If you don't read Carl Bialik's two columns at wsj.com, you should check them out. Twice a day he offers up interesting takes on things numbers and sports related. The Numbers Guy is a column on the intersection of topical subjects (often sports) and math. The Daily Fix is a daily wrap-up of interesting sports stories. Often his posts are estimates of "how unlikely was that?" (I contributed to his "How unlikely was the Steelers-Chargers 11-10 score?" article.)
This week his columns focus on football, and the playoffs in particular. Bialik notes how various sports prognosticators can go from awful to great in one year. I think this is more evidence that these guys don't know anything that the rest of us don't. They're just guessing or extrapolating the present. And they're wasting their time.
Bialik looks how unlikely that all four home teams are underdogs for the wildcard round this weekend. He also looks at how historic the Dolphins turnaround was this year.
Phil Birnbaum also chimes in on the same subject, linking to Bialik's post. He also beats me to my own punch by citing the fact that last year my system had the Dolphin's pegged as a far better team than their 1-15 record indicated, and so their turnaround may have about as much to do with luck as other factors.
The PFR Blog has a fact-filled post on the history of the NFL's playoff tie-breaking methods and which ones are most predictive of which teams will actually be successful in the playoffs. I'm not sure the sample sizes are big enough for such an analysis, but I do have a couple opinions on what methods are best.
To me, the head-to-head tie breaker is probably the most fair, or at least the most acceptable. Although one team will have enjoyed home field advantage if the head-to-head match-up was an inter-division game, at least we can say it was "settled on the field." One commenter noted that in the current system, an NFC team's victory over the Lions would be more valuable than an inter-conference victory over the Titans. That's why I'd suggest a very simple strength-of-schedule tie breaking system.
Say two or more teams are tied at 10-6 for a wildcard. A simple and fair way to break the tie would be to ask how hard were those 10 wins to come by. We could just add up the total number of wins of each of the two team's opponents, common or not, and whoever had the tougher schedule gets the playoff berth.
Football Outsiders looks at whether older quarterbacks underperform in the cold months of the season. They don't, at least according to their analysis, which is based on fantasy points. I think the more interesting thing is that QB performance, regardless of age, doesn't fall off very much at all in the winter months.
Smart Football has a good philosophical post about what is "real" football. Pounding the rock out of the I-formation all game long, right? No, says Chris. It's all arbitrary, and the rules have been changed so many times that there's no such thing as real football anyway.
I tend to agree. When I was doing research for a post about the extra point, I was fascinated by how the rules of American football evolved from rugby and even soccer. Back in the late 19th century, when football as we know it began, college teams from the northeast were constantly squabbling over which rules to use--"football" rules with a line of scrimmage, or standard rugby rules. Often the same teams would play a game with one set of rules in the morning, then another with the other rules in the afternoon!
Weekly Roundup
By
Brian Burke
published on 1/03/2009
in
roundup
Subscribe to:
Post Comments (Atom)
I don't think it's very fair to say that pre-season predictions are worthless. I take a lot of pride in the fact that not only did I put my projections out in March, but that I don't have any amazing statistical formulas or super computers that do the work for me. I take a time-tested stock analysis technique (Relative Strength Index) and apply it to the NFL. Very simple. I just have to determine whether each team is trending downwards toward an RSI of 45 or upward toward an RSI of 55, then make an estimate of a winning percentage that would have them headed in the right direction by the right amount. It's the perfect balance of objectivity and subjectivity if you ask me, and the results are competitive with any other method I've found.
See fo' yo'self:
http://www.thefootballprofessor.com/2008-nfl-forecast/
That's just it. Your predictions are just as good as any I've seen.
Well, I can see your point that it is a waste of time for game charters or whoever to pour thousands of man hours into analyzing game tape and play-by-play data just to produce a set of projections that a guy like myself can produce in just a couple hours with an Excel spreadsheet. I see what you're saying there. But I don't know that you can say pre-season projections are worthless...I mean, there has to be some value to them. Even if they're just a guide, or course there will be deviations (significant ones at that) but it is pretty amazing to hit 18 teams within 2 games of their actual record before they even play.
If I mindlessly regress last year's wins at:
'08 wins = 5.7 + 2.9*('07 wins)
I get 17 teams w/in 2 games. Plus I bet my overall error rate is at least as good as any other prediction.
It literally took me 3 minutes on Excel.
If just guess 8-8 for every team, I get 15 teams w/in 2 games.
Doug: can you regress your projected wins to actual wins, finding slope and r-squared?
R-squared, yes, I can get that. You'll have to forgive me, I'm not a math whiz...what's the deal with slope? What would that tell me? And how would I find it?
It seems to me that the odds that all 4 wild card teams are better than their division-winner opponents is simply the odds that the 3rd and 4th best teams in a conference are in the same divisions as the first and second best teams. And that should be 6/14 x 5/13 = 30/182 or about 1 in 6. So it's about 1/36 that all 4 road teams are better on wild-card weekend. Of course, once you throw in HFA, it gets murkier. But the odds still seem pretty reasonable.
Brian, correct me if I'm wrong but your simple linear regression includes intercept and slope values that you solved for after knowing the "'08 wins", right? So that's a fine analysis but it's not a fair predictor, since you know the future already. Or is your mindless regression a generic forumula you always use?
Your point about 8-8 for all teams is a good one.
No, that doesn't include 2008. It's from 2002-2005, where one season's win totals are used to predict the following season's (a total of 3 "season-pairs").
It's from this post I linked to above. The regression formula is at the bottom of the article.
The 8-8 prediction is clever, but it really depends on how you measure accuracy. Doug chose "w/in 2 games," which is a 5-game spread. If we do mean absolute error, the 8-8 predictions do very well. But we use RMSE (root mean squared error--which is what most statisticians or mathematicians would use), it doesn't hold up well.
We could make 2009 predictions right now that would be as accurate as any that SI or ESPN or FO will make in August. It's not that ours would be very accurate at all, just that nobody's are.
Think about how surprised you would have been to be told in August: Ravens, Dolphins, and Falcons in (avg 3.3 wins in '07). Cowboys, Jaguars, and Patriots out (avg. 13.3 wins).
Who's going to surprise us next year? Rams? Raiders? Chiefs? Lions? Somebody will!
Good, I hoped that the regression formula would be something established from historical data rather than fit to the current data, but you did say it was "mindless" so I wasn't sure. Sorry I didn't read the linked post before commenting.
As for MAE vs. RMSE, of course RMSE will always be greater than or equal to MAE, but if the 8-8 method appears much worse with RMSE then that's an indication of a lot of variance in the magnitude of the errors using that method, since RMSE will punish you for making a few big mistakes. Which would indicate that guessing every team will finish 8-8 is likely to be pretty good most of the time, but you'll probably get a few big mistakes.