The 5 (or so) Average-est Position Players of 2009
This June I posted a ranking of the “average-est” positional players in baseball (according to current WAR). R.J. liked it so much that he “borrowed” the idea for a post later in the summer. It’s just a toy stat, but a fun one. Simply subtract the “replacement” runs from a players contribution, and the absolute value of the remaining runs is the amount a position player is away from league average. (It’s not as straightforward with pitching.) Part of the fun is seeing the different ways one can get to be league average in relation to hitting, fielding, and positional value. So now that the season is over, who were the average-est position players of 2009?
The number after the player’s name is the players “absolute” difference from average according to stats here at FanGraphs (remember the usual qualifications and take this with grain of salt). Unlike earlier in the season, there are fewer players at the extremes of hitting and fielding value.
5) Clint Barmes (1.3). Maybe I just don’t follow the NL West closely enough, but am I the only one who gets this guy mixed up with Garrett Atkins? I shouldn’t, because despite being about as bad at the plate as Atkins this season, he’s a far better defender. I was surprised to see that Barmes has logged more career innings at SS than 2B, and has performed well there (+6.2 career UZR/150).
4) Magglio Ordonez (0.8). This is a shock, given just how public Ordonez’ struggles with the bat this season were. Well, at least the Tigers avoided triggering that massive vesting option for 2010. Oh, wait.
[Bengie Molina (0.7). It's the Giants' awesome cleanup hitter! Better than some might think... doesn't take much for a catcher. However, I decided he shouldn't count since FanGraphs WAR doesn't currently incorporate catcher defense. If you do include one of the available measures for 2009 catcher defense, Molina's clearly below average. Don't worry Giants fans, your boys aren't totally out of it...]
3) Aaron Rowand (0.7). I don’t think this is what the Giants were looking for when they signed him. Not horrible or great in any one area — slightly below average as a hitter, slightly above average as a defensive center fielder.
2) Jimmy Rollins (0.3). So close! Rollins had a dreadful year with the bat (although he rebounded in the second half), and a down year in the field (for him) according to UZR, but for a shortstop, that’s good enough to be a league-average player. I’ll be posting more on him in a few days.
1) Ryan Ludwick (0.0). At 18.1 RAR and with 18.1 replacement runs, Ludwick was a perfectly average positional player of 2009. While he was slightly above average as a hitter (5.3) and fielder (1.3), he played a less demanding corner outfield position for -6.4. It’s a far cry from 2008′s 5.6 WAR, and a lesson in regression to the mean, but he was still a bargain at $3.7M. Congratulations on your perfect adequacy, Mr. Ludwick! No wonder you made that cameo on Shaq Vs.
For Average-ness rankings of all qualified 2009 position players, click here.
19
Loved the list before, love it now. Thanks Matt.
You’re two for two with funny, interesting articles. At the risk of succumbing to the dangers of small sample size, it looks like Fangraphs made a strong acquisition this off-season.
devil_fingers = Emilio Bonifacio
Thanks… I think.
Okay, I don’t get it.
Are you saying I have a .611 OPS as a FanGraphs blogger?
No, Bonifacio started the year off really hot. Then…
It was a joke though, you’re doing great.
But only 1 for 2 in Shaq Vs. references.
I beleive you have to adjust for negative values. Yuniesky Betancurt and Emilio Bonifacio are with the top players.I think is caused by adding two negative values.
He’s taking the absolute value of the difference between “runs above replacement” and “replacement runs.”
So, the two extremes (high above, well below) will appear to be the “larger” numbers. It doesn’t make much difference to this work, as the author was interested in values which tended to zero from either direction.
Your Average is incorrect. It should be about 1 WAR, not 2.
Your site bases its RAR and WAR calculations on a season of 600 PA. That isnt average; that is a ton. That is what you want to give your stars. An average player only ends up with about 300 PA. (this season was 299.3 PA I believe). And when you do the work, you will see this holds true. ML hitters averaged 1.04 WAR in 2009. They have done so each season (with-in points) since 2002.
A player who managed 600 PA wasn’t average at all. He managed to stay healthy and keep a guaranteed position over a full season. That is something only a select group was able to accomplish. 86 total 2009 players to be exact. Fewer then 3 per team, and only about 14% of the non-pitchers in the league this season. On the list of qualified players you used here roughly 70 of them didn’t reach that mark. And if your average was really just what the qualifying players does, then the average WAR you are looking for would be the 3.1 those full-time players provided – not 2.0.
Jason and Laynce Nix (how ironic is that?) are the proto-typical average positional players for 2009. Each had roughly 300 PA of matching 9.9 RAR/1.0 WAR production. While Milton Bradley and Jack Cust represent the average players among starters. They provided production that was equal to the 300 PA of 0 wRAA and 0 UZR that the real average player provides.
Actually, “replacement” is all that is added on in addition to average. When you take “replacement” out, not matter how many PAs a player has, “0″ indicates average.
exactly.
And a replacement player is graded on a -20 runs fewer then the average player per 600 PA received.
Average wRAA = 0.
Average UZR = 0.
Average PA = 300.
So the average player = 0 + 0 + 300 PA of the -20 Runs per 600 PA a replacement level player receives, or 10 Runs overall. That is a 1.0 WAR, not 2.0.
JoeyO:
I’m not sure I understand what you’re getting at. Corrent me if I’m wrong, but it sounds like you’re assuming that WAR is calculated using replacement level as its original baseline, and “builds up” to a +20 run (~+2 wins) plyaer as average.
That’s not how it works. The “replacement” line is added in completely separately. If you leave out “replacement,” Not matter how many PAs a player has, if he’s perfectly average as a hitter, and perfectly average as a fielder (leave aside the positional adjustment issue for now, or assume his pos _ field = 0), he’s exaclty average, no matter how many PAs he has.
Maybe someone else can jump in here and tell me what I’m missing?
Matt Klaassen,
No, I am using Average as the baseline. A true average person is worth 0 Batting Runs Above Average and 0 Fielding Runs Above Average. The average positional adjustment is intended to be 0 (even if the DH adjusts it by about a bit over a run). And the average player gets 300 PA.
Put those exact stats into a fangraphs calculation.
0 Batting RAA + 0 Fielding RAA + 10 Replacement + 0 Positional
The outcome for the perfectly average player is 10 RAR, or a 1.0 WAR.
By saying average is 20 RAR or 2.0 WAR (well, 18.Something because of the DH), you are saying +20 Runs Above Replacement and +2.0 Wins Above Replacement. Keeping the Batting RAA, Fielding RAA and Positional Adjustment all at 0 means the entire +20 has to be in the Replacement field. But if you put a +20 Replacement in you are saying “kept a replacement level player off the field for 600 PA”. You cant do that if you only get 300 PA yourself. If you are average you can only keep him off the field for the 300 PA you get, and that is worth +10 runs, not +20 runs. So you cant be average and be 20 RAR/2.0 WAR unless you managed 600 PA; and you arent average if you got 600 PA. That is their problem.
Truly being average means 0’s across the board and 300 PA. Doing so leaves your equation looking like the one I provided above, with an end result of 10 RAR and 1.0 WAR. “Average” = 10 RAR or 1.0 WAR.
Hahaha, it dawned on me what is going on here. He is referring to average as a player who has his Batting, Fielding and Positional Adjustment get as close as possible to 0. The fact that they all seem to end up in the 18.Somthing RAR range is coincidence, not design – they just happened to have a similar amount of PA. So this really means “closest to neutral value for his respective position” and not “represents the average MLB player” like I was replying to after seeing the RAR’s all ending up at 18.Something.
Guess we were both right, I was just at the wrong house :)
(but why didnt you just say “he is trying to make 0 between the Bat+Fld+Pos fields” in a clearer way? Hahaha)
Did furcal have a better overall season than rollins did?
Yes, mostly due to Jimmy essentially being Clint Barmes at the plate. (seriously, .245/.294/.440 vs .250/.296/.423), and Furcal being better defensively this year.
I kinda feel like a player’s average-ness should be calculated by summing the abs. value of each component (batting, fielding, positional), and just disregarding the replacement level. That would yield the player with the most universally average skillset, versus a player like Ludwick, who is kinda above average in hitting, but below average in defense at a discounted position. I realize this eliminates a lot of players simply by virtue of position, but I guess I’d argue that’s kind of the point.
And it would appear the Bob-adjusted Average-est POY is Jose Lopez, at a staggering 3.9 abs. runs from average ( .1 B, 2.9 F, .9 P). I’d say that’s pretty overwhelmingly average, but he only comes in at 30th in the original list. Oh well. He’s followed by Mark DeRosa and Melky Cabrera
clint barmes as bad as atkins at the plate? check HR totals.