Monday, November 15, 2010

weekend review

No recruiting news and no major depth chart news (I hope)** which is good because there are other sports besides football going on.

**Not counting Ras-I Dowling's unfortunate season-ending broken ankle. Dowling's been out so much this season it doesn't really have an impact on the depth chart.

But first, football anyway. High school playoffs and not-playoffs:

Collegiate 28, St. Christopher's 16: In a rematch of last week's game, Thompson Brown's season ends in the VIS Division I semifinals. Brown, again playing tight end rather than running back, caught two passes for 16 yards.

Petersburg 22, Meadowbrook 21: The Richmond-area schools still had one last regular-season game to play. Kevin Green and Petersburg played their way into the playoffs; Green threw for one touchdown, one 2PC, 103 yards, and added 84 yards and another 2PC on the ground. This was all before the fourth quarter: Green was injured on the first play of the fourth, and returned, but only to hand the ball off.

L.C. Bird 41, Midlothian 0: Anthony Harris was 8/12 for 133 yards and two touchdowns.

H.D. Woodson 14, Coolidge 2: Just one catch for Darius Redman in Woodson's first-round win. (I totally misread the Post's recitation of Woodson's playoff scenarios last week. No bye like I managed to think they had.)

Hampton 36, Lake Taylor 7: Productive day for David Watford, who accounted for 233 yards of total offense and 3 touchdowns.

Other scores:

Hermitage 41, Patrick Henry 28 (Diamonte Bailey)
Bayside 26, Cox 15 (Ross Burbank)
Stone Bridge 55, JEB Stuart 0 (Rob Burns)
I.C. Norcom 24, Churchland 18 (Kameron Mack)
Phoebus 49, Warwick 0 (Caleb Taylor)
Good Counsel 44, Gonzaga 0 (Vincent Croce)
DeMatha 34, St. John's 20 (Kelby Johnson/Jordan Lomax)
Damascus 21, Linganore 7 (Brandon Phelps)
Archbishop Curley 15, Boys' Latin 7 (Marco Jones)
Woodland Hills 29, Mt. Lebanon 22 (Tim Cwalina)

Playoff scenarios:

- Losses mean the seasons for Cwalina, Jones, and Burbank are over. Cox's loss was to be expected and it being even as close as it was, is an accomplishment. Jones and Boys' Latin were losers of a 50/50 kind of matchup; for Cwalina, the loss in the WPIAL second round was a mild upset as Mt. Lebanon hadn't lost all season.

- Two games next week will pit future Hoos against one another. DeMatha rematches with Good Counsel for the WCAC title; DeMatha won the regular season battle in the kind of game you expect between the two powerhouses. Norcom and Hampton lost only one game each all season; they face off next week.

- Hermitage faces Varina in the Central Region Division 6 semifinals; Hermitage won the regular season game between the two, 17-12.

- L.C. Bird goes against Highland Springs in the Central Region Division 5 semifinals.

- Stone Bridge faces South Lakes in the Northern Region Division 5 semis.

- The Eastern Region Division 5 semis features the aforementioned Norcom/Hampton matchup, as well as Phoebus vs. Great Bridge.

- In Maryland, the 3A West final is undefeated Damascus vs. Quince Orchard, a rematch of a close September game.

- The DCIAA championship game is H.D. Woodson vs. Ballou.


Basketball extravaganza! The opening game against William & Mary gave us a lot to digest, mostly involving the surprise redshirt decision for James Johnson.

First, the game: it went about as well as you'd hope. UVA fans were hoping to get an answer to the question, "which freshman will make the biggest splash?" That turned out to be Billy Baron and his 5-for-6 three-point shooting, prompting every loyal Hoo fan with a rudimentary knowledge of classic movies to declare that it was "time for the old Billy Baroo."

The admittedly way too early conclusion from the game, at least as far as point guard is concerned, is that the position is well under control. Most of the team played well, as you'd hope against a rebuilding CAA team, but Baron's shooting and the overall play of Jontel Evans (3-for-4, 7 assists, 1 turnover in 31 minutes) meant that for one game at least, point guard was in good hands. (Baron is going to be on the court with Evans a lot, so he's not solely a point guard.)

But the bigger news is Johnson's redshirt. A redshirt for someone shouldn't come as a surprise. Too many candidates for the rotation, not enough spots in it. With Zeglinski out and Johnson redshirting, that leaves 10 players, and I think you'll see all of them on the court regularly until Sammy returns, at which point some other decisions will have to be made.

The impact this season is that the second-biggest guy on the team won't see the court. It'll cause matchup problems, to be sure, against big teams. But most of those teams were already going to beat hell out of us, James Johnson or no James Johnson, and they were probably going to abuse us in the post and on the boards, Johnson or no Johnson. Will the decision to redshirt Johnson cost us any wins? My guess is no. It'll probably turn some losses into ugly losses, but Johnson, like the rest of the freshman, is a 99% unknown quantity and therefore reasonably replaceable with other 99% unknown quantities (mainly Will Regan.)

Big-picture-wise, Johnson remains of course technically a member of the 2010 recruiting class, but for practical purposes this gives us the big man Bennett wanted in the '11 class. Only with a year of development in a college program rather than one more year playing overmatched competition in high school.

Tonight: USC-Upstate, which lost 66-35 on Saturday to a young and probably overmatched-in-the-Big-Ten Michigan squad. (Does that sound like anyone you know?) Upstate is probably the second-worst team we'll see all year and should present little more resistance than a speed bump.


In Euro-football news, there's good news and bad news. The good news is a home game to start the NCAA tournament; the bad news is, UVA is unseeded, meaning a road trip if they beat first-round opponent Old Dominion.

ODU is a middling CAA team, earning an at-large pick despite not even earning a bid to the four-team CAA tournament. Like UVA, their biggest strength is in net, and like UVA, they struggle to score. Biggest win and likely reason for entry into the tournament: a 2-1 win over then-#2 North Carolina. Keeper Evan Newton is the CAA's first-team goalie, with stats just slightly worse than UVA's own Diego Restrepo. Neither team is much of an offensive powerhouse, so expect a low-scoring game. That's on Thursday.

The second-seeded ladies face Ohio State in the, uh, eighth-finals on Saturday after eliminating Lehigh and South Carolina. With the region's top seed, Maryland, out in the second round, the road to the final four is laid out clearly before them; if they get there, they'll probably have to try and get past UNC, which as you know is a women's soccer school before it's a basketball school.

It's a biggish week for UVA sporting events, so a quick schedule to keep you kept up on it all:

Monday: hoops vs. USC-Upstate
Thursday: men's soccer vs. ODU, and basketball at Stanford
Friday: field hockey, a Final Four matchup against UNC
Saturday: football at Boston College, of course, and women's soccer against Ohio State
Sunday: men's soccer at Penn State, if they win on Thursday

EDIT: Oops. Forgot the AP Poll thing. It's right here and updated for this week. Two notes:

- Overvotes returned to sanity. There aren't that many. Honestly, if I had to do it again I'd have used standard deviations instead of a single inflexible number, but this is still showing enough interesting trends that the results are useful.

- Central Florida presented an interesting case. Should they be considered an ACC or SEC regional team? I went back and forth and decided on the ACC for the simple reason that the state of Florida has two ACC teams and one SEC. So it's a little more ACC country. I guess. Anyway, it turned out not to matter. Both the SEC and ACC writers overrated them by a huge, huge margin as compared to the rest of the country. So did the ACC/SEC crossover voters. A useful data point indeed. Keep in mind that it's a lot harder to get a big bias number for the 25th team than it is for teams higher up, because there's no way of knowing whether, say, a Big Ten voter thinks they're the 26th best team or the 35th best team. Unranked in a ballot is unranked and by default, is considered a "ranking" of 26th. So the fact that Oklahoma at 19th, for example, gets a bias from Big 12 voters this week of 1.23 places probably means that Big 12 voters really do consider them 1.23 spots higher than the rest of the country. Florida gets a bias of 1.18 places from SEC voters, but it should be more because being ranked 24th means you show up unranked on a lot of ballots.

If rankings went to 30, those voters that didn't rank UCF might pile a few more teams on top of them and their average would go down. (The only way that calculated average of 25.19 would stay the same is if every voter that didn't rank UCF ranked them 26th.) So UCF's number of 2.02 from the ACC (and just under 1 from the SEC, if that were the case) is astonishing.


Grateful Reader said...

Only since you yourself have now raised the subject of using standard deviations for the regional bias analysis...

If I weren't already a Virginia fan I'd be checking your blog weekly anyway just to see the update on this analysis. I've always wondered about regional bias (in basketball too, but of course polls actually matter in football), but have never seen anyone determined/insane enough to sift through the data. You, however, are obviously a tad bit obsessive, which works out nicely for me.

So... I'll just say that if you're obsessive enough to go back and redo your regional bias analysis using standard deviations, I assure you I will not object to seeing the results. I'm grateful for having any analysis, but the effect of the current updates do work out to be a little bit of a tease: there's just enough information to whet the appetite, but with the constraints of the binary definition, not enough information to draw any firm conclusions.

But of course if you don't feel like it, I don't blame you. Score one for sanity.

For what it's worth, I'm really surprised at the marked lack of bias seen thus far. It has me now hypothesizing that groupthink (i.e., when a voter uses last week's *consensus* poll as an input, if not the primary input, for his *individual* votes this week -- e.g. by "laddering" from there) is far more prevalent than I even suspected. Naturally this requires that you design yet another analysis to investigate groupthink tendencies. It's a good thing you have so much free time!

Brendan said...

I'm planning a full end-of-season roundup of these statistics once the final weekly poll comes out (but probably before the "final" poll that comes after the title game.) It'll take a couple weeks to put together - if I do the standard deviation relook, that's when it'll happen.

I would say that actually the results lately do suggest some bias. I'm not sure what a good threshold would be for the final average to reach before it can be called a good indicator of bias, but the results lately have been pretty consistent, so I'm leaning toward saying: yes, regional bias exists, but it pretty much cancels out in the end.