On the surface, the switch from Smith to Kaep seems like high risk/high reward gamble. Perhaps it is. There are certainly very few precedents for replacing a successful quarterback (e.g,, in terms of won-loss percentage) this late in the season on a winning team with serious superbowl aspirations. Advanced stats (aka Sabermetrics) may provide a different view of this. I was surprised by what I saw.
Advanced Stats, particularly "win probability added", tend to make Kaep look good. (This will be shown and discussed below.) Here I will very briefly explore the questions:
1) are advanced stats meaningful for football? (What are advanced stats?)
2) do the 49ers use advanced stats?
3) What are Kaepernick's advanced stats?
This post will only scratch the surface. We can perhaps go deeper in the discussion.
Sabermetric analysis (popularized in the movie Moneyball) is widely used in baseball. Their use in football is more difficult due to the deeply team nature of the sport. Which teams used "advanced stats" is not well known; teams tend to be secretive about this as this NY times article mentions:
"…when the Baltimore Ravens announced in August that they had hired a director of football analytics, it was a rare public signal of the growing interest among teams in weaving statistical analysis into game-day, draft and free-agency preparation…One general manager whose team does delve into statistics, but who didn’t want to be identified, wondered why the Ravens announced the hire at all.
This general manager suspects that more teams do some form of statistical analysis than are publicly known. People who work in sports statistics, and coaches and general managers, agree that there has been a shift in the N.F.L.’s guarded thinking."
(This is from: " More N.F.L. Teams Hire Statisticians but Their Use Remains Mostly Guarded", NY Times, Nov 24, 2012). (More on this below.)
One of the leading measures used in sabermetrics is WPA, which stands for "win probability added". It probably won't surprise you to hear that Buster Posey, Melkey Cabrera and Marco Scutaro had very high WPAs with the SF Giants this year (and Belt was higher than you might have thought). At each juncture in a game there is a certain probability that your team will win. Each player is credited individually for what they add, or subtract, from that. WPA can be positive of negative. (I plan to post more details about this, and its applicability to football, in an appendix as a comment to this post.)
Here are Kaepernicks stats for WPA for the last 5 games (starting with the 1st Rams game (which he finished after Alex concussion):
.47 , 1.2 , (6) St. Louis (week 10)
.40 , 10.8 , (5) Chicago
.30 , 6.9 , (11) New Orleans
.16 , -0.8 , (17) St Louis
.22 , 9.1 , (6) Miami
Sorry the alignment is poor. The first number is WPA. The second number is EPA ("extra points added") (roughly analogous to RE24 in baseball, which is "runs added" (run expectancy)). The 3rd number (in parentheses) is his rank among all NFL quarterbacks that week (based on WPA).
In terms of WPA and rank, these statistics cast Kaepernick in a very good light. Three weeks are in the top 8, the other two are in the next group of 8 (9-16). Even his worst game, against the Rams in week 13 is not abysmal. One can argue as to whether these are meaningful at all. Advanced stats are difficult to define and use in football. However, if there is a position for which they may be relevant, it would be quarterback.
It is intriguing that the NY times article quoted above refers to the Ravens, coached by Jim Harbaugh's brother, with whom Jim has said he speaks to weekly about football. One can imagine as a point of pure speculation the question: "what do you think about switching from Alex to Colin?" John: "My director of analytics tells me that there is very little downside risk." That is basically the takeaway message I got from looking at these: not as much risk as I thought.
Perhaps some of you already saw and felt this intuitively. ("I can feel it in my bones…") Perhaps some of you vehemently disagree.