Friday, March 21, 2008

Keeping Score

Driven by a fascination with how data can be represented visually, a fellow by the name of Alex Reisner has spent a lot of time studying score sheets. He has over a dozen available -- including some originals -- for study, download, and use on his site. I'm not much of a scoring guy -- I tend to concentrate more on beer acquisition, consumption, and processing while at the ballpark -- but I know many of you are. If so, definitely check out Alex's work. He has some neat statistical research tools as well, though I'll admit they're lost on a Luddite like me.

More up my alley are his road trip pages. I'm a road trip freak myself, even if having small children at home has put off my driving pursuits for a few years. When time permits, I find myself obsessively reading other people's online road trip stories. Some are good, some are bad, but almost all are interesting on some level. I found Alex's particularly enjoyable because, unlike so many, he doesn't feel the need to document every single diner, gas station, and rest area on the road, which can quickly becomes tedious. Instead, he focuses on the major themes and emotions which serious road-trippers will easily recognize, and augments them with some excellent photography and even audio.

Go there. Waste some time. You'll enjoy yourself.

1 comment:

Unknown said...

Thanks for this-- fantastic road trip photos and thoughts. Beautiful. I'm planning a road trip this summer, so it was both enjoyable and useful.

The stats pages were also great. He actually doesn't do a great job of explaining what Z-scores are, which is too bad because the better you understand them, the more clearly you can see how truly outstanding some of those top seasons were.

Here's my attempt: standard deviations (SD) are like markers tell you how far from the average this person/score/number is. They always work the same way (in a normal distribution)-- 68% of the population is within one SD in either direction of the mean. 34% below, 34% above. If you go 2 SDs from the mean, add an additional 14% on either side. So now 96% of the population is included, 48% on either side of the mean. When you get up to 3 SDs from the mean, you're including about 99.99% of the population. Z-scores tell you where in this range someone falls. (Note: this is easier to understand pictorially, I think.)

SO... if a player's Home Run Z-score is +2.00, he's 2 SDs above the mean. That means he had more HRs than 98% of players that year. That's why some of the numbers are so amazing. A +5 Z-score!? Wow! I don't know...maybe I have to be a stat geek to appreciate it :)