Bluefan Posted June 27, 2006 Share Posted June 27, 2006 Hello fellow DCPers, you may have read my thread earlier this season (circa 1 week ago) that featured some predictions based on a scant amount of data. Well, the season has progressed, more corps have competed, more scores have come in, and I have wasted even more time to bring you some updated and hopefully logical predictions. If you don't care how I came up with this, skip this huge paragraph, otherwise, read on fellow nerds! There are some exceptions, but generally this is how it goes. Before I start, I should tell you I include interpolated data. In other words if corps A scored a 70 on the 20th and a 73 on the 23rd, I would essentially give that corps a score of 71 on the 21st and 72 on the 22nd. I did this so that I can track daily progress instead of show-to-show progress, as sometimes (as we know all to well) there are a great many days between shows. I took the difference between each corps most recent and first scores this season, i.e. the progress in x many days. I then divided that value by the progress in the same number of days from the 2005 season. This gives me a unitless ratio that expresses the relation between a corps improvement this year compared to last year. I then take each corps 2005 progress which is calculated by (Final Score - First Score)/Days on Tour and multiply it by the aforementioned ratio to get a prediction of how the corps will continue to progress the rest of this season. Then, I weight this value with the 2005 progress for one main reason: if a corps is not improving as well as last year, it will likely be able to turn things around to some extent. This is usually the case, but some corps are progressing faster than last year, however I had to use the weighting for everyone (otherwise Cavies end up 9th and Esperanza charges into 10th, and other weird stuff happens too). FINALLY, these weighted values, which have the units points/day, are used to extrapolate each corps first score (since previous equations take into account recent scores) to the end of the season. So, if you are still awake, here are my 2006 predictions. 25-22. Corps that haven't competed yet (we're waiting!) 21. Pioneer 66.25 20. Magic - 72.45 19. Esperanza - 74.475 18. Pacific Crest - 74.6 17. Mandarins - 74.95 16. Southwind - 79.025 15. Blue Stars - 82.4 14. Capital Regiment - 82.675 13. Colts - 84.8 12. Spirit - 87.125 11. Glassmen - 87.95 10. Blue Knights - 88.075 9. Crown - 88.675 8. Scouts - 89.275 7. SCV - 89.575 6. Boston Crusaders - 91.425 5. Bluecoats - 95.625 4. Phantom Regiment - 96.475 3. Cavaliers - 96.75 2. Blue Devils - 98.825 1. Cadets - 100.00 Oh, by the way, I had to normalize all the scores because the predicted Cadets score is actually 100.6 therefore these scores are actually (predicted score/cadets predicted score) x 100 Quote Link to comment Share on other sites More sharing options...
Lancerman07 Posted June 27, 2006 Share Posted June 27, 2006 I think that looks pretty good other than the Cadets. Hopefully you can update this in about a week and the Cadets should come down a bit. Quote Link to comment Share on other sites More sharing options...
Gentleman Marcher Posted June 27, 2006 Share Posted June 27, 2006 Hmm...finals the exact same as last year, SCV improving but not much, Boston jumping, Coats remaining, Capital Regiment in 14th (!!!), Blue Stars in finals, and, no surprise, the big three medaling. Everything in finals, placement wise (not so sure about scores) looks pretty safe. Quote Link to comment Share on other sites More sharing options...
KungFuCharlie Posted June 27, 2006 Share Posted June 27, 2006 Hmm...finals the exact same as last year I think the last time that happened was '87-'88. Quote Link to comment Share on other sites More sharing options...
Bluefan Posted June 27, 2006 Author Share Posted June 27, 2006 I think the last time that happened was '87-'88. Placements aren't the same, just the corps that are in finals. Or has that not happened since 87-88, that would surprise me. Quote Link to comment Share on other sites More sharing options...
cire Posted June 27, 2006 Share Posted June 27, 2006 The Grinder is more realistic, and proven to be correct. not too sure about this method Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.