Friday

Numbers - It's all in what we make of them

Today's New York Times brings three stories showing, once again, that statistics are what we make of them.

First, continuing the theme of statistics and medical news, a front page story about how a promising gene-based approach to treating lung cancer turned out to be, in the stark words of the story, "wrong. [The] gene-based tests proved worthless, and the research behind them was discredited." The errors were found by a pair of statisticians who found that researchers had moved a row or column of data over by one in a spreadsheet. The story is on the front page of the print edition.

Second is an Op-Ed, "What Does '3,000 Hits' Really Mean?" telling a story about the kind of consistent production, year after year, Derek Jeter needed to even approach 3,000 hits. Only 27 players in all of recorded history have done that. The Op-Ed takes the numbers and turns them into a narrative, and it's a pretty nice tale.

Last but by no means least is a sports column about the Atlanta pitcher Jair Jurrjens, who has produced an ERA much lower than his statistics would predict. You can't model the entire world, or even what happens in a baseball game. Neil Paine, the author, floats a theory that Jurrjens success is "smoke and mirrors," by which I assume he means it's a statistical fluke that won't hold up when he's pitched more innings. But Paine also raises the possibility that hitters can't hit Jurrjens' pitches hard--and maybe the statheads will come up with another measure to account for it. Any thoughts? Post them in the comments.

No comments:

Popular Posts