A second look at Ian Kinsler

Statistics current through Sept. 13.

People all around the world of fantasy “hate” Ian Kinsler. It’s time for that to end.

In the preseason, I received a lot of criticism for ranking Ian Kinsler as my number two fantasy second baseman for 2011 behind Chase Utley, who was still healthy at the time. Some critics legitimately focused on the level of risk imposed by Kinsler’s health; I thought he was a top five or 10 second baseman even if inundated with injuries. For what it’s worth, Kinsler has appeared in 142 of the Rangers’ 148 games this season.

Other critics focused on his raw stats, claiming that Kinsler had never proven himself on par with the likes of Robinson Cano. In reality, as I pointed out, the two players rated very similarly over the past four seasons:

Name GP AB AVG HR SB R RBI
Ian Kinsler 498 2058 0.280 78 95 372 263
Robinson Cano 640 2477 0.305 87 14 369 363
Kinsler prorated 640 2645 0.280 100 122 478 338

Despite a clear differential in games played, they have comparable absolute home run and runs scored production, with Cano having a noticeable edge in batting average and RBI. Kinsler owned a huge advantage in stolen base totals. Cano had a substantial lead in games played over Kinsler.

Health is obviously a factor of value, but when you consider how close the two are in value even when one is constantly making trips to the disabled list, comparing their rate stats is not irrational. If we prorate Kinsler’s numbers to the same number of games played as Cano, Kinsler produced the better rate stats overall. Hence, my claim that Kinsler’s ceiling and floor were high enough, despite an ankle-injury-derailed 2010, to warrant top three consideration.

Eight months later, my widely panned ranking has turned out to be more or less correct. Kinsler has been the third best second baseman this year per Yahoo (second best, per my Z-Score sum calculations), behind Cano and Dustin Pedroia. Over the past 30 days, Kinsler has been the No. 1 fantasy second baseman. He has also been an elite player overall. So far through 2011, Kinsler’s season qualifies him as a top 30 overall fantasy player (No. 28, by Yahoo’s calculations) and top 10 fantasy hitter (via my Z-Score sums). Over the past 30 days he’s been even better, and Yahoo values him as a top 10 overall player (No. 7) over that period.

Cano (Yahoo’s No. 14 ranked overall, my No. 9 overall hitter in baseball via Z-Score sums), who I ranked as my No. 3 second baseman, has been the better fantasy baseball player. Thus, my argument of Kinsler over Cano was incorrect…right?

In terms of their results, yes. Objectively speaking, no matter how you slice their numbers on the season (which is honestly all that really matters), Cano has been the better fantasy player. Yahoo has their values pegged much farther apart than my Z-Score sums (below), but Cano’s produced more in the relevant fantasy categories:

Name Team G AB PA AVG HR SB R RBI ZAVG ZHR ZSB ZR ZRBI ZSUM
Robinson Cano Yankees 146 577 624 0.305 26 8 96 111 +1.04 +2.14 +0.20 +2.49 +3.04 +8.91
Ian Kinsler Rangers 142 569 663 0.251 28 25 108 72 -0.23 +2.39 +2.25 +3.01 +1.42 +8.83

Z-Score sums are determined by taking the difference between a player’s value in a given stat from the league’s mean production in that stat, and then dividing by the standard deviation. This gives you a strong measure of relatively in evaluating how good players are in a given statistical category. Z-Scores, for example, are a useful tool for answering the question of which is more “valuable” in a vacuum—20 RBI or five stolen bases. By taking the Z-Score for each Roto category and summing them, we get a Z-Score sum which tells you what each player’s overall relative value is.

To get an accurate Z-Score sum, you need an accurate pool of players. The pool of useful players from which to calculate Z-Score sums will vary wildly from league to league, depending on the number of teams, the number of players per position per team, the format of play, etc. To simplify an election bias, I simply included all players with 100 or more plate appearances to determine the relative value of hitters for this exercise.

On one hand, this inflates the value of players who have more counting stats, as the players near the bottom of the plate appearance threshold —who probably have little fantasy value—drag down the “mean” and increase the standard deviation. At the same time, this effect occurs equally on all players in the pool. Hence, while the means and standard deviations may not be accurate representations of the true fantasy means and standard deviations of any given league, they have a relatively similar effect for rankings purposes. Alternatively explained, though the absolute Z-Scores numbers may be skewed, their relativity should be reliable for our purposes.

Mathematics aside, the point remains salient. Cano has been better, but their value has been pretty close and both have been pretty excellent.

But Kinsler should be doing a lot better—at least if you buy into BABIP-luck theories. So let’s look at BABIP luck-neutralized production of Cano and Kinsler and see who should be faring better.

To account for BABIP luck, I calculated the xBABIP of Kinsler and Cano using THT’s own xBABIP tool. Using each player’s xBABIP, I then calculated the number of hits each player should have expected to produce irrespective of luck. Using this expected hits total, I recalculated the player’s expected batting average.

Then, using the differential between each player’s expected hits (xH) total and his actual hits (aH) and holding each player’s production rates constant, I calculated an expected stolen base, runs scored and RBI differential to be added to/subtractedfrom each player’s actual 2011 numbers. I ignored the BABIP effect on home runs, as those are not balls in play and because I did not want to have to go through each player’s detailed game log to find “robbed home runs” and then also account for “lucky” home runs.

I also performed the same analysis and “number crunching” using each player’s career BABIP for comparison.

Here is each player’s relevant BABIP data:

Name 2011 BABIP 2011 xBABIP Career BABIP
Ian Kinsler 0.242 0.308 0.283
Robinson Cano 0.320 0.324 0.322

Using some mathematical “reverse engineering,” here are how Cano and Kinsler’s BABIP-luck neutralized 2011 stats stack up using their career BABIPs:

Name Team AB H HR R RBI SB AVG ZAVG ZHR ZSB ZR ZRBI ZSUM
Robinson Cano Yankees 577 177.0 26 96.4 111.5 8 0.307 1.1 2.1 0.2 2.5 3.1 9.0
Ian Kinsler Rangers 569 162.7 28 117.2 78.1 27.1 0.286 0.6 2.4 2.5 3.4 1.7 10.5

And using their 2011 xBABIPs:

Name Team G AB xH HR R RBI SB AVG ZAVG ZHR ZSB ZR ZRBI ZSUM
Robinson Cano Yankees 146 577 178.0 26 96.9 112 8.1 0.308 1.1 2.1 0.2 2.5 3.1 9.1
Ian Kinsler Rangers 142 569 174.6 28 122.8 81.9 28.4 0.307 1.1 2.4 2.7 3.7 1.8 11.6

As you might notice, when you strip out Kinsler’s poor BABIP luck, whether you use his career BABIP or 2011 xBABIP, he should have been a superior fantasy producer. In fact, Kinsler’s xBABIP adjusted stats would place him as the fifth overall player, behind Matt Kemp (13.5 Z-Score sum), Curtis Granderson (12.9), Jacoby Ellsbury (12.6) and Ryan Braun (11.6). That’s a hair ahead of Jose Bautista (10.2). Even if we hold Kinsler’s counting stats constant and adjust only his batting average to reflect his career BABIP, Kinsler’s aggregate production this season outweighs Cano’s. As it should be evident, Kinsler is an elite player of the highest order when healthy, even if he did not put it all together in 2012.

Of course, results are all that matter in fantasy, and in that regard, Cano has had the better 2011 season. However, for the future, it is clear that my initial hypothesis—that a healthy Ian Kinsler has top-tier player upside—rings true. For 2012, neither Kinsler or Cano will come cheap. However, Cano and Pedroia will likely cost more, meaning Kinsler, if healthy, could be a top 10 player at a relative discount

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: What’s the effect of the new college bats?
Next: A celebration of Fenway’s centennial »

Comments

  1. Steve Stein said...

    About this “BABiP luck” thing – I’m not sure I totally buy it.  If a player’s BABiP is consistently under .300, don’t you have to start thinking it’s not all attributable to luck, but something about the player’s hitting that’s to blame?  Perhaps he’s a little bit lacking in the “hit ‘em where they ain’t” department?  Maybe he’s a little too predictable in where his batted balls go, to the point where good scouting can reduce his average?

  2. Jeffrey Gross said...

    @Steve,

    A player’s BABIP is unique, and not all are .300. Kinsler’s career rate is .283. hence, the adjustment for both career and xbabip rates. there is regression towards one’s own unique babip, but you are correct in that such regression is not always towards a mean of .300

  3. Eddie said...

    In point leagues which value walks and hitting for extra bases, Kinsler has been the better player this season. I know that’s a small subset and all, but it’s something.

  4. Sean said...

    Jeffrey, as a Rangers fan, I can tell you with reasonable certainty that Kinsler’s low BABIP is indeed a result of his uppercut swing, which leads to a large amount of popups. However, that hitting style works for him, and if he evened out his swing his home run totals would likely plummet.

    I know I’m biased, but in my opinion Kinsler is one of the most underrated players in baseball. The problem from a fantasy perspective is that two of his most valuable assets, walks and defense, don’t show up in traditional fantasy formats.

  5. Jeffrey Gross said...

    @Sean/Ross,

    I do not deny he has an uppercut swing that induces higher popup rates. His season and career rate (just under 11%) is a point or 2 higher than league average. However, even with that considered, his BABIP should be somewhere between .285 and .305

    I honestly love Kinsler. He’s great

  6. Samuel said...

    I love Kinsler. I’ve had him on every team for the past three seasons, even when I had to take him in I think it was the 2nd round? One year.

  7. Ross said...

    His BABIP is so low every year because he has a tendency to alter his swing path at time and swing for the fences, leading to pop ups. If he’d level out his swing he’d have more line drives and ground balls, both of which would lead to more hits and a higher BABIP.

  8. jessef said...

    Ian Kinsler

    2011
    home: .290 / .399 / .517, wRC+ 158
    away: .203 / .289 / .413, wRC+ 90

    and this isn’t at all a small sample thing.

    career
    home: .308 / .395 / .525, wRC+ 147
    road: .241 / .313 / .409, wRC+ 92

    Can anyone who watches him everyday explain what is going on with this?

  9. jessef said...

    @YP:

    So you’re suggesting that a BABIP difference of .065 over 2800 evenly split plate appearances is the result of luck?  I know BABIP is highly variable, but that’s more than two full seasons of data for each and a huge difference in BABIP.

  10. Jeffrey Gross said...

    This is from Derek A.:

    This does look extreme, but adding on to what Joe was saying, think about their division as well.

    So, they play no road games in Arlington, but about 1/3 of their total road games vs the AL West, all of which are pitchers’ parks.

    This used to happen with the Rockies, where everything conspired to make the Coors split seem more drastic than it really was. When you play home games in a great hitters’ park,m not only does it boost your home splits, but you lose the chance to get the benefit of that park as part of your road stats. Then, with the imbalanced schedule, the overall balance of the parks in your division have a disproportionate impact on those numbers. The Rockies had some of the worst hitters parks in their division, and they suffered from this confluence of circumstances big time. Texas could be rehashing this to some degree.

    Nelly Cruz career splits:
    Home .306/.366/.579
    Road: .238/.297/.436

    Michael Young
    .325/.373/.492
    .282/.326/.410

    Hambone’s got about a .130 OPS differential as well.

  11. YP said...

    Kinsler’s home/road splits are weird as he hits about the same amount of homers on the road and home…it all comes down to his luck on the road is just terrible but there is no discernible reason for it.  I will say, after the first month or two, he has hit much better on the road.  But his babip for everything from grounders, liners, flyballs, etc is terrible on the road despite his power numbers not really being inflated by Arlington.  In 2008, he did hit over 300 on the road, however.

  12. jessef said...

    Okay, that makes some sense.  But these wRC+ numbers are supposed to be park-adjusted.  Is it assumed that Kinsler is playing all of his road games in neutral parks?  I’d think they’d be adjusting his numbers based on the actual parks he’s played in.

    If so, demonstrating this point with other Rangers and Rockies players just suggests to me that the accepted park factors are insufficiently accounting for how much those parks help(ed) hitters.

  13. Derek Ambrosino said...

    demonstrating this point with other Rangers and Rockies players just suggests to me that the accepted park factors are insufficiently accounting for how much those parks help(ed) hitters.

    Quite, possibly, and some more than others too, I’d presume.

    Parks don’t impact all hitters equally, as you know. I think that guy, Michael Schnell, did some work on this. My boy csh19792001 from my baseballfever days used to toss around some more specific data to support this too. So, perhaps, it’s a bit of this, a dash of that, and a sprinkle of the other, and it all combines to really blow out Kinsler’s splits.

  14. jessef said...

    Yeah, pretty weird.  I wonder how he’d do if he played on another team.  I’d guess that his splits would be less pronounced

  15. Spencer Silva said...

    I play in a league that values walk totals and dings you for K’s. In my league, Kinsler has been by far the best 2B this season. His BB/K ratio is the best in the league, which-in my opinion- is the paramount sign of a gifted hitter. Yes, Alberto Callaspo and Juan Pierre probably have great BB/K ratios too, but they come without 30+ HR power. How many 30+ HR guys that you think of can strike out less than 80 times in a season? I’m pretty sure I can count them all on one hand.

    In leagues where OBP and K’s count, Kinsler is your man. Pedroia is your other man, but he won’t steal as many bases and smash HR’s…

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>