things, a putative sabermetric analysis of Ryan Howard's inability to deal with the shift in the free commuter rag, The Metro. It's a publication that is worth every penny: aside from their inability to master spell check in Word, their sports editor earlier this season proposed solving the Phils' early season woes by starting light-hitting Michael Bourne and batting him leadoff yet! This is a publication that has a great deal of difficulty covering the previous day's game highlights in the two columns they allot among the credit repair and escort ads.

Howard, a pull-heavy lefty bat, has been facing defensive shifts since his rookie year. The opposition did seem to step up its use in the middle of his MVP year last season, but to say that Howard's woes are due to the shift when he a myriad of issues have cursed him this year, seems laughable at best. That doesn't stop the Metro from quoting Howard, without editorializing, that he "believes that he would be near .400 if it weren't for the additional defensive help opponents place on the right side of the infield. "I've lost about 13 or 14 hits so far to the shift,' he said. 'That's lots of hits.'"

Howard has been injured and played injured at the beginning of the year. He was moved up to the third spot in the lineup by manager Charlie "I Need a Friggin'" Manuel to start the year. And he started the year in a good old-fashioned slump. It didn't matter how much of a shift the defenses used when Howard was striking out so often. Also, whenever Howard seemed to start rolling, opponents would just walk him which caused him to be less patient at the plate.

Even so, Howard is in the top five starting first baseman in the majors based on OPS at .937. He is two points ahead of Albert Pujols, whose average is fifty points higher but who has almost identical on-base, slugging numbers, home runs, and RBI numbers to Howard. The biggest difference between the two are the strikeouts, with Howard's K's (68) more than doubling Pujols' (32).

Here are Howard's breakdowns per month with his batting ratios and strikeout and walk rates:

OK, so Howard's strikeouts numbers are up this year (26% of his plate appearances in 2006), but they have gone up after April as Howard has started to be more effective at the plate. Note that as his walks have fallen, he has become more comfortable at the plate. Expect opponents to start walking him more regularly as teams start to realize that he is back.

Anyway, back to the nutty Metro article, in which they cite the percentage of singles among his hits last year (54%) as opposed to this year (only 46%). That sound very persuasive, right? His singles are, of course, ending up in the opponents' gloves due to the nefarious shift.

Did I mention that he has just 46 hits so far in his abbreviated 2007 season. The difference between last year's pace and this year's is four hits, four hits! Four hits are nothing really, but when you consider that his doubles have gone up from 14 to 20%, you can make a better case for four of his hits skidding through the infield and becoming seeing-eye doubles.

However, the real issue is not the makeup of his hits but his general lack of hits in general. In each of his three previous seasons, Howard has averaged a hit in 26% of his plate appearances. This year he is at 20%. And yes, most of the loss has happened with singles. Here are Ryan's hits represented as the percentage of plate appearances:

But where have those singles gone? Are they in shifted infielders' gloves as the Metro opines or did they go somewhere else?

Maybe if we look at the other stats (represented as percentage of the total plate appearances), the picture will get clearer

Yr	BB%	K%	IBB%	HBP%	SH%	SF%	GIDP%
2004	5%	31%	0%	2%	0%	0%	5%
2005	9%	29%	2%	0%	0%	1%	2%
2006	15%	26%	5%	1%	0%	1%	1%
2007	19%	29%	8%	0%	0%	1%	3%
Total	14%	27%	5%	1%	0%	1%	2%

The seven-point increase in walks and strikeouts (combined) more than accounts for the five-point drop in singles. One thing that is a concern, however, is the his times grounding into a double play have trebled since last year. Given that he went down with a left quadriceps strain, perhaps his speed was affected, but I am still surprised by the big increase in GIDPs.

Instead of plate appearances, let's look just at the Balls in Play (BiP), that is, the plate appearances that involved balls that could have been fielded by the defensive. In other words, home runs, walks, strikeouts, and times hit by a pitch are ignored. Here are Howard's stats represented as a percentage of BiP:

Again the singles are down and the GIDPs, sac flies, and doubles are up. It seems that earlier this year he had doubles and flyball strength but wasn't getting the ball over the fence due to ineffectiveness and injury. As for the singles disappearing, it seems that a good bit were converted to double play groundouts. That doesn't seem to support the shift overly affecting his game since they difficulty of turning a double play goes up when the defense is shifted out of position. Besides, shifts are used less frequently with runners on base to prevent against the stolen base.

In a typical shift the third baseman plays a to the left of second base, the shortstop plays the normal second base position, and the second baseman plays a sort of short fielder (sometimes the shortstop is shifted to this spot and the second baseman shifts slightly to the right). A ball hit to the short fielder is likely not to cause the runners to be doubled up. A ball to the normal second base position now becomes that much more difficult because a) the third baseman is now out of position as the pivot man on the throw to first and b) there is no one backing up the third baseman on the throw to second. This is another reason why the shift is rarely used with men on base: it takes you out of the double play.

So where are we? The shift, though annoyance, appears to have very little to do with what is—or rather what had been—ailing Ryan Howard. He has been frustrated by an increase in walks. The lack of decent balls to hit made him less patient or at least rustier at the plate and led to more strikeouts. (I say rustier since his pitches per plate appearance have gone up from 4.06 in 2006 to 4.20 this year. Them again a pitch-around walk requires at least four pitches so that increase may be a bit deceiving.) He has also been a disappointment to a certain degree given his increase in double play groundouts, but again, they have little to nothing to do with the shift.

I leave you with two stats. First, I combined Howard's singles, doubles, triples, sac flies, and GIDPs (represented as BiP percentage) and found that they have remained remarkably consistent over the last three years. So one could argue that the singles have gone to one or more of the other categories.

Also, I list the ground ball percent for balls put in play. Note that this is at an all-time low for Howard. So not only is he grounding into more double plays, but he is also getting under balls and flying out. The end result is—surprise!—he's having a worse year than last year, getting under balls and hitting double play grounders.

Then again, it is still early and we are still talking about a small sample of data. By the end of the season, it may all seem like a blip on the radar screen. Given the way he's been hitting since his return and his offensive ability I would not bet against him.

Oh, by the way, Howard is not even in the top five in NL All-Star voting so far. I can see the argument for Fielder and Pujols ahead of him, but his omission is indefensible. It's almost as bad as the talk going around that Barry Bonds will not make the All-Star roster even though he is arguably the best player since Ruth, he will probably break the career homer record later this year, and he is in the top three (behind Magglio Ordonez and A-Rod) in OPS in the majors. If Howard and Bonds are left off the roster in favor of Nomar Garciaparra and Carlos Lee, they may as well not hold the game.