A Pomeroy-influenced peek into the Pac-10

by bruinhoo on Jan 2, 2009 7:11 PM PST Comment 3 comments

With the start of the Pac-10 regular season now upon us, I thought that another view toward conference play might be in order. Ken Pomeroy's CBB ratings are an informative look at the current state of D-1 basketball.

A quick note to start; while his ratings are, IMO, as good a rating formula as is currently available for the college game, at this early stage in the season, it is not perfect for the purpose of comparing teams into the future. His methodology uses season team performance updated on a daily basis - while as good a way to accurately gauge performance as we have, can present problems when looking into the future, particularly when teams have varying mixes of youth and experience, whose maturity and skill may grow at different rates (for example, UNC have a highly talented and experienced team, having returned most of last years' Final Four team. While they will undoubtedly get better as the year goes by, the positive effect of additional games and practices is likely less for them as for a team composed largely of freshmen, or other first-time collegiate starters). This is likely just a long way of saying that I like the current update of this data to look at a few points of interest heading onto conference play, rather than guessing who is going to win.

As of this afternoon, the projection for Pac-10 play is as follows:

Arizona St: 14-4 (17-1); 9-0, 8-1

UCLA: 12-6 (14-4); 9-0, 6-3

Washington: 11-7 (10-8); 8-1, 2-7

Stanford: 10-8 (10-8); 8-1, 2-7

Cal: 10-8 (9-9); 7-2, 2-7

Washington St: 9-9 (9-9); 7-2, 2-7

USC: 9-9 (9-9); 7-2, 2-7

Arizona: 9-9 (9-9); 7-2, 2-7

Oregon: 4-14 (2-16); 1-8, 1-8

Oregon St: 2-16 (0-18); 0-9, 0-9

The first W-L record is the overall conference record projected for each team, factoring in some level of luck (or lack of) and other variables, and is independent of specific matchups. The second record is the product of the sum of the projected results for all Pac-10 conference games. The third record reflects the projected home record for each team, while the fourth record reflects the projected away record for each team.

An interesting subtext to the dataset is a strong home-court advantage that comes through the projections. Pomeroy gives home-court has a defined, fixed value which may not take into account that some venues give a greater real advantage, but provides statistical stability.

Today's data set projects that Arizona State will win all but one road game (@ UCLA), and that UCLA will lose three (@ ASU, Stanford, UW). No other Pac-10 team is projected to win more than two road games in conference; the Bay Area and Washington schools, as well as SC and Arizona, are projected to win two (all - @ Oregon, OSU); Oregon is projected to win only @ OSU, and OSU is projected to win none

Related to the last point is the projected weakness of the Oregon schools, particularly regarding the ease which Pomeroy sees the rest of the conference successfully navigating the Oregon road trip. The ratings do provide for some wiggle room accounting for luck and other variables throughout the season, but the only home game which an Oregon team is currently projected to win is Oregon hosting OSU. In reality, I cannot see every Pac-10 team sweeping the Oregon swing, but regardless, this is an indication of just how far behind they are from the rest of the Pac.

The pronounced home-away split present in these figures leads me to note a clear division among conference teams; Arizona State and UCLA at the top, the Oregon schools at the bottom, with the rest of the conference residing in a middle tier which currently appears both larger, and more evenly matched than in recent years.

Among the six teams in the middle of the conference, all are projected to defeat each of the remaining middle-tier teams at home, and lose to each team on the road, as well as UCLA and Arizona State, but win at Oregon and Oregon State. The uniform nature of the home-away splits tells me that, as of now, the middle of the Pac-10 is wide open for the taking. A team with room to grow and mature, such as WSU, or continue to learn a new coaches' system (Arizona) can make a strong run through the conference, and into March.

This is a FanPost and does not necessarily reflect the views of BruinsNation's (BN) editors. It does reflect the views of this particular fan though, which is as important as the views of BN's editors.

4 recs | Comment 3 comments

Story-email Email Printer Print

Comments

Good post hoo

Recommended as always.

I think my only caveat here is that we have to be a little cautious about reading into the stats given the kind of schedule other teams have played this season. I know the critics will say look at UCLA’s. But as we have seen over the years Howland has a method to his scheduling. As for Stanford’s scheduling we simply have no idea how Dawkins operates.

Also Arizona’s victory over Gonzaga is looking less impressive every day, while Michigan is emerging as a viable tourney threat out of the Big-10.

These numbers will get really interesting after the first half of pac-10 season is over.

Please keep it coming. Always learn something by reading this type of posts despite the caveat referenced above.

by Nestor on Jan 3, 2009 9:34 AM PST reply actions 0 recs

Agreed

Schedule has a bit soft for other Pac-10 teams. Our only two losses have come against quality opponents, and both losses have been pretty damn close, both at times when our young freshman had not yet quite grasped the college game.

They’re still raw, but we can all see how this team has developed and grown from the losses at Texas and Michigan to what we saw last night. I’ll trade those two losses for a 6 game winning streak in March and that’s what CBH is all about.

by Bellerophon on Jan 3, 2009 11:03 AM PST reply actions 0 recs

Trends

Very interesting post, b-hoo. As you point out, this rating system is updated day to day, so it gives a good sense of where a team currently stands, but does not necessarily project well into the future.

I don’t know how UCLA was rated previously, but it would be interesting to see where we were after the Michigan and Texas games as compared to where we are now, and then where we will be in 4 weeks. I would expect to see a trend of improvement that continues throught the Pac-10 season, reflecting the development of the freshmen and the guidance of CBH.

greg in denver - UCLA guy for life

by gbruin on Jan 4, 2009 6:38 AM PST reply actions 0 recs