Roundup 1/16

Does a playoff bye increase a team's chances of winning in the divisional round? Jason Lisk reminds us that, accounting for team strength, probably not.

Jason also shows that Vinny Testaverde was better than most people think if you consider how bad the teams around him were.

More on QB comebacks from Scott Kacsmar.

Last week we saw the Ravens stomp the Patriots throwing only 10 times for 34 yards. Chase Stuart looks at the reverse. How close have NFL teams come to never running the ball in a game? Could we see something like that in today's Cardinals-Saints game?

I'm not sure if I linked to this already, but here's some really thoughtful stuff from Chase on whether defenses can be thought of "good against the pass" and "good against the run," or are they just "good?" The articles are full of game theory implications. Ideally, teams should balance their schemes so they are not significantly worse in one facet or the other. I just don't think that's the reality of how coaches implement their schemes, or at least limited by the personnel they have. If they did (or could), we'd see them geared to stop the pass far more often on 1st and 2nd down, and ready to stop the run more on 3rd downs.

In the last Roundup, I briefly touched on a debate about how well draft position correlates with QB per-play performance. Dave Berri, who found that the connection is very weak, wrote me to clarify his findings. He makes a good case that for QBs who qualify with at least 100 attempts, draft position does not correlate with per-play performance. He uses his own measure of performance and the NFL QB rating as performance metrics, but I've confirmed the finding if you just look at Adjusted Yards Per Attempt. Here is a nice summary of the debate from the WSJ's Carl Bialik.

Carl points out that using a minimum qualifying criterion causes a big problem. QBs who aren't good enough in practice, pre-season, or spot-relief to be given enough playing time to accumulate 100 attempts would tend to be the worse QBs, presuming coaches have any ability to discriminate between good and bad passers. These non-qualifiers tend to come from the later rounds. What Carl is referring to is known as survivorship bias, which is the reason you should never believe the claims of mutual fund companies (or odds makers). Basically, it means that the worst performing cases in research studies fail to exist long enough to be considered, biasing upward the performance of their category as a whole. In this case, only QBs good enough to stick around past 100 attempts survive to be considered.

I tried to solve this problem a couple years ago by assigning the non-qualifiers the equivalent stats of the 95th percentile passer (who turned out to be Danny Wuerffel in my database). The result was a clear connection between draft position and per-play performance. Dave Berri rightly pointed out that this may penalize non-qualifiers too harshly, biasing my results. I've gone back and looked at the data several different ways, and I was surprised just how little draft position matters. I'll have some interesting findings to post closer to draft season.

Survivorship bias is a common problem in sports studies because the worst players are either on the bench or on the street. It presents a particularly difficult challenge in aging and injury studies. The players who to peak later than most are the ones still around to be counted late in their careers, making peak ages appear later than they typically are. Likewise, only players who tend to stay healthy are still around to be counted, complicating even the simplest injury studies.

Here's another way survivorship bias rears its head in sports analysis.

I think "momentum" is one of the most commonly overused and superstitious concepts in sports analysis. It's mostly a run of good luck. But what many people call momentum can also be a product of health, fundamental improvement, or other unmeasured team qualities. Here's a look at the December and January regular season records of the recent Super Bowl opponents.

Could the prevalence of dome teams in the NFC be creating a quality mirage?

Did the 40-year old Brett Favre just finish his best regular season ever? Was Aaron Rogers better this year than Favre ever was? Neil Paine tells us.

Football Outsiders has been posting a 'stat of the day' recently. Some good nuggets in there. Here's a look at who leads the league in "failed completions," complete passes that don't count as successful plays according to the Hidden Game of Football system. Here's who leads in QB hits. Which teams have been the most imbalanced between offense and defense? Which cornerbacks are best at stopping the run?

The Book of Odds looked at how unlikely last week's three wildcard rematches were. Here's more from Book of Odds on football.

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

6 Responses to “Roundup 1/16”

  1. Ian says:

    Is the 90 points conceded (so far, with one quarter to play) by Arizona a playoff record? If not, what's the highest number of points conceded by a team in the playoffs?

  2. William says:

    I'm not a fan of the "December and January regular season record" stat. Each team gets 16 games, so let's use the Week 14-17 records. No team gets a bye after Week 10.

  3. William says:

    The 1992 Buffalo Bills allowed 103 points, the 1989 Denver Broncos allowed 99 points, and the 1985 New England Patriots allowed 94 points.

  4. Anonymous says:

    Brian, been watching your site since week 9 , great stuff.
    I’m working on my own model but finding it difficult to
    Find game data. Currently using Iskoe’s stats. What do you think?
    Anyone know where I can find an archive of past Iskoe’s stats.

  5. Ian says:

    William - thanks. The 92 Bills did cross my mind as I remembered them conceding 52 in the Superbowl.

    Funny how each of the previous teams made the big game.

  6. James says:

    Funny, or does it make sense that the more games a team plays the more likely it is to have points scored on it?

    I'd bet those three teams just happen to be the outliers that allowed points AND won. The more interesting question would be average number of points allowed per game, min. 2 games played.

Leave a Reply

Note: Only a member of this blog may post a comment.