On April 29 2006 I ran a half marathon in Lieshout. During the weeks before that I had asked readers of GliaWebNews to predict my time and rank. The following predictions were made:
| Time | Timeoff | Rank | Rankoff | Timerank | Rankrank | Total |
|---|---|---|---|---|---|---|
| 5890 | 9 | 30 | 6 | 7 | 6 | 13 |
| 6441 | 542 | 30 | 6 | 5 | 6 | 11 |
| 6323 | 424 | 22 | 14 | 6 | 4 | 10 |
| 6600 | 701 | 30 | 6 | 2 | 6 | 8 |
| 6453 | 554 | 14 | 22 | 4 | 3 | 7 |
| 6480 | 581 | - | 99 | 3 | 1 | 4 |
| 3534 | 2365 | 1 | 35 | 1 | 2 | 3 |
Times are given in seconds. The actual time and rank were 1:38:19 (5899 seconds) and 36.
The table also gives for each participant the number of seconds they were off (Timeoff), the predicted rank (Rank), the number the rank was off (Rankoff), the reversed rank of Timeoff (Timerank), the reversed rank of Rankoff (Rankrank), and the sum of Timerank and Rankrank (Total).
The winner with a Total of 13, and 9 seconds off, is Edward Vanhove from Belgium.
It seems the predicted times are a somewhat better indicator of one's ability to predict than the predicted ranks.
It seems the raw number of seconds one is off (Timeoff) is just as good a predictor of IQ than is the Total score. Timerank is a less good predictor of IQ, probably because the information about exactly how many seconds one is off is lost in ranks.