Team WinDivison MakePlayoffs MakeSuperBowl WinSuperBowl ARI 3.0 48.9 3.0 0.9 ATL 3.9 10.2 1.5 0.7 BAL 13.9 35.3 3.4 0.7 BUF 4.2 13.3 1.8 0.6 CAR 79.5 87.5 15.4 9.2 CHI 1.7 12.5 1.4 0.4 CIN 44.1 71.3 11.0 4.7 CLE 0.1 0.4 0.0 0.0 DAL 42.8 46.8 4.3 1.1 DEN 76.8 95.9 25.4…

2014 Record 2015 Record My record 2016: Total (weeks 1-16) – SU: 12-4 ATS: 10-5-1 O/U: 5-10-1 Week 1 – SU: 12-4 ATS: 10-5-1 O/U: 5-10-1 Carolina at Denver Prediction: Broncos 23-20 (58%) Pick: Broncos +3  Total: Over 40.5 New England at Arizona Prediction: Patriots 24-23 (52%)Pick: Patriots +9 Total: Over 44 PUSH Tampa Bay at…

One of my colleagues, Tim O’Brien, showed a slide in the first week of his classes describing how he first found out about R: Here is that email from 1996 (TWENTY years ago!!!) where Tim asked Ross Ihaka how to get R, and Ross helped him out.  #amazing #history Cheers.

There is a general election happening right now and the NFL season is starting.  Let’s talk about the candidates chances in terms of NFL games. At the The Upshot they track 8 different presidential forecasts, 5 of which give actual percentages.  Currently (at 11:07am on September 2nd) Clinton’s win probabilities range from a high of 94%…

Based on @predictit state electoral college markets, Clinton has about a 94.1% chance to win in November compared to a 5.6% chance for Trump (with a 0.3% chance of a tie).  Details of how I estimated these probabilities are here and the code is on github here. Overall in the past week, there has been a small shift…

Recently, did some cool stuff by scraping the data from predictit.org estimating Clinton vs Trump win probabilities using data from their state markets (GitHub code here). Last night, I decided to try to get some polling data and Huffington Post makes their polling data available through a very easy to use API in JSON format…

“There has been a great deal of hype  surrounding neural networks, making them seem magical and mysterious.  As we make clear in this section, they are just nonlinear statistical models, much like the projection pursuit regression model described above.” – Page 350, “The Elements of Statistical Learning”,Hastie, Tibshirani, Friedman. Cheers.

JSM 2016 was last week in Chicago (the greatest city in the world).  As always, it was awesome. First, here is a link t a bunch of slides from JSM that were compiled by @kwbroman Sunday This year I was delighted to get to put together the first (hopefully annual!) JSM Data Art show.  This…

Optimizing Fanduel in R

Sunday, July 31, 2016 12pm: Setup JSM 2016 Data Art show. 4-6pm: “For the Love of the Game: Applications of Statistics in Sports”—Contributed CC-W184d Monday, August 1, 2016 8-10am:   “Advanced Methods for Statistics in Sports”— Contributed CC-W175a 11:30am-12:20pm “Contributed Poster Presentations: Section on Statistics in Sports” —Contributed CC-Hall F1 West Rating Offensive Production in Baseball: A…