Brendan Ryan’s WAR As A Paradigm For Why To Not Always Blindly Trust Some Stats

If you looked up the most valuable shortstops to date in the 2012 season, by their FanGraphs WAR, you would not find Brendan Ryan’s name on the list. That is because Brendan Ryan has not been sent out to the plate often enough to qualify for the batting title or FanGraphs’ criteria. However, if we were to sort shortstops by WAR on FanGraphs without a qualifying number of requisite plate appearances, Brendan Ryan, and his 1.7 fWAR, would rank as the 8th most valuable shortstop in baseball.  (For the uninitiated, a 0.0 WAR, or Wins Above Replacement, indicates a replacement-level player. A minor league callup or journeyman like, say, Brooks Conrad who is freely available several times a year. 2.0 WAR is a league average player. 5.0 WAR is All-Star territory.)

According to baseball-reference.com’s version of WAR, Brendan Ryan is not the 7th most valuable shortstop in baseball. He is the 26th most valuable position player in all of baseball. B-R states that Ryan has already put up 2.7 Wins Above Replacement, which, through only 253 plate appearances, is approaching MVP caliber play over a full season of 650 plate appearances.

Let me reiterate real quick: Ryan isn’t qualified for the batting title because his manager won’t send him to the plate very often because his manager thinks he’s not his best option on a very frequent basis.

According to Baseball Prospectus’s WARP (Wins Above Replacement Player), Brendan Ryan has been worth 1.0 wins. They rank Ryan as the 13th best shortstop in baseball. This puts Brendan Ryan on pace to post league-average production in only a limited number of plate appearances this season.

Have I mentioned that Brendan Ryan’s manager does not send him to the plate very much? At a time when a lot of people are coming to sabermetrics and very few are grasping the depth and nuance of some stats, WAR included, it would be easy to latch on to Ryan’s WAR and say “Free Brendan Ryan” and follow that with thousands of words about how stupid Ryan’s manager is and why Ryan should be played more.

But the truth is Brendan Ryan is arguably the worst hitter in baseball among non-pitchers who play regularly. Since coming to the league in 2007, Ryan has a career .289 wOBA. The shortstops with a worse wOBA than Ryan since then are Robert Andino, Alex Cora, Ronny Cedeno, Bobby Crosby, Omar “Hey Keith Law Is He A Hall Of Famer?” Vizquel, John McDonald, Josh Wilson, and Cesar Izturis. Of that list, only two of them are within 400 plate appearances of Ryan since 2007. Most of them have half as many. That’s because they are horrible (by MLB criteria, not, like, compared to you or me or a single-A player) at hitting baseballs.

Brendan Ryan actually gets a lot more plate appearances than he should because his glove is so good. If only 8 shortstops the last half-decade have been worse at the plate, only one, Nick Punto, has been better in the field. Ryan is truly a defensive whiz kid. His defensive stats are  mostly driving his WAR(P) numbers to their current levels at the Big Three sabermetric sites.

I say mostly because each site does credit Ryan with having a positive contribution on offense this year, which baffles the mind.

2012 Ryan: .182/.285/.271

2012 AL: .255/.321/.410

2012 MLB: .254/.320/.405

2012 MLB Shortstops: .258/.312/.377

2012 Edwin Maysonet: .250/.297/.350

That Ryan is way behind MLB average should come as no surprise. His Batting Average is below the Mendoza line and his On Base Percentage is virtually unplayable. That Ryan is even worse compared to his league, which has higher salaries than the NL and does not send a pitcher to the batters’ box, is even less surprising.

However, if Brendan Ryan is a positive contributor on offense (and B-R has Ryan at 0.6 offensive WAR, Prospectus has him at a more modest VORP ((WAR for hitting)) of 2.9) then he should have, I would assume, a marginally better batting line than the30th-best MLB shortstop. That’s Edwin Maysonet, who, among SS with 50 or more PA this year has the line above. I’m not a statistician by any means, but in a league in which, on average, SSs hit .258/.312/.377, a player whose triple slash is .182/.285/.271 cannot by any stretch of the imagination be considered a positive at the plate. When we put Ryan up against Maysonet, the picture-postcard-definition of a Replacement Level Player, and find Ryan is 68 points of AVG, 12 points of OBP, and 79 points of slugging beneath Maysonet at the plate, what we have is Brendan Ryan: Sub-Replacement Level Hitter. There are 15 shortstops after Maysonet and ahead of Ryan. Ryan is 16 spots below replacement level at the plate.

Whatever positional adjustments or other the Big Three are making in their WAR that is causing Brendan Ryan to be considered an offensive asset is, at best, in need of repair. If he were close to league average or better than Maysonet we could explain away him having a small positive at the plate due to playing in Safeco, a notoriously tough park for hitters. If he were not trailing the 30th-best shortstop by over 40 points of wOBA we could do that. But Ryan is not at that level. He’s dreadful at the plate. Awful. Were it not for that glove, that wonderful glove, Ryan would be unplayable at anything over triple-A.

The next question that comes to mind is how much defense one player can possibly contribute. Dan Turkenkopf, in the recent “Extra Innings” book from BP, noted that even the most prolific defender makes only two extra plays per five games. Over 162 games then the most plays a defensive player is likely to make is 64.8. Each play, though, is not a run saved. Depending on the leverage and run expectation of the situation, making a play could have more or less value. On average, a single is worth 0.48 runs. Brendan Ryan, a shortstop, is unlikely to stop many doubles or triples and certainly no home runs. If he makes 64.8 extra plays at 0.48 runs in a full season, Ryan will save just over 31 runs. Which would be good for 3.1 WAR, at the rate of 10 runs per win.

(A quick note, from a few weeks after this was posted. It’s been mentioned  by someone on a message board that we should credit Ryan with taking away a hit AND credit him for an out on each play. That’s basically double jeopardy for stats. He’s turning a ball in play into an out, not a hit, on an “extra play.’” We can credit him for taking away a hit, or we can credit him for adding an out, but you can’t credit him for both. That would be like crediting a hitter who records a base hit for two hits, as if he both created a hit and avoided an out. When a fielder takes away a hit, he takes away the hit, and that’s it, even though the out is recorded. He’s moving the situation from “expected hit” to “not expected hit” in terms of an “extra play.” If the play is not an “extra play,” the assumption is that the out will be routinely made. Making the “extra play” just takes the expectation of the ball in play back to a normal level.)

Ryan will make some higher-leverage plays worth more that 0.48 runs. He will also make some lower-leverage plays worth less than 0.48 runs. We’ll go on the average of 0.48 (culled from Tango.)

Certainly Ryan will make some double plays which will result in good defensive plays he makes in either starting or turning the play. That won’t add up to more than a handful of runs. I’ll say at his very most prolific, over a full season in which he’s the most active defensive player on the field, Ryan can add 4 wins.

The problem is Ryan is a shortstop and won’t be the most active. Outfielders are more active. Ryan’s team, the Mariners, has the lowest ground ball rate among all pitching staffs in baseball. And Brendan Ryan doesn’t play all the time because his bat is so bad. So Ryan can’t post a 4-win defensive season because he lacks the opportunity to do so. Ryan has fielded 157 balls in his zone so far this year, 21st in baseball among shortstops.

Ryan has only played in 74 games. Going by the “2 plays every 5 games” rule Turkenkopf came up with for the games most prolific defender, in his given opportunity the most Ryan could have created is 29.6 plays, or 14.2 runs. That’s 1.4 WAR. Not 2.7. Not 1.7. The most defense Brendan Ryan has added is 14 runs, and he has absolutely positively 100% for sure been a sub-replacement-level hitter to the tune of at least a half-win this year, meaning he’s somewhere around the 1 win, or WARP, Baseball Prospectus is giving him.

Ryan is league average, and he’s only league average because is glove is absolutely spectacular. He’s arguably the best defensive shortstop in baseball, and unquestionably one of the top 8-10. Unfortunately, he’s also one of the worst hitters in baseball, getting us back to that level of league average production. He’s certainly not a 2.7-win player through 250 plate appearances this year. That’s a joke.

 

About these ads

2 thoughts on “Brendan Ryan’s WAR As A Paradigm For Why To Not Always Blindly Trust Some Stats

  1. Dan says:

    Fangraphs has Ryan at -12.9 wRAA. Baseball Reference has Ryan as a -8 on their “Rbat” stat, which my understanding is their hitting portion of their WAR calculation. Something is definitely fishy here. After looking at it a bit, you may be correct that their definition of a “replacement player” is way too low. For example, Ryan seems to be getting an extra 0.9 WAR from Baseball Reference in the stat called “Rrep” (every player gets this, evidently). This stat does not appear to be based on anything he’s done, but is instead based on some comparison to a team with a 0.320 winning percentage. Well, there are no teams currently in baseball with that low of a winning percentage to compare players with, but they say it is an adjustment of +22 runs per 650 PA. In Ryan’s case he gets 9 runs just for appearing at the plate 265 times. So that washes out his -8 Rbat, and gets him to +1. They also have him as a positive baserunner, and they give him a bonus for playing a premium position (yes, even on hitting stats he gets that bonus). So that’s why his offensive WAR is a positive number. All these adjustments they make, plus premiums he gets for playing SS are helping his numbers greatly.

    Brendan Ryan may be an outlier that breaks the models these sites are using to create WAR. It’s probably not at all normal for a hitter as bad as him to even be in MLB. So my guess is that he doesn’t fit their mathematical models.

    • themiddle54 says:

      The idea that a guy who plays a position really well that most guys can’t play isn’t lost on me. And Brendan Ryan might be the best SS glove in baseball. But he’s not good enough with that glove to be an MVP-grade guy over a full year. That’s due to those positional adjustments and where the replacement level bar is being set.

Say Something Constructive

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 44 other followers

%d bloggers like this: