I love New Year's Eve. Even if it's a "one-second holiday", it's one of my favorite calendar events of the year. After all, each New Year is an opportunity for renewal and fresh optimism. Didn't like the last year? Make some resolutions and work to make the next year even better. Weird as it may seem, this post falls in line with making 2009 better. Frankly, I have come to believe that people are expecting too much from Marquette this year. Yet if you look at the projected Pomeroy schedule predictions, we presently see a projection of Marquette finishing 9-9 in conference. The numbers are telling us something different than expectations.
Going 9-9 in conference may be a shocking scenario to some, even inspiring an interesting thread on MarquetteHoops.com. I wanted to take my own look at things because of a few concerns with the way that Pomeroy does his predictions. First, I don't think that his numbers take into consideration such factors as home versus road. It's no surprise that teams play better at home, but the statistics back it up. For example, Marquette averaged 1.06 points per possession of offense last year at home and on the road. However, the defense at home (0.91 ppp allowed) was much better than on the road (1.07 ppp allowed). In fact, only three teams in the Big East last year had winning road efficiency margins (UL, UConn, and Georgetown), and their statistics were still worse on the road.
In addition, I'm pretty sure that Pomeroy also doesn't account for matchup-specific situations. In other words, what if two winning teams match up, but one is good at forcing turnovers and one is lousy at protecting the ball? What if a team is great at getting to the Free Throw line, but it doesn't matter for the other team's defensive efficiency?
Therefore, I ran a model that looked at these two issues in particular with the following results.
Unfortunately, using this model, the end result is pretty much the same as Pomeroy's. The model that was run also comes up with a prediction of 9-9. With a deeper look, the expected win percentages seem about right to me. Look at the percentages in the following categories:
Home wins (high confidence) - Cincy, DePaul, St. Johns
Home wins (tougher games) - Villanova, WVU, Georgetown, Seton Hall, UConn
Home loss - Syracuse
Road wins (high confidence) - NONE. It's the Big East, dummy!
Toss-up games - USF, DePaul
Winnable road games - Rutgers, Providence
Unlikely wins - Notre Dame, Villanova, Georgetown, Louisville, Pittsburgh
Seems about right. Probably the only quibbles I have right now are the likely win against UConn and how low the percentages are for ND, @GU, @UL, and @Pitt. With more data, those percentages should get better, but does anyone really expect us to win those games, except maybe in South Bend?
However, the second area where Pomeroy's prediction is missing is a confidence interval. In other words, a prediction of 9-9 is useless without some sort of range. Therefore, I ran a monte carlo simulation with the predictions that are above.
This story is more doom and gloom. According to the monte carlo simulation:
- There is almost a 2/3 chance that the team ends up with nine wins or less
- Eighty percent likely that the team only gets ten wins
- That means only a 20% confidence that the team even matches last year's win total of eleven wins
- Hoping that the team ends up with thirteen wins and a shot at the Big East Championship? 2.8% chances are the odds.
- Or, if you really want to get freaked out... there's a one in four chance the team finishes with seven wins or less.
Personally, this is a tough pill to swallow. With four 1000-point scorers plus Lazar Hayward on the roster, I had high expectations as well in October, calling for
22-24 regular season wins (we were 22-8 in regular season last year), finish somewhere between 2nd and 5th (11-13 wins) in the BIG EAST, and win their first game in the NCAA tournament.
However, once I began looking at the numbers and prediction model with last year's data, my expectations were reduced to a record of 22-9 (11-7). This was still far less than the MUSCoop community.
Why is Marquette on a trajectory towards a worse finish this year, despite all that experience? I don't have the answer to that question. Our defense is clearly worse, but it's unclear if it's coaching, the transition of systems, the lack of height, or any combination. It's also clear that Marquette (#47 Pomeroy ranking) has not kept pace with the rest of the Big East. There are five teams in Pomeroy's top 10 ranking, and seven teams in his top 30. The rest of the Big East has gotten better, while Marquette is playing worse. It should be no surprise, then, that the predictions are for MU to finish worse.
Sure, maybe there's a chance that the team will start playing better. The team is adjusting to the new system, and we'll get 13 min per game from Otule and Fulce (combined). I haven't given up hope on the 20%, but for now, my New Year's Resolution is to expect less.
Happy New Year
edit: evidently I can't add