South Carolina's prospects for the rest of the year
The Massey Ratings is one of the less heralded computer components of the BCS. The ratings have a predictor based on games to date, with past predictions frozen as they were at the time the game was played.
I have generally found Massey to be more accurate than the better known Sagarin. Massey's forecast for S.C. State was Carolina with a 99% chance of winning and an expected outcome of 39.2 to 5.8, which is essentially what the score would have been had our late fumble recovery not been reversed. Here, according to Massey, is Carolina's likelihood of winning each remaining game:
vs. Kentucky 79%
at Alabama 22%
vs. Vanderbilt 82%
at Tennessee 54%
at Arkansas 56%
vs. Florida 25%
vs. Clemson 61%
If you assume the most likely outcome for each event, Carolina has a great 9-3 year capped by a win over Clemson. Someone better with statistics than me would probably tell you the expected outcome with these projections is slightly less than 8-4, as three of the expected "wins" are by a rather small margin.
However, any true Sandlapper should just go ahead and book your ticket for the Georgia Dome with the Gamecocks' impressive .007% prospects of going 11-1.
5 comments
|
0 recs |
Do you like this story?
Comments
That actually looks about right,
except for that Clemson game. I guess the Massey Ratings don’t take into account one team’s inexplicable habit of losing to an inferrior rival at the end of the season. Anywho, that 25% chance of beating Florida is looking pretty good. An 8-4 year given the pre-season predictions sounds pretty good to me.
Dum spiro spero - "While I breathe, I hope"
State motto of South Carolina
by The Feathered Warrior on Oct 8, 2009 1:17 AM EDT reply actions
What jumps out at me is how unlikely the computers see an upset in the Kentucky and Vandy games being...
Both teams have really dropped off since last year.
Garnet and Black Attack: A Blog by and for Gamecocks Fans.
We've lost twice to Vandy
but we’re an 82% shoe-in this year.
On the other hand, we’ve beaten UT 2 of 3 (and the one loss was because the Vols got three breaks on their final drive, including a false start that let them re-kick the game winning FG) … but we’re only 54% likely to beat them.
I’m not sure I’m buying the Massey ’puter program.
I don't get Tennessee in a lot of ways this year
For instance, they are a slight favorite (-1.5 or so) over Georgia this year. I haven’t seen anything yet to suggest the Vols should be favored over anybody from the top eight in the conference, and I wouldn’t put money on them against the other teams at the bottom.
I suspect the computer has them overrated because of SOS, but people (theoretically) watch the games. We should know better.
by GwinnettGamecock on Oct 8, 2009 3:31 PM EDT up reply actions
Sample size.
Unfortunately, football is the worst sport to apply these predictave models because of the small number of games played. I’d also like to see the methodology behind this model too. And anything from 40 to 60% should be considered “tossup.”
--Robert
by a gamecock fan on Oct 8, 2009 6:04 PM EDT via mobile reply actions

by 














