Spurred by a recent blog post, I looked up "Best is the enemy of good". Coaches, teachers and statisticians often look at tables like this:
1 | Harriet | 96 |
2 | John | 93 |
3 | Jill | 92 |
4 | Bill | 91 |
5 | Mabel | 91 |
6 | Phil | 89 |
7 | Zee | 88 |
8 | Martin | 88 |
9 | Gil | 87 |
10 | Jack | 76 |
11 | Mary | 74 |
12 | Joan | 70 |
Scores on tests and in sports are often (but not always!!!) produced by processes that look like these balls sifting through this sort of set of pegs:
If you take the ball in the right-most slot (the "best scorers") in this set of results) and paint that ball so you can track it, you will usually find that ball more often in the central slots on the next time you pour them through the pegs. It takes good luck to be in the right-most slot and the painted ball probably won't do so well the next time through.
It is common for one of the balls or one of the players or one of the students to tend to score high consistently but not to be the top scorer. The coach or teacher may not ever notice how high that "good" scorer tends to be. If we only look at the "best", we may not notice the "good".
