Understanding Ryan Howard’s WAR

Ryan Howard‘s WAR is not so flattering as beat writer Ryan Lawrence of the Delco Times pointed out in a tweet earlier today.

A few reasons I can’t comprehend WAR: Jamey Carroll, Ike Davis and Jeff Keppinger all ahead of Ryan Howard.

Because WAR isn’t exactly the easiest thing to comprehend, I thought I’d take this opportunity to explain why Howard’s WAR is low as well as expound on the fairness of Lawrence’s take on WAR. First, the educational stuff.

What goes into a hitter’s WAR? On FanGraphs, there’s batting runs, fielding runs, positional adjustment runs, and replacement level runs. In layman’s terms, it accounts for offense and defense, debits or credits the player for his position, and then compares the player’s production to a replacement-level player.

Batting runs are based on wRAA, which stands for weighted runs above average. wRAA is acquired by subtracting the league average wOBA (weighted on-base average, which is equal to the league on-base percentage) from the player’s wOBA, then dividing by 1.15, which scales the number to mimic on-base percentage. Finally, you multiply by the player’s total plate appearances. So, for Ryan Howard, his wRAA is ((.369-.325)/1.15)*537 which gives us 20.5.

If you are playing along at home, you may notice that Howard’s wRAA doesn’t match up with his batting runs total. That’s because FanGraphs adjusts batting runs for park factors but does not do so for wRAA. Howard plays in a hitter-friendly park, so his number is depressed from 20.5 to 19.0.

Fielding runs are based off of the sum total of UZR at each position. Since Ryan Howard has only played first base, all we’re looking at is… first base. His UZR there is negative nine runs. There’s your fielding runs. Note that it is raw UZR and not UZR/150, which is UZR scaled to 150 defensive games. And, of course, there is the huge caveat of using just one-year’s worth of UZR data — it is a bit unreliable as it takes about two and a half years for this type of defensive data to stabilize. Dave Cameron of FanGraphs gives UZR an error bar of +/- five runs, so Howard could be between -4 and -14.

Positional runs credit or debit the player for playing more or less demanding positions. A +5 UZR is better from a center fielder than from a corner outfielder. This is done by adding or subtracting runs from the fielder (all are per 162 defensive games):

  • Catcher: +12.5 runs
  • First Base: -12.5 runs
  • Second Base: +2.5 runs
  • Third Base: +2.5 runs
  • Shortstop: +7.5 runs
  • Left Field: -7.5 runs
  • Center Field: +2.5 runs
  • Right Field: -7.5 runs
  • Designated Hitter: -17.5 runs

Ryan Howard would be debited -12.5 runs but FanGraphs further scales it based on his playing time. Currently, Howard’s positional adjustment deduction is -9.4 and will change very slightly over the next 20 games based on his presence on the field.

Finally, we have to deal with replacement runs. A replacement-level player is expected to produce -20 runs per 600 PA. Ryan Howard thus far has accrued 537 PA, so his replacement runs total is 17.9: (20*537)/600.

Then you take all of your separate run totals and add them up.

  • Batting runs: 19.0
  • Fielding runs: -9.0
  • Positional runs: -9.4
  • Replacement runs: 17.9

Howard’s total is 18.5. That is runs above replacement. To get to wins above replacement, simply divide that by ten. Why? Dave Cameron explains:

RS^2/(RS^2 + RA^2) = Pythagorean Winning Percentage. So, if a team scored 775 runs and allowed 775 runs, they’d have a .500 Pythag Win%, or 81 wins and 81 losses – even amounts of runs scored and runs allowed should lead to something like an even record. Not as scary as it sounds.

What happens if we subtract 10 runs from the runs scored column, so that we now have a 765 RS/775 RA team? Pythag spits out a .4935 win%, and .4935 * 162 = 79.95 wins. So, instead of 81 wins, you’re now expected to win just barely less than 80. By subtracting 10 runs, you lost a fraction more than one win.

Howard’s WAR, then, is 1.85. FanGraphs only rounds to one decimal place, so it shows up as 1.9.

FanGraphs compiles WAR slightly differently than Baseball Reference, which uses Rally’s WAR. For specificity, it is best to denote which type you are using: fWAR for FanGraphs WAR and rWAR for Rally’s WAR. I use fWAR but it is more a force of habit than anything else. rWAR is just as reliable.

Bat Field Rep Pos
Howard 19.0 -9.0 17.9 -9.4
Carroll 3.4 0.2 13.1 3.1
Davis 10.4 8.4 17.2 -9.2
Keppinger 3.8 -2.0 16.6 2.0

If you look at the leaderboards for WAR, it should pass the smell test. Most of the players at the top are players widely regarded as great players. Every so often, you may find someone who jumps out at you, like Andres Torres. Some people, like Ryan Lawrence, will see those fishy-looking players and completely dismiss the stat altogether.

i just can’t seem to take it seriously if Jamey Carroll & Jeff Keppinger are rated higher than Howard.

What Lawrence should have done before tweeting is look at each player’s value tab on their FanGraphs page. To the right is a breakdown of each player’s RAR.

The reason why Carroll is higher than Howard in RAR is because Howard is judged to be a poor fielder by UZR and because Howard has played a much less important position. Carroll has logged most of his defensive innings at shortstop and second base.

Similarly, UZR rates Davis as a great defender and that is the sole reason why he is ranked higher than Howard in WAR. Keppinger hasn’t been as bad defensively and has played more important positions.

Instead of Lawrence saying…

it is hard to take the stat seriously when i see those names ahead of Howard

… he should have looked up the numbers listed above, or at least asked someone who knows about WAR to explain the disparity instead of letting his confirmation bias take over and cherry-picking names to discredit the stat. If you had asked him before he noticed Carroll, Davis, and Keppinger, I would bet that Lawrence would have told you that WAR generally had a vast majority of players ranked appropriately.

Most people are going to look at Howard’s .282 AVG, 29 HR, and 96 RBI and think that he’s having a good season offensively. And he is — it’s just not a great season. With a .345 OBP and .525 SLG, he should not be in the same stratosphere as Albert Pujols, Joey Votto, and Adrian Gonzalez. Last year? Sure. 2006? Yeah, absolutely. But not 2010. So people see Howard’s 1.9 WAR and the company he’s in, and they balk as Lawrence did. When you dig deeper though, you see that it is completely justified.

Worley Should Get Nod Over Kendrick

The Phillies had been uncertain as to who would start this afternoon’s match against the New York Mets. Initially, their list included three pitchers: Kyle Kendrick, Vance Worley, and Nate Robertson. Robertson was obviously removed after allowing six runs in two thirds of an inning on Wednesday against the Florida Marlins. So, the Phillies were choosing between Kendrick and Worley. They chose Kendrick.

Kendrick has been, well, horrible in his last six starts, spanning 30 and two-thirds innings. His ERA is a bloated 7.04. While he has been BABIP unlucky (.359), he has a lackluster 10-to-9 strikeout-to-walk ratio. By trade, he is not someone that misses a lot of bats. A .359 BABIP wouldn’t be nearly as detrimental if he were striking out an average amount of batters (about 6.0 per nine innings). Instead, he has been averaging fewer than three strikeouts per nine in his past six starts.

Worley, on the other hand, is someone who can miss bats. In his Minor League career spanning over 380 innings, he compiled a 6.6 K/9 and 2.4 BB/9, both respectable numbers, slightly above the average. In his first Major League start on Monday, he struck out five, walked one and allowed just two runs in five innings of work.

There is no reason not to start Worley today. He is, by all measures, a better pitcher than Kendrick. What has Kendrick done to earn this vote of confidence?

This dilemma is somewhat reminiscent of the team’s handling of Scott Mathieson earlier in the season. Despite Mathieson statistically being an upgrade over about half of the bullpen, the Phillies kept him stuffed in the Minors. When they finally called him up, he was the victim of some bad luck and was subsequently sent right back down to the Minors. The team even invoked a rarely-used roster loophole to do so.

As great as things have turned out in recent years, sometimes the team makes you wonder if they really are still stuck in the Stone Age in terms of player evaluation. Even using rudimentary metrics, Worley is clearly a better candidate than Kendrick. With 20 games left in the regular season, the team should want to maximize its chance to win in each. The Phillies are passing up a golden opportunity to do so by not replacing Kendrick with Worley today.