A respected Columbia professor disputes the findings that trailing at halftime makes a team more likely to win. Andrew Gelman's rebuttal is over my head, but seems to boil down to different ways to interpret the same data, and he includes this instant classic of a line: "they do some ridiculous overfitting with a 5th-degree polynomial. Don't even get me started on this sort of thing." Someone had to say it, and he said it.
ESPN (!): Wednesday Bullets
Comments