Sweet 16: Purdue-UConn Preview, Analysis, and Prediction

I would say that I think I just died and went to heaven, but Indiana only won one conference game this year so that can’t be true. But I’m still pretty excited right now, and wouldn’t you know it: it has to do with sports; and not just sports, but finding another tool to help analyze sports (specifically college basketball in this case) at ridiculously granular levels.

I am talking about the Game Predictor at ESPN.com, which is powered by the fine folks at TeamRankings.com.

I have not yet had a chance to explore all of the offerings at TeamRankings.com, but for sports and stats geeks like me it looks like a little slice of paradise. I’m on a bit of a time crunch this morning and this week so I’ll probably have to wait until this weekend to really get in there and explore, but I am very intrigued. (Especially since my bracket tanked this year and I obviously need all the help I can get.)

Anyway, bringing it back to today.

This morning, I am going to begin the first of my 8 official Sweet 16 game predictions. I don’t know that I’ll get through all 8 games by the time they tip on Thursday, but I’m going to try. And instead of offering off-the-cuff analysis based on my knowledge and gut feelings (as I did in my Sweet 16 preview), I am going with objective analysis — and using the Game Predictor to do so.

Quickly, how the Game Predictor works:

• You choose up to five team stats from a list of about 40 or 50.
• Then you choose two teams.
• The Game Predictor spits out odds of winning, a confidence level, shows you which team is stronger in each category you have chosen, and then presents historical examples of games that most resemble the one you have chosen.

Here are the five stats I have chosen for my analysis of each of the eight Sweet 16 games. We’ll see how well they predict the Elite 8:

1. Offensive Efficiency Rating (points scored per possession)
2. Defensive Efficiency Rating (points given up per possession)
3. Assist/Turnover Ratio
4. Free Throw %
5. Defensive Field Goal %
6. Note: The system automatically incorporates an adjustment for strength of schedule.

Let’s jump right into the game that kicks off the Sweet 16: UConn v Purdue

UConn-Purdue Preview and Prediction

First, here are the particulars for the UConn-Purdue Sweet 16 game:

• StubHub: West Region Sweet 16 Tickets
• West Regional Breakdown
• Date: March 26
• TV Time: 7:07 PM on CBS
• Site: University of Phoenix Stadium in Glendale, AZ
• Announcers: Dick Enberg and Jay Bilas
• Point Spread: UConn -6 1/2
• Over-Under: 134
• My prediction (made Monday before using Game Predictor analysis): UConn

Here is how the five stat categories stack up between UConn and Purdue (with a screenshot and the results typed out, in case the image has trouble loading):

• Offensive Efficiency: UConn – 1.101 | Purdue – 1.024
• Defensive Efficiency: UConn – 0.890 | Purdue – 0.879
• Assist/TO Ratio: UConn – 1.278 | Purdue – 1.315
• Free Throw %: UConn – 0.676 | Purdue – 0.706
• Def. Field Goal &: UConn – 0.374 | Purdue – 0.389

If you go just by the five statistical categories chosen, it looks like Purdue should win right? Well, here is how Game Predictor sees this game:

• Odds to win: UConn – 75.3% | Purdue – 24.7%
• Most Likely Final Score: UConn – 74.2 | Purdue – 67.1
• Odds to Cover Spread (Purdue +6.5): UConn – 53.4% | Purdue 46.6%
• Confidence Level: 4 out of 5 stars

Obviously Game Predictor is a big fan of the Huskies, and something tells me it has a lot to do with two things:

1. The automatic strength of schedule adjustment included in the model. With UConn being from the Big East, a conference that this season has one of the highest strength ratings of any conference in history, the Huskies and all other Big East teams will obviously have an advantage — as they should.
2. It appears that the model takes historical comparisons very much into account. In each of the historical comparisons listed, the UConn comparison is a very high seed (mostly 1s and 2s) while the Purdue comparison falls anywhere from a 5 to a 13. So the model must provide some kind of advantage for the higher/favored seed.

Without knowing more in-depth info about the model used to generate these predictions, it is hard to analyze it any more than I have. I tried out different stats for this matchup and while the odds and scores fluctuated, the confidence level and ultimate outcome stayed pretty much the same. And honestly, that’s about what we should expect.

UConn is a better overall team and has had a better season than Purdue, and the Huskies have been dynamite thus far in the tournament. While I would not be totally shocked to see Purdue pull off the upset, the Huskies are rightfully a 6.5 point favorite and will most likely be moving on to the Elite 8.

For Purdue to win, they must do the following:

1. Get Hasheem Thabeet in foul trouble.
2. Get to the free throw line and take advantage of their 70%+ free throw shooting.
3. Make shots – if Thabeet is controlling the paint, the Boilermakers will have to hit the 10-15 foot jumpshot, as well as knock down threes.
4. Be efficient on offense. We know that Purdue will play solid D, as they were the 4th most efficient defensive team in the country this year. However, Purdue’s offensive efficiency was only 103rd, while UConn’s was 13th. This is perhaps the most significant statistical discrepancy heading into the game.

Purdue should obviously try to slow down the pace of the game, but they are not going to be able to play this in the high 50s, low 60s. UConn is just too talented and playing at too high a level on offense right now. Purdue needs to be comfortable scoring 70 points and winning against a good team, which could be a major challenge.

As much as it pains me to say this, Purdue is a very good basketball team this year. They embody a lot of what I hope Indiana looks like in the coming years. They are tough, physical, play solid D, chock full of guys who played high school ball in Indiana, and know how to grind out wins.

Against a team like UConn, however, that may not be enough.

I agree with the Game Predictor on this one. Purdue is certainly capable, but it will undoubtedly take their best effort of the season to knock off the Huskies, the most dominant team in the first and second rounds of the tournament. And even Purdue’s best effort may ultimately fall just short of a win.

I hate Purdue but I always root for the Big Ten in the tournament — so either way, I’ll be happy with the outcome of this one. I expect UConn to win, but I think Purdue plays a great game, keeps it close, beats the spread, and convinces a lot of people that they are a top 10 team heading into next season.

Who do you think with win the Sweet 16 matchup between Connecticut and Purdue?

• Purdue
• UConn

View Results

 Loading ...

E’Twuan Moore/Matt Painter photo credit: REUTERS/Brent Smith (UNITED STATES)

Tags: 2009 NCAA Tournament, College Basketball, march madness, Purdue Boilermakers

---------------

If you enjoyed this post, please consider sharing with your friends: