2A looks wide open right now.
- Thread starter GilliamRatings
- Start date
I like the Gilliam Ratings but it seems like it takes at least 4-5 seeks before they get enough data to crunch good numbers. For example Union is 3-0; beat Richlands; PF and PA are about the same; AND Richlands is #2 and I Union is way down the list somewhere. Overall pretty good indicator during the second half of the season. Anyone familiar with how they get the ratings?
I think the limited data set affects every objective rating system. With only three weeks into the season, you have to rely on data from last year's results, at least to a small extent. Unfortunately, we're finding out that some teams have greatly improved while others have fallen off, particularly in 2A. Marion is clearly a different team than they were last year. Gate City hasn't looked very good either, but their schedule is brutal, making it hard to tell if they are worse or the competition is just way better.
As for the Richlands/Union example, I think the Gilliam ratings will hold true in the end. If the two teams were to play each other 10 times, I'd be willing to bet Richlands wins more often than not. Union got the better of them in a very close game, but are they the better team?
I was thinking this very thing, other than the fact that the 3 state champions in 4A, 5A and 6A look pretty darn strong, there seems to be a lot of shuffling at the tops of all the classes right now.
Richlands and Union are so similar. I think Richland's defense is better but Union is deeper esp. at RB. They have at least 4 good RBs with a good scrambling QB. 2A west is gonna be deep with Appomattox, Buckingham, Giles, Glenvar, Union and Richlands. Possibly a couple others as well. Havent heard much from 2a east this yr. Would like to hear who are the favorites so far.
Usual suspects in the east: Clarke, Goochland, King William. All three have lost at least once already, though. Thanks Essex! Strasburg is undefeated but hasn't been tested. Bruton and Buffalo Gap may surprise some folks. I'm particularly bullish on Gap. I expect them to finish second behind Riverheads in the Shenandoah district. Bruton will have to run the gauntlet of the 3A/4A Bay Rivers District which will prepare them for a deep playoff run. Wilson will make the playoffs, but they are a mere shadow of last year's team. There are some quality teams in the east, but I doubt any of them are playing on the same level as the top teams in the west.
All that matters is who is clicking come playoff time. Richlands was 7-3 last year in regular season but starting clicking toward the end of the year. If you told me week 3 last year that Richlands would go to the semis i wouldve never believed it. I just hope all teams can stay healthy.
Thanks for posting. Sounds reasonable for the most part and I now understand.
My only issue with the formula is that teams that run up points with keeping first teams in on weak competition after the game has been decided (usually late in games) are going to be overrated in this formula. Teams that show sportsmanship and/or develop players by substituting early and mostly run the ball with big leads may be underrated with this system. That variable seems unaccounted for in the equation and leads to . Is there a way to account for that or would it make it too complicated?
Funny thing was people who weren't in the area ranked Wilson really high going into the season.... Now not so much...
All based on prior years early like OB said
Actually, there is a blowout correction that dampens that down. As you outplay what is predicted you get a law of diminishing returns. If a team gets It's hard to get much return (though you get a little a bit) for running it up big time. I would also say the flaw is the opposite of what you say. Teams that put in their twos probably get a little underrated.
It's much harder to move up beating on a bad team than you might imagine.
The biggest mismatch the computer picks all year is 107-0. The favorite in that game must win by at least 74 points just to hold their rating and by 108 in order to move up.
Also, if you do get yourself overrated one week, well the next week you've set the bar much higher for yourself and you're likely to come right back down to earth.
How about not doing a ranking until 3 weeks or so into the season? Basing this year off last year is a poor starting point, for the most part.
I will entertain myself however I darn well please.
Thin skin is not well suited for posting on message boards, My point was that if the purpose is to rank THIS years teams, why not wait until each team has played 2-3 teams to get a gauge of who is who? If your sole purpose was to entertain yourself, congratulations!
Because statisticians have found no variable that correlates better in predicting this years success than the previous year's success. This is, of course, proved every year when the computer picks at least 2/3 of the week 1 winners correctly with no data yet. I also have plenty on my website to explain exactly how accurate you can expect these ratings to be each and every week of the season.
My point is still this. I like doing it. So I do it. When I get bored with doing it I stop for a decade.
I'm a numbers guy myself, love the work you do with the GilliamRatings.
I am sorry. I thought that statement seemed as if you were a little peeved. By all means, Keep Calm and continue entertaining yourself!
