Under the Radar: Deserving MVP Candidates
As the battle rages on between Jacoby Ellsbury, Jose Bautista and Curtis Granderson, or Matt Kemp, Justin Upton and Joey Votto, I felt it was necessary to take a peek at a few players who have flown under the radar in their respective league’s MVP balloting — at least on the basis of how much press they’ve received.
Ian Kinsler – Texas Rangers 2B
Kinsler has been rock-solid at the keystone for a Rangers club that has had a bit more trouble putting away the American League West than most expected. Nonetheless, with a pair of former studs having down seasons – namely, Chase Utley and Dan Uggla – Kinsler has stepped in to join Dustin Pedroia at the top of the ranks of the WAR leaderboard, checking in at 6.7 with just a handful of games remaining. In terms of value to the team, Kinsler certainly has it, as he’s played in all but six of his club’s games, and leads the league in plate appearances. This has been pretty noteworthy, considering Adrian Beltre, Nelson Cruz, and Josh Hamilton have each missed 30-plus games for various reasons for the Rangers.
Alas, Kinsler’s MVP candidacy is hurt first and foremost by his .253 batting average. There are certainly enough Triple Crown-worshipping scribes in the BBWAA that will knock Kinsler down a few pegs based on a statistic that weighs singles and home runs as equal. Similarly, Kinsler’s 72 RBI tie him for 33rd in the American League, which will no doubt make Ian a prisoner of circumstance, rather than a non-entity when it comes to driving teammates home. Finally, those writers who have progressed to the OPS-side of the aisle will likely see Kinsler’s 826 OPS as little to write home about, both in the context of his career (pales in comparison to his ’08 season) and his league (30th among AL hitters with 200 or more plate appearances). Still, it’s far and away the best season of Kinsler’s career; one that saw him finally combine an excellent season at the dish with one with the leather, and for that Ian should be commended. It’s been a good year for Ians on the baseball field.
Ryan Braun – Milwaukee Brewers LF
It truly is inexplicable why Braun isn’t garnering more attention in the NL MVP race. He falls in line with a few of the standard, if antiquated prerequisites. First of all, he’s hitting .331, good for second place in the Senior Circuit. Second, he’s closing in on 30 home runs and 100 RBI, both of which tend to be looked upon very favorably in final balloting. Finally, despite the club’s best efforts to do the contrary, it’s looking pretty obvious that the Hebrew Hammer’s Brewers are going to run away with the NL Central for the first time in the club’s history. In fact, this will be the club’s first division title in nearly 30 years, since the 1982 group of Harvey’s Wallbangers won the AL East and lost the World Series.
Even when considering Braun in a more modern light, he’s still a very legit candidate. Braun paces the NL in wOBA, SLG, and wRC+, and all by a healthy margin. So what gives? Well, I don’t entirely know. The media seems fixated on having three top candidates in each league, which leaves Braun on the outside looking in. Similarly, Braun will never be in the consideration for a Gold Glove award. Still, he’s no Mr. Kemp in that regard either, yet it’s the Dodger whom is garnering plenty of attention for a team that is only still in contention to win as many games as it loses. To summarize, I can’t really say for sure why Braun isn’t a legit candidate. He fills all the stereotypes and preconceived notions that an MVP seems to require, yet falls short on the hype. With a contract that could extend through the 2021 season, I’m guessing he doesn’t mind.
Alex Gordon – Kansas City Royals LF
The final candidate, at least in my view, is more of a feel-good candidate. Once-burdened by the weight of being branded as the “next George Brett”, Gordon has finally come into his own as a hitter, setting career highs in each triple-slash category, home runs, RBI, and doubles, and has done so by a sizeable margin. It couldn’t have come at a better time for all parties involved either, as the crowd of super prospects descends upon Kauffman Stadium, bringing hope to a team that’s been dormant since the late 80s, save for a brief glimmer in the Tony Pena era.
Perhaps the most drastic jump for Gordon has been in his WAR, where he had accumulated 4.7 wins coming into 2011. This season, Gordon has been good for 6.0 – good for 13th among all hitters – while drastically improving his baserunning and defense as well. No, Gordon may not even garner as much as a tenth-place vote in the balloting. After all, the Royals figure to finish fourth in a topsy-turvy Central division, and there still may be apprehension about whether or not this breakout is for real, but one thing remains for sure: Gordon has had one heck of a comeback season. I think Alex will take that 2011 Comeback Player of the Year award on his mantel any day.
Honorable Mentions – Ben Zobrist, Shane Victorino, and Miguel Cabrera.
In addition to Rotographs, Warne writes about the Minnesota Twins for The Athletic and is a sportswriter for Sportradar U.S. in downtown Minneapolis. Follow him on Twitter @Brandon_Warne, or feel free to email him to do podcasts or for any old reason at brandon.r.warne@gmail-dot-com
Strawman on Braun. I’ve seen him mentioned in many places.
I was under the impression Braun was getting attention and Votto was mostly being ignored. Maybe I missed something?
I agree. I haven’t seen anything so far this year (prior to this mention) that had Votto anywhere near MVP contention. If the Reds were reversed with the Brewers in the standings, I think there would be a lot of people making the case for a repeat MVP for Votto.
I’ll third that. I’ve seen a ton of talk about Braun and zilch on Votto.
Ya Braun seems like hes getting as much talk for mvp as anyone in the NL. Upton i the guy never hear anyone talk about
My mistake there fellas. My perception of that may have been skewed. I was under the impression that Kemp and Upton were getting a lot of pub (along with Halladay), and that Votto was also in that mix as the defending MVP.
Sorry for any mix-up. Must not be reading the same places as everyone else. 🙁
I thought Prince Fielder was getting lots of pub for MVP?
Like the Zobrist H.M., but he goes completely cold for weeks at a time.
Are there any players that don’t?
Albert Pujols, save the beginning of 2011. He may be the only one.
miggy
I’ve noticed that. It’s hard not to love his versatility, but also hard not to be bewildered by that inconsistency.
My perception is that Braun and Upton are the top candidates in the NL. Actually, I think the most underrated is Joey Votto and Kemp is in a situation similar to Bautista.
How impressive, Lester. I’m dazzled that you read so much about baseball.
Nice to see some Kinsler love in the MVP discussion. There are obviously better candidates, but he deserves more than the zero ink he’s getting. Kudos.
Thanks. He’s been excellent this season and deserves every bit of it.
Meh. He’s been good, not “excellent”, or MVP-caliber. Too much of that WAR is in the defense. And he simply hasn’t hit that great.
You can make a case for “Team MVP”, however.
You could counter this argument by pointing out that with a modest BABIP increase from his .242, even if you took away a good portion of his defensive value, he would look like a very palatable candidate. Not sure why its fair to take points away because he’s overachieved in one statistically volatile aspect of his performance without giving him credit for the one in which he’s so significantly underachieved. Give him a .260-.270 BABIP and he’d be hitting something like .275 / .375 / .500, .385ish wOBA and have a shot at finishing 30-30 (may fall a few short on SBs). Take 5-7 defensive runs away from that and he’s still a 7ish win player, but the amount of extra MVP consideration he might be getting under that circumstance is disproportionate to what amounts to an essentially non-existant value difference.
Mark, look at his pop-up percentage. That is why his BABIP is so poor. It’s not bad luck. Just has a huge upper-cut. Also why he hits a lot of bombs.
His pop up percentage is lower than his career average….
Yeah IFFB% certainly does not explain the low BABIP, there’s very little chance his .242 BABIP is not luck deflated. And its not as if his big defensive year is an outlier either, it may be a bit inflated but he’s basically at the same defensive level he’s been at for the last few years.
Another name under the radar in the NL is Hunter Pence. I’m not talking about his rise with the Phillies, but the whole season. He was lost on a terrible ballclub in Houston and its good to see him on a contender where his talents actually matter. Not saying he should win it, but he deserves consideration.
No, not really. These fringe-ish guys above are 6 WAR guys, Pence is at what, 3.5, 4 WAR?
A good year for Ians? Tell that to the guys who drafted Ian Stewart on their fantasy teams.
I don’t know if I’ve ever talked to such a guy.
What about Miguel Cabrera? Currently 2nd in the AL in OBP, OPS, wOBA and RC+ and if you believe in being “clutch” he is 2nd in WPA and leads all hitters in batting with RISP. Obviously the fact that he is a 1B and is sub par defensively hurts his value but for all the talk of Verlander carrying the Tigers this year, I don’t think I’ve heard anybody bring up(atleast on a national level) how good Cabrera has been this season. With that said I don’t think he deserves the MVP but I think he is certainly deserving of atleast getting some consideration.
Never mind, just saw you gave him an “honorable mention”, completely missed that the first time I read it.
If you don’t think a player should win the award, then they shouldn’t be “in the conversation”.
Using what logic?
Saying “I don’t think he deserves the MVP” means you don’t think he should win it, and therefore he shouldn’t be mentioned as an MVP candidate to begin with.
Yeah I worded that wrong, I guess I shouldn’t have said he didn’t deserve it, but instead said that he wouldn’t be my choice. And the reason I think he should be in the conversation is because even though I wouldn’t pick him I think his numbers are good enough that maybe somebody else would.
In the interest of full disclosure:
If I went back and re-did the column, I’d probably put Miggy C in there instead of Braun.
I think a lot of very casual baseball watchers think Kemp is excellent with the glove in centerfield, when the metrics and scouts would probably disagree. Even hardcore baseball fans are probably not watching many Dodgers games this year, so just automatically assume Kemp is a Gold Glove-caliber player out there. So he is probably getting huge credit for something that is not really true. There is a strong tendency for writers to assume every tall, athletic-looking African-American that plays the outfield is just a superstar on defense. Not all of them are Willie Mays…
Depends what defensive metrics you’re looking at. Kemp may not have range, but his outfield assist numbers are very good.
This is just silly.
If any second baseman in the AL deserves MVP it is Cano. He really has produced more runs and therefore provided more value than Kinsler (and Pedroia). The RBI and runs scored are a record of what actually happened, not what you think should have happened.
Until the Rangers start getting credit for the runs you think Kinsler should have created cetris parabus, he simply hasn’t been as valuable as the strangely underrated second baseman in the Bronx.
Are we really talking about RBI and runs scored on Fangraphs? Cano has a very slight edge on Kinsler in wRC+ and wOBA despite Kinsler’s .243 BABIP. Take into account Kinsler’s superior defense at second, and he’s clearly been better.
damnant quod non intelligunt
Under that logic, Ryan Howard is better than Albert Pujols and Joey Votto this year.
The above comment was directed at Jason, not Mr. Jones.
It is clearly the case that “better” does not necessarily mean “more valuable”. Even if you think Kinsler is better than Cano (I think he’s not even close, but that is a separate issue), he has not been more valuable (either due to luck or opportunity, or skill when it counts). Anyway you cut it, Cano has outproduced Kinsler in what the team finds valuable (runs).
There is no evidence which player is a better fielder one way or the other, so I am only considering Cano’s superior offense.
you mean how Kinsler gets on base more than Cano and scores more runs (which ultimately is the job of a baseball player) means Cano is still better? Kinsler has better power, OB skills, defense and baserunning. Cano only beats him in average, not to mention Kinsler has put up immense value while being one of the unluckiest hitters in the majors. So tell me, when did batting average become so valuable?
I just took a look and it appears at a glance that Ryan Howard is not more valuable than either Votto or Pujols this year. He just doesn’t get on base enough or run fast enough to score many runs. …plus those 163 strike outs are horrific. But their values look pretty close this year.
Now if we are talking about skill level than I put Howard way below Pujols and also well below Votto.
YP,
The job of a batter is to score runs and drive runs in. The small advantage Kinsler has at scoring runs is more than offset by his vast inferiority at driving them in. The reason for this is Cano’s superior power (much greater slg%) and his ability to get on base with hits rather than relying on walks (Cano has outpaced Kinsler in hits and extra base hits by far). That is how you drive in runs. Cano is just better at it.
I really like Kinsler. I think he’s the second best 2b in baseball.
Kinsler is a leadoff hitter and is 2nd to all leadoff hitters in RBIs. His ISO Power is about the same as Cano’s so their actual power is roughly the same, not to mention Kinsler has a higher OBP despite hitting 50 points lower.
YP,
I can’t help where Kinsler bats in the lineup. Maybe he would be more valuable if his team utilized him better. I don’t know.
Cano hits 5th and leads all of baseball in RBI for players that bat 5th. He might lead all of baseball in runs scored for players who bat fifth too. If he hit 4th or 3rd he’d probably have more RBI and runs scored yet. But he doesn’t.
MVP is for what actually happened, not what you think would happen all things being equal.
You are overvaluing walks.
Completely agree. Cano has the best season out of any 2B in the AL. Runs are a useless statistic, bu the RBIs are pretty legit. Kinsler and Pedroia both play on top-3 teams in Runs Scored, so they can’t blame a poor offense on inferior RBIs number
And Kinsler’s defense is not any better than Cano’s. UZR is a very immature statistic (literally), and cannot be used to determine defensive skills
Wow…there is so much wrong with this post. I watch most Rangers games…there is almost no play Kinsler doesn’t make at 2nd.
Runs scored are not a useless statistic when we are talking about value. The Yankees certainly find Curtis Granderson’s crazy number of runs scored to be valuable!
It is a limited statistic when we are talking about skill, however, since often times, runs scored are not independent of other people’s performance.
With awards like the MVP, we want to recognize what actually happened, not what you think should have happened independent of all game conditions.
We don’t determine the winner of games by adding up team linear weights at the end of games. Sometimes the losing team outhits the winning team. We accept that.
Sometimes a less skilled player outhits a more skilled player. I accept that.
Try watching any Ranger game and you’ll see Kinsler turn an amazing double play. And the reason Kinsler doesn’t have the most RBIs is that he is a LEADOFF hitter!
Jason, I happen to disagree with you about how the word “valuable” in “Most Valuable Player” should be interpreted (I’m in the “Most Outstanding Player” award camp), but leaving that aside and taking your view…
Even if you don’t think they’re as flawed as the typical fangraphs-ite does, surely you do see that runs scored and RBI are still only an approximate measure of value to the team (even if we give a batter all the credit both for his R and for his RBI). Some runs are scored (or driven in) in blowouts, others in close games.
Given this, how do you feel about Win Probability Added (WPA) as a statistic for your version of the MVP? I know it’s SABR-y, but based on the arguments you’re making for Cano, it seems to fit far more closely with your idea of “value to the team” than RBI + R.
Squelchtoad,
That is an interesting idea and would be interesting to look at. However, I suspect it would be biased too heavily towards players on marginal teams that play a lot of close games. I don’t see any good reason for this.
I actually think teams want to win in blowouts because it generally makes the next game easier. For example, when the Yankees win two close games in a row, then play a third close game they often will not have one or two of their three top bullpen pitchers available. The Yankees were recently crippled in this way against Baltimore and they dropped two extra innings games in a row to them.
So I am not in favor of penalizing a player for contributing to a blowout since it might contribute to a win tomorrow.
“would be interesting to look at”
Nothing easier: http://www.fangraphs.com/leaders.aspx?pos=all&stats=bat&lg=all&qual=y&type=3&season=2011
As you can see, it actually tends to favor guys on good teams, rather than guys on “marginal teams” (this makes sense more wins –> more win probability added by the team as a whole over the course of the season –> more WPA for its individual players).
As for blowouts, sure, wearing out a staff has value. That’s why patient offenses are valuable. But that’s a very small effect. It has a lot, lot less value than making sure your team wins the game at hand. Don’t try to tell me that Cano driving in a run in a 9-1 game does more for the Yankees’ playoff chances than Cano driving in a run in a 3-4 game. And that’s not even considering runs driven in (or scored) in a LOSING blowout. If Cano homers during the 8th inning of what ends as a 3-hit complete game for Justin Verlander with the Tigers leading 9-0, that’s not doing a heck of a lot for the Yankees’ chances of winning either the current or the next game against the Tigers (let alone their playoff chances).
Squelchtoad,
As I suspected (and contrary to your assertion), it seems to be biased towards the marginal teams with run differentials around zero. More specifically, it seems to be biased towards teams that win despite not having large run differentials (marginal teams). This makes sense. If your team has huge run differentials, you will have more “meaningless” at bats throughout the course of the year. The two teams with top run differentials and the most wins don’t have a player in the top ten.
Obviously there are situations where runs are less valuable than others such as the ones you describe. What about runs that you deem to be very valuable because they greatly increase your chances of winning the game, but then you lose none the less? I don’t have an answer I’m just curious.
“The two teams with top run differentials and the most wins don’t have a player in the top ten.”
That’s because they (especially the Phils) have depended more on run prevention rather than run scoring to produce their differential (relative to, say, the Red Sox). I neglected to post WPA for pitchers, but if I had, you would have seen big numbers for all of the Phils’ big three, as well as for Sabbathia. The Phils are #1 in MLB in total pitching WPA, the Yankees #3 (Atlanta is #2).
But maybe you think it’s biased against the *stars* on top teams. I think that’s empirically false as well. Here are the top 5 MLB teams in run differential. I don’t think any of them qualify as “winning without a large run differential.”
1. PHI
2. NYY
3. BOS
4. ARI
5. MIL
Together, they make up exactly 1/6 = ~16.7% of all MLB teams. Together, they have 3 of the top 10 (30%) and 7 of the top 20 (35%) of top WPA batters. They have 7 (!!!) of the top 10 pitchers (70%) and 9 of the top 20 (45%).
In short, players on teams with big run differentials are likelier than average to show up on the WPA leaderboard.
Now, I personally do think it’s *valuable* when a player does all he can to help his team win even if it doesn’t eventually do so (“don’t demean Cano’s go ahead HR when Rivera blows a save”), but if you don’t, you surely see that that, too, is far *more* of a problem for R+RBI than for WPA. Teams don’t make huge “win probability” comebacks all that often (there’s a reason it’s called “win probability”), so it’s actually pretty rare that a player can accumulate monster WPA in a losing effort. On the other hand, teams often bring in crappy pitchers when they’ve got a big lead, so it’s easy to score (and drive in) lots of runs in a LOSING blowout.
Btw, if you’re interested in “doing all you can in a losing effort,” you might enjoy this: http://www.baseball-reference.com/blog/archives/10759)
Squelchtoad,
I don’t know where you got that list of the top run differential teams, but its wrong. The real list is as follows:
NYY +207
PHI +183
BOS +156
TEX +139
Those are the only teams above 100. The rest are pretty marginal over the number of games they’ve played so far:
TAM +80
DET +61
ATL +60
MIL +59
ARI +53
STL +53
Whoops. Well, in that case, the top 4 teams make up ~13.3% of baseball teams. They possess:
2 of the top 10 batters (20%), 5 of the top 20 (25%), and 5 of the top 20 (all in the top 10) pitchers.
Squelchtoad,
I pulled the data into R and did the correlations between both WPA and team run differential and WPA and team winning percent. To do so I used all of the 148 players included in the FanGraphs WPA column, however I removed from these all the players traded during the season. Team run differential and winning percentages were as of this evening.
Here are the correlations:
1) WPA and run differential
Pearson’s product-moment correlation
data: WPA and DIFF
t = 2.8548, df = 135, p-value = 0.004987
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
0.07384541 0.39068679
sample estimates:
cor
0.2386058
R^2 = 0.057
There is a weak but positive and significant correlation between WPA and team run differential. However team run differential only explains 5.7% of the variation in player WPA.
2) WPA and team winning percentage:
Pearson’s product-moment correlation
data: WPA and PCT
t = 2.803, df = 135, p-value = 0.005808
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
0.06953935 0.38701294
sample estimates:
cor
0.2345197
R^2 = 0.055
There is also a weak, positive, significant correlation between WPA and team winning percentage. Team winning percentage explains only 5.5% of the variation in player WPA.
Additionally, I plotted both of the correlations and they both look like clouds.
CONCLUSION:
We were both about equally wrong. Though we were both correct in predicting the correlations, the correlations are so weak as to be effectively meaningless relative to what we were arguing about.
LESSON: It pays to do the maths!
Squelchtoad,
Ugh. I was thinking about it wrong. To test my hypothesis I should have been correlating WPA with the absolute value of run differential since I am predicting that the greater your average game differential, the lesser your WPA (negative correlation).
Consequently I correlated the absolute value of run differential with WPA and produced the following results:
Pearson’s product-moment correlation
data: WPA and Abs_Diff
t = -0.0481, df = 135, p-value = 0.9617
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.1717353 0.1636901
sample estimates:
cor
-0.004139064
R^2 = -1.713185e-05
CONCLUSION: I was totally wrong and you were less wrong.
LESSON: It still pays to do the math.
Thanks for running the numbers, and for being willing to alter your views.
Alex Avila
Adam Dunn!
If you’re going to mention Ian Kinsler, you’ve got to honorable mention Howie Kendrick, who has been almost as good on a much worse offense.
But Kinsler is a good example of why I remain skeptical of WAR, especially fWAR, which seems to over-value defense and position. I have a hard time believing that Kinsler has contributed more runs to his team than Troy Tulowitzi, Ryan Braun, Adrian Gonzalez, Miguel Cabrera, and Albert Pujols, to name a few.
his defense has been stupendous, he plays the 2nd hardest position on the field and he’s going to possibly have another 30/30 season (unlikely but close) not to mention he walks more than he strikes out. That’s value.
Why? As a Red Sox fan, I have no problem with believing Pedroia has helped the team win more than Adrian Gonzalez. Simply put, hes just a better player.
I see no reason why it couldn’t be the same with Kinsler.
If defense weren’t that important, you’d have guys who look like AGon all over the diamond.
I’m not saying defense isn’t important, just (perhaps) not as important as fWAR makes it out to be. If you were building a team from scratch and, all things being equal (salary, etc) would you really rather have Kinsler (6.7 fWAR) than Tulowitzki (6.5), or Kendrick (5.9) than Miggy Cabrera (5.8)? Of course we have to look at their careers, ages, performance trajectories, and so forth, but even as an Angels fan I can’t see how Howie Kendrick (.291/.342/.475, .356 wOBA) is having a better year than Miguel Cabrera (.332/.435/.560, .421 wOBA), even though Kendrick (17.6 Fld) is a much better defensive player than Cabrera (-5.0).
If I understand WAR correctly, one of its utilities is that you get to ask questions like: “If you placed X player in a lineup of average players for their position, how good would they be comparatively?” So the question would be, if you placed Howie Kendrick as the 2B on a team of exactly league average players, would that team be fractionally better than a team of Miguel Cabrera and average players? Obviously their fWARs are so close that the teams–according to “WAR Theory”–would be virtually identical, but I can’t shake the feeling that the Miggy teams would win more games.
Baseball Reference’s WAR has them quite different: Cabrera at 6.0 and Howie at 4.4. To me that intuitively seems more accurate, even if I’m pulling for Howie as an Angels fan!
Think of it this way; if Miquel Cabera got hurt you could find a way to find a decent replacement (obviously not Miggy caliber but a decent hitter). If Kinsler got hurt finding a decent hitting replacement would be much much more difficult
Yes, I get that — but that seems more true in fantasy baseball than in real life. I guess the question is whether the difference between Cabrera and Kendrick and league average (or even replacement) players at their position is the same, as fWAR suggest, or different, as rWAR suggests. I find that time and time again, rWAR seems to correspond with what my intuitive sense says — especially when looking at numbers other than just WAR. This goes for pitchers equally well – last year Cliff Lee led the majors among pitchers with a 7.1 fWAR, with Halladay second at 6.6. I’m assuming this has to do with Lee’s much lower FIP (2.58 to 3.01), despite Halladay’s superior ERA (2.44 to 3.18). Lee also had a LOB% of 67.9% to Halladay’s incredible 82.7%. So you could argue that Lee was actually better because he wasn’t as lucky as Halladay.
But my view is that you can’t just look at WAR, whether fWAR or rWAR, to explain everything. All things tolled, Halladay was a better pitcher last year (IMO) – better in that the end results were that he prevented more runs from scoring. I’m not so simplistic as to say that all that matters are RBI and Runs, or Wins and Saves, but that the “truth” is somewhere in the middle between uber-advanced sabermetrics and traditional stats.
One more note about Kinsler. We shouldn’t forget that his home/road splits are extremely lopsided, both for this year and his career. Does that factor into whether he’s an MVP candidate or not, if a player is actually poor (a .291 OBP) for half of his games (on the road)? Home/road splits are another dubious stat – but again, my point is that we can’t rely too heavily on any stat but must look at the whole picture. In that sense, I think there is a compelling argument that Howie Kendrick has been equal or better than Kinsler.
While the year has been kind to Ians Kennedy and Kinsler, if your name is Ian and your last name does not begin with the last name K, it has been abysmal – in addition to Stewart, Ian Desmond has also struggled.
I could go back and edit, but I don’t wanna. 🙂
If Kinsler had an even neutral BABIP instead of the insane bad luck he’s had (dude hits a line drive to a fielder seemingly every game), he’d be the MVP frontrunner IMO.
He’s hitting the ball in the air well over half the time. If you’re going to have a low BABiP, that’s one way to do it.
his Flyball rates are in line with most of his career other than last year (way more GBs) and 09 (a ton of flyballs which his BABIP made more sense). His BABIP should be much closer to 280ish. His babip on grounders, line drives and fly balls are way below his career averages. You don’t explain all that just by saying he hits fly balls especially when his IFFB% is actually slightly lower than his career averages.
Quick…what’s his ShHA!!?
troy tulowitzki says hello
How is he? I hear the weather in Denver is great this time of year.
wow, a LOLCOORZ comment coming from an FG writer? shouldn’t you know better than that, or are you just a huge douche? because if you didn’t know, his H/R splits are .407 wOBA 151 wRC+ at home against .376 and 130 on the road.
go ahead, try and tell me the guy tied for 4th in the league in WAR and within a half-win of the leader isn’t the friggin definition of an “under the radar mvp candidate.”
“I like this team and I think this player from this team is the best. Because I like this team.”
Who do you think will end up with the NL MVP this year? And if everything stays the same, what are Upton’s chances!
“Still, he’s no Mr. Kemp in that regard either”
Matt Kemp is a below average CF. Braun is a well below average LF. I would hope we’re not discounting Kemp for his defense more than Braun, even after the positional adjustment.
I’m a huge Yankees fan, but anyone saying Cano is better than Kinsler is out of their mind. This year at least. That said, Cano MIGHT have more talent. But that won’t matter until he stops swinging at everything. Last year he had his highest BB% and “coincidently” (not) had his best year. I don’t know why he couldn’t carry that into this year. While it is fun to watch him drive pitches well out of the strike zone, it makes me crazy when he won’t make pitchers prove they can throw a strike. As for his fielding, I’ve almost given up on him rediscovering his random 2007 pretty-decent-amount-above-averageness.
Despite his poor approach at the plate, Cano is still an elite hitter. He just gets so many hits and extra base hits that he more than compensates for the much less valuable walks he is foregoing. It might be a short career for him if he can’t maintain his ability to drive the ball consistently into his old age. However, right now, Cano is the best hitting second baseman in baseball.
Random is right for the fielding metrics. Don’t pay attention to them. The error in the measurement is likely greater than the differences they are trying to measure.
Why exactly are walks so much less valuable? You don’t make an out, you make the pitcher throw more pitches, and you’re still on base.
Jacob,
Walks are less valuable than hits because they are less likely to lead to a run than even singles are.
I’m thinking Kinsler’s low RBI’s and low batting average are related, you can knock in runs with singles too.
What the ???? Kinsler’s WAR on Baseball Reference is 4.8 and 6.7 here??? Are the calculations really that different???
It is largely differences in the random number generator stuck on the end of WAR (i.e. the fielding metrics).
“Hi, my name is Jason and statistics that don’t agree with my point of view are stupid.”
Jimmy,
Sometimes fielding metrics agree with my point of view and sometimes they don’t. My reason for ignoring them all together has nothing to do with this. I ignore them because there are many parameters that go into them and they all have an associated error. Because of the nature of the data used in fielding metrics, the error is likely to be large. In fact the error may be as large, or larger than the differences they are trying to measure. We don’t know the magnitude of the error involved in fielding statistics because they are not calculated. Until they are, no one should pay them any mind.
Jason,
That’s not even true actually. At least in Kinsler’s case, the fielding numbers on B-R essentially agree (7 runs on B-R vs. 10.5 runs on fangraphs). The difference is actually in the offensive valuation. Kinsler gets just 13 rbat on B-R and 24 batting runs on fangraphs, which is where the vast majority of difference comes from, so its a difference in the offensive linear weights systems, not so much the defensive valuation.
I also understand what you’re saying about fielding metrics, but when they a) largely agree and b) have for many years, they have value. Fielding can’t be simply ignored, its still a significant part of player value that its at least as ineffective to simply ignore it as it is to recognize flawed systems. When the systems agree, its absolutely worth including in a discussion, and with Kinsler they do.
Mark,
Thanks for correcting me about the defensive numbers not being the cause of the discrepancy between the two valuations.
It is incorrect that should accept the defensive metrics “when they agree”. Either you accept them when they agree and when they don’t agree, or you reject them in either case. A random number generator will also produce results that sometimes agree with each other.
I agree that defense is important and needs to be considered. I am just not willing to do so with a number denuded of error bars when I suspect the error bars to be huge.
Do you think a UZR of 10 is really better than a UZR of 7? How do you know? It’s like accepting a poll without being given the margin of error.
I’m not arguing for them to be treated as precise measurements, I am saying they shouldn’t be treated as “random number generators” though. The commonly accepted idea is that you need three full years of data to accurately gauge a player’s “true talent” defensively based on numbers like UZR and DRS. Well, when both systems essentially agree over three concurrent years and there’s only marginal variation within the three years, you can make an assumption. The assumption may lack precision, but the error rate is tolerable enough to make a value judgment. Is it possible the congruence is random, but when a system isn’t designed to be random, the higher the degree of congruence the greater the likelihood the result is legitimate and not random.
Its not all that different from trying to make a prediction based on a more precise descriptive statistic–the larger the sample the more likely you are to make an accurate prediction. The only difference here is we’re creating a valuation, not a prediction, and for some reason it seems to peeve people when a valuation isn’t strictly precise, even though absolute precision isn’t really necessary in developing judgments as long as we’re willing to be flexible in said judgment and not treat it as absolute truth (which, even with a high degree of precision, such judgments often aren’t anyway, if ever).
There was an example that this discussion reminds me of in the book “Black Swan” where the author asks if a coin is thrown 100 times and lands heads each time, what are the odds that it will be heads on the 101st throw? The scientist’s gut reaction is that its still only 50%, but unless it can be explicitly assumed that this is a truely balanced coin, it may actually not be 50% at all, it may be far more likely that the 101st throw will be heads, since we don’t truly know that the flips are random. It could be a loaded coin, or some kind of fake coin that features heads on both sides. There is a point at which this possibility becomes more likely than the possibility that it is a “fair” coin when observing enough throws. Is 100 heads in a row possible with a fair coin? Sure, but its very unlikely to happen with a fair coin.
The problem logicians tend to have with arguments like this is that they are examples of the “affirming the consequent” fallacy, but consequent affirmation simply isn’t a form of deduction, it can still be valid evidence for induction, and specifically the type of induction that the science of statistics is based on. If 100 throws isn’t enough, how many would it take to convince you that a coin was not fair? Would 1,000 be enough? 1,000,000? These outcomes are also technically possible based on random variation, but the greater the sample, the stronger the evidence becomes that the coin is not, indeed, fair.
Over a single season, I wouldn’t treat a UZR of 10 and a UZR 7 all that differently, but if I had five years of data where one player rated at 10, 11, 8, 13, 9 and another where a player rated 7, 5, 8, 5, 3, I would be pretty confident the former was the superior defender by 3-5 runs per year or so.
Mark,
We can test the hypothesis that the coin is fair. We don’t have to guess. A fair coin is described by a binomial process with probability of 0.5. The probability that a fair coin never shows heads in 100 flips is effectively 0. So we reject the hypothesis that it is a fair coin. The probability that a fair coin never shows heads in only 10 flips is .001. So even with ten flips we can reject the hypothesis that the coin is fair.
Similarly with your UZR examples, we can test if player A is greater than player B. I’ve used a simple T-test to compare your hypothetical that A is greater than B (one tailed):
Welch Two Sample t-test
data: a and b
t = 3.7559, df = 7.999, p-value = 0.002789
alternative hypothesis: true difference in means is greater than 0
95 percent confidence interval:
2.322476 Inf
sample estimates:
mean of x mean of y
10.2 5.6
Given the numbers, there is only a 2.8% chance that player A does not have greater UZR than player B. So we accept the hypothesis.
You see, these are things we can calculate. We can also calculate the error in UZR. We can do this! Until we do, I see no reason to take it seriously. We don’t have to guess! Why do we insist on guessing when we don’t have to?
Jason,
I’ve spoken with a number of people about some of this over the last day. They are average fans of both the Yankees and Red Sox. Each person agreed that if you put Ian Kinsler, Robinson Cano, and Dustin Pedroia into a lot and redistributed them among the Rangers, Sox, and Yankees in any order then they would be happy.
I think there are compelling arguments that all 3 could be the best 2b in baseball. When you say that Cano has the most RBI of any player batting fifth its all about scenarios. If you put Kinsler in the Yankee lineup in that same spot he’s putting up similar counting stats, though definitely not the same batting average, Cano is the better pure hitter. If you put Cano in the Sox lineup behind Ellsbury, Pedroia, Gonzalez, then by your math, his numbers would skyrocket because those three have a much higher batting average than Jeter, Granderson, and Teixeira. They might also have a higher combined OBP, Im on my phone and not doing the legwork.
The only reason Kinsler isn’t leading his batting slot in RBI is because of an MBPS caliber year from Ellsbury, who is the recipient of a deeper Red Sox lineup that has higher quality hitters lower in the order.
Or course we don’t control their batting order spots, but its something you have to marginalize, to a degree, when evaluating worth.
Jimmy,
I generally agree with you with respect to player abilities. However, this isn’t, in my mind, the same as “value” with respect to MVP voting. I think with respect to value we have to look at what actually happened.
I’ll 2nd Alex Avila. With VMart hobbled, Avila has started the vast majority of games at catcher this year and continues to put up big numbers offensively. Take him out of the mix and the Tigers replace with a minor league catcher.
In 5 years we’ll all be laughing at WAR (in today’s version) like we do at Wins, Saves, and RBIs.
Ryan Roberts checking in