:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/47048888/usa-today-8249660.0.jpg)
Today's Friday Forum finally gets around to that ever-popular means of determining expectations for a team in a given year: Win shares!
For the uninitiated, win share projections are done by estimating the percentage chance a team has to win a given game. We've run win share projections here at Alligator Army in 2014, 2013, and 2012, and while everyone was a little low in 2012 and a bit high in 2013, many folks pegged Florida's 2014 season pretty accurately.
While FlaGators, bless his heart, consistently ran his projections at the end of July, I wanted to wait until just before the season began to let injuries and possible attrition shake out. Here are mine, with reasoning below.
| Opponent | Win Share |
|---|---|
| New Mexico State | 1.00 |
| East Carolina | 0.85 |
| at Kentucky | 0.75 |
| Tennessee | 0.50 |
| Misssissppi | 0.30 |
| at Missouri | 0.35 |
| at LSU | 0.25 |
| vs. Georgia | 0.25 |
| Vanderbilt | 0.90 |
| at South Carolina | 0.45 |
| Florida Atlantic | 0.95 |
| Florida State | 0.40 |
| Total | 7.00 |
I relied on Bill Connelly's Projected S&P+ Rank a fair bit for my win shares, and I adjusted pretty heavily for home and road games — which is why the hardest games I see on the schedule are a road trip to LSU and a game in the state of Florida against Georgia, both projected as top-10 S&P+ programs, but the hardest home game will not be against defending ACC champion Florida State, which is projected to be 17th in S&P+, but Mississippi, projected to be eighth.
The number I think might strike people as most high is Kentucky's, but the Wildcats are projected to be 52nd in S&P+, suggesting that their hype may surpass their performance. That said, I also don't see any likely juggernauts on the schedule — and Florida beat a similar Georgia team last fall, after all, and was very close to beating a similar LSU team at home — so I'm comfortable putting no values lower than 0.25 on any game, which, of course, raises my overall estimate.
And while I fiddled with the component numbers just a little and made sure to give all of my values on a 0.05 gradation, coming out at exactly 7.00 wins is a happy accident.
Those are my win shares. Let's see yours.