BP Comment Quick Links
| Home | Unfiltered | Articles | Newsletter | Statistics | Fantasy | Events | Radio | Glossary | Search |
![]() |
|
|
|
May 6, 2003 Doctoring The NumbersHot Starts, Part IIIWelcome to the third and final instalment of my look at the meaningfulness of the first few dozen games of a team season. (Go back and review Parts 1 and 2 here. There will be a test later.) This final article looks to merge a team's starting record with its established performance over the past few years, to come up with a formula that most accurately projects its final record based on the available data. Warning: If you thought Part 2 was laden with too many equations, you're not going to like Part 3 any better. I ended Part 2 with a projection that the Royals, based on their 17-5 start, are projected to finish with about 97 wins. The folly with that logic should be self-evident, but let me share some evidence with you to make the point a little more clear. When the Royals' record reached 13-3, my inner circle of fellow Royals fans finally got serious about questioning whether such a strong start really meant anything in light of the team's 100-loss season in 2002. I decided to look for comparable teams throughout history that had gotten off to a similar start. Using my database of all teams from 1930 to 1999, I found a total of 75 teams that started the season either 12-4, 13-3, or 14-2. Sixty-three of those teams, or 84%, finished above .500. As a group, they finished with a .545 winning percentage. But it's not all roses. Because I then whittled down that group to look only at those teams that had played less than .420 ball the previous season, which corresponds to a 68-94 record or worse. I was left with five teams:
|
40 minutes of my life gone...the formula in Step 2 has an error. It should be P + not, P *.