## Friday, April 07, 2006

### Sliding Scale

Predicting your score on the LSAT is kind of like predicting tomorrow's weather. You can see a warm front advancing and you know it will probably rain, but there are just too many variables and you can't be certain. The LSAT poses the same problem.
• The questions change, and a different number of students take each test.
• As a result, each test's scaling changes.
• As a result, the corresponding percentiles change.
In three tests, I've scored within a four-point range. If I took the test for real tomorrow, it's reasonable to figure I would place within that range — but it could also be two or three points outside, which would widen the percentile possibilities to anywhere from 83–96%.

If I can narrow and lift my score just a few points into a predictable 168–170, that should translate to a percentile range between 97–98%. That's the difference between a 13-point range versus one single point's gamble — and when you're facing a three-year investment that's going to sink you into a short lifetime of debt, and since high scores can translate to lower tuition, that can be all the difference in the world.

If you're wondering why I've been anxiously sweating a few measly points, or if you're a friend wondering why I seem to have disappeared from the face of the planet lately, that's the answer. I've got a three-hour window to score \$100,000. I'm basically planning a heist, and it requires homework.