Friday, November 30, 2012

The Case for Kaepernick

For the past week or two, the hot topic in the NFL has been the quarterback situation in San Francisco.  Some think Alex Smith should start, while others think Colin Kaepernick should start.  Personally, I don't think there's a wrong answer.  I think the Niners would be perfectly fine with Smith starting.  But with that said, here's the case for why Kapernick was the right choice.

For starters, a Niners blog post points out that "the 49ers have [had] fewer really bad plays – and more really good plays – with Kaepernick at the controls".  One stat given in that blog post is that, in his two starts, Kaepernick has 10 completions of 20+ yard passes, while Alex only has 22 such completions in the eight full games he has played.  This is reflected in Kapernick's 9.9 yards/attempt, compared to 8.0 yards/attempt for Smith.  Here's some other stat comparisons of Smith's eight starts to Kaepernick's two:

Smith:
Completion percentage- 70.0%
Third down efficiency- 38.0%
Touchdown percentage- 6.0%
Interception percentage- 2.3%
Sack percentage- 10.0%

Kaepernick:
Completion percentage- 66.7%
Third down efficiency- 40.0%
Touchdown percentage- 6.3%
Interception percentage- 2.1%
Sack percentage- 2.0%

People think Kaepernick was chosen because he's a high-risk, high-reward player; that he can provide the big plays that Smith rarely does.  The thing is, he hasn't been high-risk (so far).  Smith's calling card has been ball security.  Kaepernick has been just as proficient in that area, but hasn't needed to take sacks or throw shorter passes to achieve that.  Smith wasn't benched just because he was injured.  He was benched because Kaepernick has been able to play Smith's style and more.

Of course one could (wisely) point out that Kaepernick's stats consist of a very small sample size.  It's possible that defenses eventually adjust to him and that he becomes less effective.  But here's the thing, if that happens, going back to Smith is easy.  It's not like Smith is going to struggle if they go right back to him in three or four weeks.  But on the flip side, what if Harbaugh had gone with Smith over Kaepernick and then there was a visible dropoff from Kaepernick?  The media and fan base would be clamoring for a change.  And if that happened, and a change was made at that point, then going back to Smith is completely off the table.  Right now, Smith isn't benched because he was playing bad; he's benched because Kaepernick happens to be playing better.  If you went back to Smith and then benched him, it would actually be an indictment of his play.  And that you can't come back from.  You can go from Kaepernick to Smith.  But if you went from Smith to Kaepernick, you can't go back to Smith.  In that instance, you're stuck with Kaepernick and would have to hope he really is good enough to win in the playoffs.

Which is better: going with Kaepernick and having Smith as the safety net or going with Smith and having Kaepernick as the safety net?  Harbaugh didn't just go with the option that had the biggest upside, he went with the one that also had the safest plan B.

Saturday, November 3, 2012

Fool's Gold

The Gold Glove winners were announced this week.  The big headline was Mike Trout losing to Adam Jones.  To see how questionable of a choice that was, I decided to look up some stats (Defensive Runs Saved, Ultimate Zone Rating, Fielding Percentage, Range Factor, and Defensive Wins Above Replacement) and compare the two.  Here's what I found:




DRS
UZR
FPCT
RF
dWAR
Jones
-16
-6.7
.982
2.75
-1.3
Trout
23
10.6
.993
2.70
2.2


Trout had such a big statistical advantage it's laughable.  The one knock I read against Trout came from a comment on an ESPN article, which brought up Trout shifting to LF when they would bring in Peter Bourjos for defensive purposes.  The problem with that point is that it doesn't take into account how good Bourjos is defensively.  Being not as good as Bourjos automatically means he's not as good as Jones?  That would be like saying Jordan wasn't as good of a defender as Dumars, because he wasn't the best defensive player on his team.  You wouldn't fault Jordan for playing with Pippen, so why would you fault Trout for playing with Bourjos (who, by the way, did have the highest UZR of anyone with a minimum of 500 innings played)?

Even if you held that against Trout, that doesn't mean Jones was the right choice.  Denard Span had 20 DRS, an 8.5 UZR, a .989 FPCT, a .289 RF, and 2.4 dWAR.  Not only did Jones have the lowest DRS of any qualified center fielder and negative defensive wins, he also committed the most errors.  Yes, he had 439 putouts, but he still only had 54.9 putouts per error.  By comparison, Trout had 132 putouts per error and Span had 84.8.  So regardless of whether you thought it should or shouldn't go to Trout, it clearly shouldn't have gone to Jones.  Here are some other questionable choices:

Carlos Gonzalez over Martin Prado:



DRS
UZR
FPCT
RF
dWAR
Gonzalez
-13
-8.5
.982
1.71
-1.9
Prado
12
10.7
.984
1.72
1.7


Chase Headley over David Wright:



DRS
UZR
FPCT
RF
dWAR
Headley
-3
6.0
.976
2.67
0
Wright
16
15.4
.974
2.50
2.1


Jimmy Rollins over Clint Barmes:



DRS
UZR
FPCT
RF
dWAR
Rollins
-8
4.4
.978
3.83
0
Barmes
13
14.4
.972
4.24
2.1


Finally, the biggest travesty of all, Andrew McCutchen over Michael Bourn:



DRS
UZR
FPCT
RF
dWAR
McCutchen
-5
-6.9
.997
2.44
-0.2
Bourn
24
22.4
.995
2.59
3.0


McCutchen over Bourn?  Bourn is statistically the best defensive outfielder in baseball.  How did he lose to someone with subpar stats?  Even if you don't think advanced metrics are 100% conclusive, you can't ignore significant gaps.  If Bourn had 24 DRS and McCutchen had 19, you might be able to make an argument that stats don't tell a complete story.  But a 29-run difference?  That's hard to write-off.  All of these questionable decisions involved at least a 19-run difference.  Is it really reasonable to expect a margin of error of 20?  These players chosen were questionable at best and preposterous at worst.  I hope these perplexing Gold Glove selections don't carry over to the AL MVP, and that the rightful winner prevails.