Monday, August 04, 2008

News and Notes, and a start to the Big East Preview

In case you didn't get a chance to see it, ESPN.com has a Shootaround focusing on the Big East. There is a team capsule for every team in the league and a feature article about the new, revised BET that now includes every team in the league. There are also vignettes about Keno Davis (Providence), AJ Price, the difficulty of exiting the BE basement, Shrek, and the likelihood of getting nine teams into the NCAA tournament. Also, Jay Bilas picks Marquette to finish fifth.

A fellow stathead at Villanova by the Numbers has an entry about Returning Minutes Experience. Instead of breaking down the total returning minutes, greyCat looks at the quality of experience. It's an interesting take on the quality of returning players, and it's not just because he ranks Marquette #1.

====

Given all that, I wanted to take my own look at a Big East Preview. We will consider the following areas based only on Big East statistics. Today will just be a quick introduction to the way I'm approaching the analysis. It's obviously not infallible. If it were, then where would be the fun of a preview? Onto the stats being considered:

#1 - A team that underperformed or overperformed based on last year's stats

Based on a team's Offensive and Defensive efficiency during conference, they should end up with a projected win-loss record. Often, the projection will end up not as a nice even integer, leaving the final results as a record of coaching ability (or lack thereof) and/or luck. Sometimes, a team will end up outside the standard deviation as a significant underperformer or overperformer. In particular, these are teams we will watch closely because we expect them to return closer to form.

Example - based on last year's stats, Notre Dame should have finished with a record of 11.7 - 6.3 (0.648) in conference. Instead, they ended up with a record of 14-4 (0.778). That gets a red flag.


#2 - Consistency (or Inconsistency) of the team last year


Consistency is both good and bad. If you're a bad team, it's not very acceptable to be consistently bad. On the other hand, if you're a good team, it's great to be consistent. I'm really looking more at inconsistency than anything. The best way of looking at consistency is that inconsistent good teams lose more than they should and inconsistent bad teams win more than they should. Teams that return a large number of players should become more consistent.

Example - Marquette was the most inconsistent team in the Big East last year. They also finished under their projected win total. This year they are extremely experienced and may end up more consistent.

#3 - Quality of Junior and Senior players (Using data provided by Villanova by the Numbers and then modified)

As mentioned above, Villanova by the Numbers had an interesting look at the quality of returning minutes. It basically factors in the team's percentage of returning minutes while looking at if those players are also Seniors. I modified the percentages slightly to include Junior and Senior returning minutes. This is obviously not a perfect way of looking at returning minutes, but I think a team playing with mostly Seniors and Juniors will beat an equally talented team playing with mostly Freshmen and Sophomores.

Example - UConn returns the highest percentage of Junior and Senior minutes. This is a big bonus.

#4 - Regression of a team towards 0.500

There's a statistical theory in Dean Oliver's "Basketball on Paper" that says a team tends to regress towards 0.500. Intuitively, this makes sense. It's pretty darn hard to stay at the upper echelon of the league on a consistent basis. Players leave, and no matter how talented a team or coach may be, eventually that team gets pulled towards 0.500. It's a mark of a good coach to keep the team performing consistently above this trend. Likewise, it's also hard for a team to stay bad for a consistently long time, if for no other reason than the team can get a new coach and offer players the promise of playing time.

Example - Providence should have finished with a record of 7.1 - 10.9. Based strictly on the numbers, PC should expect a 9% increase in win % for a record of 7.8 - 10.2. Considering that the team actually ended up 6 - 12, PC is looking at a two game bump just based on stats... and the makings of a "Keno Davis Turnaround Story" sometime in Feb.

#5 - Quality of incoming players based on RSCI

I'll fess up... I didn't feel like looking up every single team for notable transfers and incoming juco players. So instead I went on the basis of the RSCI Top 100 list to see which teams were bringing in consensus top 100 talent. In a lot of situations, this can offset the loss of starting minutes. However, as seen by our beloved Marquette team, which introduced two juco players that expect to compete for playing time, there are some limitations to using this format.

Example - Georgetown lost a lot of quality players last year, like some guys named Hibbert and Wallace. However, they offset that by adding the RSCI #65, #66, and #79 players as freshmen. That's a big bonus.

====

For every section, I compiled the stats for last year's conference play, and then tracked teams on each category. Special attention was paid to situations where a team was more than a standard deviation above or below the mean. Using this criteria, we'll tackle each of the teams in the Big East to see where the various teams may be influenced.

4 comments:

  1. I think ND may be in for a very disappointing (relative to expectations) season, and those stats help confirm that for me. On most top 20 teams, the loss of Rob Kurz would not be all that significant. However, their lack of a replacement for him through two years of awful recruiting makes Kurz's loss very painful. Combine that with a much more difficult schedule and I can easily see ND in the 10-8 or 11-7 range in Big East play.

    ReplyDelete
  2. It is nice to see someone put some thought behind a pre-season prediction rather than the standard dart-throwing.

    Have you considered looking at the teams past performance in relation to your predictors? For example, has a particular team not followed a trend towards .500 because of their coach, monetary support, etc.? Or has a team under or over performed each year over the last few years because of an oddball scheme or coaching decisions?

    This might influence the weight you assign to your predictors for particular teams.

    ReplyDelete
  3. Oliver, it's a good suggestion. However, it was a lot of work to pull the data together for all 16 teams and I don't know if I feel like doing it for the 06-07 or 05-06 years. I'll definitely track the predictions against actuals this year.

    ReplyDelete
  4. Great work, Henry. I was thinking like Oliver too. Thanks again.

    ReplyDelete

Disclaimer: We welcome alternative opinions on CrackedSidewalks. However, this is not an open forum without moderation. If what you post fails to be intelligent or productive, we reserve the right to remove your comment from publication without hesitation.

Anonymous comments will be scrutinized.

The opinions, beliefs and viewpoints expressed by forum participants on this web site do not necessarily reflect the CrackedSidewalks Team.

Note: Only a member of this blog may post a comment.