Jason uses our EPA model to assess red zone pass locations. Excellent analysis. Excellent visualization.
Keith's efficiency leaders through week 9.
Chase gives us a report on "sack factor."
An update on team injury report calibration. Questionable and Probable designatees play more often than expected.
GB's playoff chances depend on Rodgers returning for the Thanksgiving day game vs DET.
Novel idea for restructuring MLB.
Are the 2013 Jaguars the worst NFL team since the merger?
The problem with infographics.
Game theory has something to say about whether to box kick in rugby. And, using analytics to prevent rugby injuries.
Husband and wife duo wrote MLB's schedule for 20 years. Awesome mini-documentary. Helmet-knock: Punk Rock OR.
Improving the world's most famous game.
I thought this post on the average length of PhD dissertations was interesting (with a great viz). They range from under 100 pages for biostatistics to just under 300 pages for history. It struck me that the length of dissertation is directly proportional to a certain quality of each discipline. So I modified the chart to reflect my theory.
Speaking of BS, Jeff doesn't like it when different statistical approaches to ranking teams result in different rankings:
Now we KNOW that the Chicago Bears aren't the fourth best team in pro football. That jumped right out. ... Again, we KNOW that the Cincinnati Bengals aren't the best team in the NFL.
This stuff reminds me of a Penn & Teller episode on Showtime a few years ago about Feng Shui. Here’s a link to it. As you watch, mentally substitute in the word “analytics” whenever they say “science.” And, you can probably do the same for “efficiency” when they say “chi.” And, if you want…when the practitioners are explaining what they do…imagine Aaron Schatz of FO trying to explain why the Bears are #4, while Brian Burke is trying to explain why the Bengals are #1.Um, excuse me? Would someone care to tell me how exactly we KNOW the Bears are not the 4th best team? Or how we KNOW the Bengals are not the best? Because the power rankings at ESPN told us so? How do we know the Bengals aren't on the precipice of a decade-long run of dominance like never seen before? We don't. One thing we do KNOW is that, so far, after a measly 9 weeks, the Bengals are the most efficient team in the metrics that have recently been most predictive of future wins. "This stuff" isn't any more complicated than that. "KNOW" has a very high threshold in data science, or any science for that matter. First rule of analytics club is there is no KNOW. If you begin your search for truth by assuming your conclusion, well, good luck with that. End of rant.
Hey brian, does your efficiency model take into account fumbles? I want to know if there is a correlation between forcing fumbles in the first half of the year to the 2nd half of the year. Would a model predict that the Chiefs would be good at forcing fumbles (not necessarily recovering them) again for the rest of the season.
The Bears have had a really easy schedule too. The only tough game was against New Orleans @ Chicago. And even their upcoming schedule is really easy.
Here is how Jeff KNOWS the Bears and Bengals are terrible:
"Sloppiness Score: 5 times the number of turnovers…plus the number of incomplete passes. This gives you a great sense of execution for the teams involved in a game.
Stat Score: is the scoreboard equivalent of what the yardage stats were saying. The formula is 2(rushing yardage) plus 1(passing yardage) times 0.67 divided by 15. This formula plus the impact of turnovers and special teams will usually get you pretty close to the actual scoreboard margin."
As you can see from the text, there is a testable hypothesis, measurable data points, and repeatable tests leading to a theory or new hypothesis.
Cheers,
J