My take on Howard's '08

So I'm actually an M's fan (but I did predict the Phils to win the WS last year!), but there was a discussion on fangraphs.com about how the phillies and mets compared, especially each team's "Big 4". Reyes/Rollins, Wright/Utley, and Santana/Hamels all compared very evenly it seemed, but the Beltran/Howard comparison seemed to meet alot of resistance, many claiming that Beltran was far superior to Howard and the comparison was silly (as Howard tends to be irrationally overvalued by people like the BBWAA and the general media, the sabr community tends to lash back with an equally irrational undervaluation of him). So I did some looking into it using Howard's fangraphs page. I have no idea if this is a very saber-oriented community or not, I haven't so much as read a post, but either way I went pretty in-depth and figured I'd come on and share what I found out. Since I'm not sure how saber-oriented you guys are here I'll give you a quick description of some of the less-obvious stats because I'm proscrastinating my homework, if you still have any questions I'll try to clarify them.

WAR -  "Wins Above Replacement" - a giant conglomerate stat that quantifies hitting, baserunning, and fielding in terms of wins provided over a replacement player. The idea is that a team of "replacement level" players would win about 52 games in the major leagues - replacement level being rule 5 draft guys, waiver wire pickups, trade-throw ins, super-cheap FA's, etc. As an easy example, a team with 24 replacement level players (or 0 WAR players) and a 4 WAR player would then be expected to win 56 games. 2 WAR is generally considered a league average player, at least for position players. Google it if you want to know how it's calculated, it's an intricate process.

O-swing% - percentage of balls outside the strike zone swung at

O-contact% - percentage of contact made on balls outside the strike zone

Z-swing% & Z-contact% - the same as above, but with balls in the strike zone

LD%, GB%, FB% - percentage of balls hit that are line drives, groundballs, and flyballs

HR/FB% - percentage of flyballs that are home runs. More useful for determining how lucky pitchers are actually, but the application here is that if a hitter's HR/FB% usually stays pretty constant over large sample sizes, so if a hitter typically has a 20% HR/FB%, but is on a streak where he's hit little or no HR in a while but is still hitting lots of FB, it's assumed that he's just getting "unlucky" and the HR's will come back without any real approach change. Conversely, if the same historic 20% HR/FB% player is on a tear where he's hitting 50% HR/FB he'll be expected to come back to earth very shortly. (BTW, Howard has the best HF/FB% in MLB in the last 3 years)

BABIP - Batting Average on Balls In Play - Essentially, a measure of how many non-home run batted balls turn into hits. It's typically calculated as (hits-HR)/(AB-K-HR), and is another stat used to tell if a hitter has been lucky or unlucky, as it, like HR/FB, tends to revert to that player's historic average. The theory is that hitters only have so much control over where exactly they hit the ball onto the field, and that if a hitter is hitting alot of line drives but they're going straight to fielders (lowering his BABIP), it's mostly "bad luck", and it'll average out in the end as the hitter will revert to his historic BABIP and balls will begin finding holes instead of gloves again. In essence, it's a calculation of how much the opposing defense effects the hitters traditional stats - a lower-than-normal BABIP means the defense is turning alot of balls into outs, through some combination of good defense and luck, and a higher-than-normal BABIP means the opposite.

Hope you enjoy the post, and thanks for finally getting Moyer the ring that he deserves.

Howard’s stupidly overrated by the traditional media, but that doesn’t mean we have to stupidly underrate him here. He’s still a very good player, and honestly not so far off from Beltran that it’s ridiculous to compare the two. In 2006 Beltran put up a WAR of 7.2, while Howard went 6.9. Beltran again took a less than half-win advantage in 2007, 4.7 to 4.3. In 2008 Beltran stayed pretty consistant all-around with a 6.5, while Howard struggled to a 3.4.

So what changed this year with Howard? For one, he started swinging more - going from a pretty consistant career average of ~25.6% O-swing and ~70.7% z-swing to 26.7% and 73.7% - swinging 104.3% more often at balls outside the zone and 104.2% more at balls insid the zone, nearly identical numbers, so it doesn’t appear he was losing strike-zone recognition, just swinging a little more across the board. He posted the highest O-contact of his career with 40.9% (which gut intuition tells me for his player type is is not a sustainable skill, but he has shown improvement in each of the last 2 season so who knows), and his z-contact and overall contact fell right between his 2006 and 2007 numbers (the 2007 numbers being the lower of the set). I don’t know exactly what to make of this, other than he’s swinging more, and connecting more. It would be a problem if that boosted O-contact% was inducing more weak grounders and bloopers, but as his batted ball data shows nothing really supports that:
year LD% GB% FB%
2006 21.9% 41.9% 36.2%
2007 24.3% 31.5% 44.2%
2008 22.3% 41.5% 36.2%

2007 looks like the fluke here. And while he’ll certainly never match his 5.2% IFH% of 2006 and will probably continue to hover around 2%, that’s only a small handful of singles (3, maybe 4), and that’s offset by the idea that it’s very possible he’ll see his FB/HR regress (positively, that is) from 31.8% up more towards his career average of 34%. Not a huge movement, but it wouldn’t surprise me for that to fall around 33%, which would add 2 home runs on to last years total.

And then there’s the go-to “did he just get unlucky?” stat, BABIP. And, true to form, given his consistancy in batted ball type, it appears he was unlucky, as far more of his batted balls found a fielder’s glvoe than normal. FAR more. He hit to the tune of a .289 BABIP last season, a huge distance below his lifetime .334 mark (he hit .363/.336 in 06/07). If we’re to believe in BABIP at all we have to think that his ‘08 numbers were definitely hampered by bad luck (just as ‘06 was likely boosted about an equal amount by good luck). Regress his BABIP to a reasonable .320 and he adds on 11 hits. Assuming those 11 hits follow his career percentages of non-HR hits (73.5% singles, 24.7% doubles, and 1.7% triples), that gives him 8 singles and 3 doubles, and raises his line from .251/.339/.543 to .269/.350/.566 - I’m not going to bother with wOBA calculations, but this is a definite boost.

The first real thing that jumps out at me as a potential Howard-controlled problem is his BB% dropping from 15.7%/16.8% in 06/07 to 11.7% in 2008. THIS I do believe is likely a result of his increased swinging, and is the biggest indicator of a problem I can find so far. While the increased swinging doesn’t appear to be damaging his contact rate or batted ball type, it took away from his walks. Whether he actively decided to swing more, and the BB% decrease is a product of that, or if he’s losing his strike zone recognition and is thus swinging at more balls and drawing less walks, I don’t know. It seems unlikely however that he’s losing pitch-type recognition or losing significant bat speed, as that should (it seems to me) reflect itself in his contact percentage and batted-ball data as well. His K% was back in line with his career norm around 32% with a 32.6% mark, as compared to his extremely out-of-place 2007 37.6% mark. This correlates to the increased contact percentage I would imagine.

One more thing that’s possible regarding his BABIP is that his increased swinging is, while it doesn’t register in batted ball percentages, inducing weaker hits in general. Instead of hard-hit grounders up the middle and towering-fly balls to the track, he’s not getting good contact with the ball and hitting more weak infield grounders and lazy, bloop flyballs that find gloves easier. This seems a very plausible scenario, and is why I’m waiting so desperately for hitFX, to answer these questions.

In conclusion, I suppose it depends upon why you believe he underperformed last year. Best case scenario was the increase in swinging was a conscious decision, and he scales it back this year and returns to the 5+ WAR area as the walks return and his batted ball luck improves. Middle case scenario is that he’s simply losing judgement of what’s a strike and what’s a ball, but still has the ability to recognize the pitch type and drive the ball just as authoritatively as before, and just suffered bad luck. This would mean the BB% is likely to stay down, but the singles and doubles should come back, and he bumps back up to the 4-5 WAR area. Worst case scenario is that he’s losing his bat speed and is swinging more because he’s having to start earlier and “guess” the pitcher. As far as I can tell though this should manifest itself in an increase in K% and a decrease in contact%, neither of which were observed in 2008, and in fact the opposites were true as both increased over 2007. It’s possible that he simply “guessed right” alot (emphasis on “alot”), and if this is the case then scouts and managers should figure it out pretty quickly (bat speed is the type of thing old-school scouting types pick up on quickly, and if I can figure this out then I’m sure the new-school stats types have know it for months now) and start mixing in more offspeed and bendy stuff, in which case he’s just at the beginning of a long, painful decline. I can’t find anything in the stats to suggest that’s the case though, and I’m more apt to believe one of the first two.

While they’re not as close in talent as Reyes/Rollins, it’s far from absurd to compare Howard to Beltran. Everone but Oliver predicts the walks start to come back, and James/Chone/ZiPS project his 2009 wOBA to fall right in the middle of his 2006 and 2007 numbers, while Oliver and (the strictly mathematical) Marcel project his wOBA to fall in the middle of his 2007 and 2008 numbers. While I would take Beltran over Howard, it’s far from asinine to think that Howard’s WAR could, once again, be within .5 of Beltran’s (or perhaps even better than). He’s no MVP like the BBWAA would have you believe, but he’s still a very, very good player, and Beltran is not so far out of his league to make the comparison ridiculous.