Saturday, November 15

Stat Boy Saturday: Looking Ahead to the Playoffs, Part II

Another week with our honorary stat boy, Zach Fein of Fein Sports. A contributor with as much stat muscle as anyone in the business. You may not understand what he's says and use it for your fantasy leagues, but if you're in the mood to be baffled and confused, then boy does he have you covered. Because life is one big spreadsheet. This week he continues last week's article on fantasy playoff performances. It's good to know one of our own isn't falling into the trap. Read Tosten's News From Ball Street one more time. Now read it again. See any mention of fantasy playoff schedules? I'm sad to say that Ken Daube of is the most recent fantasy pundit to fall into the trap of looking ahead to stretch-run and fantasy playoff strength of schedule. As ESPN's own numbers guy, shouldn't he know better? Last week I looked at that topic and indeed found there was no correlation of regular season and fantasy playoff performance. I concluded that just because a team was poor against the pass from Weeks 1 through 12, doesn't necessarily mean that they'd do poor against the pass for the remainder of the year. This week, I did the same for individual players. In the sample I used the top 12 quarterbacks and the top 15 running backs and wideouts from 2005-2007 (not a huge sample size, I know, but the Web site from which I got this data had week-by-week game logs dating back to only 2005). Since some players don't have much playing time in Week 17, I used Weeks 1 through 11 as the "regular season" and 12 through 16 as the "fantasy playoffs." Here are the correlations for each position, in fantasy points scored*:

Fan. Points Correlation
Pos. r r2
QBs 0.30 0.09
RBs 0.23 0.05
WRs -0.04 0.00
As with the previous article on this subject, there's no meaningful connection between regular season and fantasy playoff performance. Wide receivers' playoff production is essentially a pick-a-number-from-a-hat situation—you can't predict it whatsoever. Now, here're the correlations if you split each position up into three tiers—the top 12, middle 12, and bottom 12 in the 36-player quarterback sample, or equivalent to the QBs ranked Nos. 1 through 4, Nos. 5 through 8, and Nos. 9 through 12; and the top 15, middle 15, and bottom 15 in the 45-player running back and wide receiver sample, or equivalent to the top 5, etc. You might want to take this with a grain of salt as the sample size is so small, by the way.
Correlation of Tiers
Pos. r r2
Top Tier
QBs 0.58 0.34
RBs 0.02 0.00
WRs 0.12 0.02
Middle Tier
QBs 0.45 0.20
RBs 0.32 0.10
WRs -0.12 0.01
Bottom Tier
QBs -0.25 0.06
RBs -0.43 0.19
WRs -0.48 0.23
One thing first: All positions in the bottom tier have a negative r. This is most likely due to all the players (such as Ryan Grant, Earnest Graham, Ladell Betts, et al) whose late-season performance vaulted them into the top 15. The highest correlations go to the quarterbacks, but an r-squared of .34 still is not very significant in predicting a player's playoff performance. What does all this mean, then? It means that there will still be those fantasy playoff flukesBilly Volek circa 2004that raise a borderline playoff team in your fantasy league to the championship. It means that there will still be those fantasy playoff bustsChad Johnson Ocho Cinco Johnson circa 2007—that screw over the No. 1 seed to a first-round loss to the fourth-ranked team. It means that there will be fantasy experts that'll be berated for ranking Player A over Player B, after Player A throws three picks and no touchdowns and Player B throws for 300 yards and two scores, even though the stats say that it's extraordinarily difficult to predict a player's playoff performance. And it means that, once again, you'll lose before reaching the finals due to your No. 1 pick's putting up single-digit fantasy points for the first time all year. It happens.

No comments: