Recent Posts

Pages: [1] 2 3 ... 10
North Region football / Re: FB: North Coast Athletic Conference
« Last post by Bombers798891 on Today at 08:12:05 am »
In talking about returning starters/production, I think it's also tricky because, especially on defense, there's a lot of noise.

ETP already mentioned the various alignments that require different guys to come in. But it goes way beyond that. Any of the 11 guys can get a sack, INT, or tackle on any play. Multiple guys can get credited with a tackle or sack, and it might not even be clear who did what on a gang tackle. Or, if a guy is obviously making his way to the sidelines and I give him a shove just before he does, and he falls as he's going out, maybe the SID decides to give me a tackle, maybe they don't. Then, you factor in special teams, where there could be another 60 tackles and a couple of fumbles to go around. And because some defensive stats (INTs, sacks, fumble recoveries) are relatively rare, a couple garbage time ones by a guy can really overstate his production relative to his teammates.

At least on offense, despite different alignments and packages, only one guy gets each carry, catch, or pass, and there's no judgement calls about what happened, and with several hundred catches and carries, and thousands of yards, you don't really have a guy who can fluke his way to a large percentage of that. I actually feel like I'd want to use "starters" for defense, and production for offense.
The latest around the nation podcast is out featuring Coach Caruso.

I'm excited to listen later today and getting pumped up for the season to start!

Thanks for posting the info about all the returners.
Very interesting. Good stuff 


You're right on regarding the two favorites. Clearly a separation between the two and the rest of the conference.
I can understand your desire for the season to get started, but let's wait a while until some more football like weather can arrive.   :)
Are Spring workouts underway yet?  Has anyone seen anything at the SC Times on the J's?
Yea, spring was a couple months ago mate. Being in the wrong hemisphere will do that to a guy.... ;)

No, fall workouts have not yet officially begun.

Ha!  Doh.
Are Spring workouts underway yet?  Has anyone seen anything at the SC Times on the J's?
Yea, spring was a couple months ago mate. Being in the wrong hemisphere will do that to a guy.... ;)

No, fall workouts have not yet officially begun.
West Region football / Re: FB: Minnesota Intercollegiate Athletic Conference
« Last post by OzJohnnie on Yesterday at 09:58:26 pm »
Are Spring workouts underway yet?  Has anyone seen anything at the SC Times on the J's?
West Region football / Re: FB: Wisconsin Intercollegiate Athletic Conference
« Last post by BoBo on Yesterday at 09:16:21 pm »

"Lock Haven (D-II PA): After 41 years in the coaching profession, assistant head coach / linebackers coach John Miech will retire at the end of the moth. Miech is the former head coach at UW-Stevens Point (D-III) where he went 156-84-2 before moving on to UW-Whitewater (D-III) where he helped win two national titles."
Northeast Region / Re: NESCAC Hoops
« Last post by kate on Yesterday at 06:31:56 pm »
Sure hope that Drew is happy in their "new" league.  One of these days, we have to have a discussion on the merits of starting new leagues.  That would be different from switching leagues for financial, number & types of sports offered and distances having to be travelled. 
My model's preseason predictions for the WIAC:

Team  Overall  Conference  Pool A Probability 
Whitewater  8-1  (6-1)  46.18%
Oshkosh 8-1  (6-1)  36.64%
Platteville  7-2  (5-2)  13.20%
Stevens Point  7-3  (4-3)  3.62%
Stout  2-7  (2-5)  0.19%
LaCrosse  5-5  (2-5)  0.07%
River Falls  3-7  (1-6)  0.06%
Eau Claire  2-8  (1-6)  0.04%

No surprises at the top here, with the only three teams to make the playoffs for the last few years at the top.
The model has LaCrosse rated as the slightly better team than Stout, but home-field advantage factors tilt more in favor of the Blue Devils.
(Don't worry about the total wins not adding up too much. If I listed each team's average win-loss records, as opposed to their singular most likely record, the number of total conference wins would be equal to the number of conference losses. Its a trick of the rounding.)

Can I ask a question about how the model works?

I'm going to anyway. 

Does your model actually project the individual game results within a bunch of individual simulations, then average the whole works to spit out a projected average win-loss record?  Since you mentioned rounding, and also I see that UWW and UWO are both listed with a projected WIAC record of "6-1" that leads me to believe your projected records are derived from many simulations.  If I've understood correctly, I assume that this would be something like 1,000 simulations where UWW beats UWO 55% of the time, but the handful of other scenarios like UWP or UWSP beating UWW nudged it just over the halfway mark and gave UWW an "average" of more than 0.5 conference losses.

Basically, I'm trying to illustrate the distinction between "1 simulation says UWW will beat UWO but lose to someone else lower in the standings" vs. "over 1,000 simulations UWW beat UWO in 55% of scenarios, and that plus the small handful of losses by UWW to other teams in other scenarios tipped it such that both teams had a conference loss in 500+ of the 1,000 simulations, rounding us off to an average conference record of 6-1 for each."

10,000 simulations actually.

(Things are about to get a bit nerdy)
Each team starts with their preseason Offensive and Defensive ratings, from which I can make a score prediction for Week 1. Using a random number generator and the observed error distribution of score predictions from previous seasons, a game score is developed for each game in Week 1. In each simulation, the games are simulated sequentially, which means the simulated "results" of Week 1 affect the predicted score for Week 2, because each team's rating is updated after every week (this allows for a more realistic distribution of outlier results, i.e. teams dropping from the Top 25 in the preseason to 3-7 by season's end). Rinse & repeat 9,999 times.

So you're absolutely right about your assumptions about the UW-W/UW-O situation, and did a more concise job of explaining it than I probably would have.
North Region football / Re: FB: North Coast Athletic Conference
« Last post by HansenRatings on Yesterday at 05:42:15 pm »
If any of you can find play-by-play data from a centralized location for all of d3, let me know. The best thing I can think of to approximate "percentage of snaps played" would be some combination of K1*(Games Played)+K2*(Games Started). My first assumption is to use K1 = 1 and K2 = 2 (a starter plays twice as many snaps as someone who merely appeared in the game), but I feel like it will probably skew more towards 3:1. I actually started scraping some of that data from the NCAA Stats site today.
If I can get a good number of teams input how ExTartan was describing, and the results are no more accurate than a coach saying "we're returning X of our best 11," I'm probably not going to spend too much time pursuing this further.
Pages: [1] 2 3 ... 10