At first thought, I assumed something like 68%, 67%, 66%, 65%, 51% should be 3, 2, 2, 2, 1. That is how I treated the first few weeks of this when I was asked to go back and retroactively find what Ken would have scored. Basically, my thought process was that 51% gets an obvious 1, the rest are all nearly equal and all get 2, and then you have 1 left over and have to bump 68% up to a 3.
After further thought, even a case like this one with four nearly unique prediction %'s still should be scored 4,3,1,1,1 if you truly want to best play the odds. The math is rather straightforward, and it shows an interesting distinction between math and "gut reaction". I'll fully admit my initial "gut reaction" was wrong. Glad that the scoring happened to come out the same to-date, and I'll use the proper method going forward.
After further thought, even a case like this one with four nearly unique prediction %'s still should be scored 4,3,1,1,1 if you truly want to best play the odds. The math is rather straightforward, and it shows an interesting distinction between math and "gut reaction". I'll fully admit my initial "gut reaction" was wrong. Glad that the scoring happened to come out the same to-date, and I'll use the proper method going forward.
Comment