Saturday, October 3, 2009

Northwestern's scoring drives

78 yards, field goal
18 yards, touchdown
5 yards, field goal
10 yards, field goal
11 yards, field goal
78 yards, touchdown

If anybody ever tells you that offenses and defenses can be judged by points scored and points allowed, tell them they are wrong and point them to today's game.

Saturday, August 29, 2009

IC Rankings, Preseason

As the new college sports season starts, I want to introduce a Rankings system I developed a few years ago. I start out by ranking teams in the reverse order and going forward from there ranking them according to the same patterns that I have observed in the Coaches' Poll for much of the past 15 or so years. I'll go into the nerd talk and what IC stands for in the next few paragraphs. Those who don't care, just skip the italicized and check out the rankings. Just don't come crying when they don't look "right."

The idea for these rankings developed somewhere around 2004. The 2004 season was significant for a few reasons. It was the only season in the BCS era where three major conference teams finished the regular season undefeated. Purdue's four regular season losses were by a combined 10 points. I realized after the season that even had the Boilermakers won those four games, they would have suffered the same fate as Auburn and been left out of the dust. Also, in 2004, I was taking differential equations at Purdue and learned the significance of an initial condition. It was then I realized that the Coaches' poll can be approximated as a differential equation with the preseason rankings as an initial condition. Ooooh... nerd talk. I warned you.

I'll give a basic example of a differential equation. Timmy starts driving west at 60 mph for 200 miles. Where does he finish driving? The answer, of course, depends on where Timmy started driving. If he started driving in Kansas City, that is your initial condition and you can now figure out where he ended. Now, how on earth does this apply to the Coaches' poll?

Each week after week one, the Coaches' poll typically moves teams up or down from their previous ranking based on whether they lost or didn't lose. Some exceptions occur for big wins, but that is the general trend. So, if a team ranked #1 in the preseason ranking, as USC was in 2004, they can go undefeated and make the BCS Championship game. This does not hold true for somebody outside of the top two in the preseason rankings (see unranked Auburn in 2004). So "IC" stands for initial condition and the rankings are simply developed to make a point about the significance and lunacy of preseason polls based on nothing but hype and program prestige helping to decide a national championship.

Without further ado, here are this year's preseason rankings, the bottom 25 teams from last year's FEI Ratings.

  1. North Texas
  2. Idaho
  3. Washington State
  4. Washington
  5. San Diego State
  6. New Mexico State
  7. Tulane
  8. Western Kentucky
  9. Texas A&M
  10. Florida International
  11. Utah State
  12. Wyoming
  13. Hawaii
  14. SMU
  15. Indiana
  16. Louisiana Monroe
  17. Iowa State
  18. San Jose State
  19. Toledo
  20. Miami U.
  21. Army
  22. Louisiana Tech
  23. Eastern Michigan
  24. Florida Atlantic
  25. Arkansas State
Also recieving votes: Louisiana Lafayette, UAB, Middle Tennessee, UNLV, Kansas State, UCLA, Kent State, Syracuse, Mississippi State, Colorado, Colorado State, New Mexico, Fresno State

Let the games begin!

Friday, March 27, 2009

Postgame Randomness

  • I was a little put off by some of the rhetoric coming from Jay Bilas near the end of the game. All the talk about Purdue exceeding expectations and being ahead of schedule was pure ignorance. Anybody associated with the program will tell you this year was a disappointment, albiet brought on by injuries. Fortunately, it looks like everybody expected to return will return. Add in some front court depth in the incoming recruiting class and improved health, and things should improve.
  • Lewis Jackson continued to kill the team. The four turnovers were not a killer, however I cannot begin to count the amount of open teammates he passed up before driving to the hoop and missing a layup. Hopefully he has to earn playing time next year, because he certainly didn't this year.
  • Matt Painter clearly believes the way to neutralize a dominant opposing big man is to have our big man fire up jumpers. JaJuan Johnson was not only shooting from out there yesterday against Hasheem Thabeet, but also last week against Jon Brockman. This was also in full force two years ago with Carl Landry against Greg Oden.
  • The game marked a two game (at least) losing streak against teams days after recruiting violations coming to the forefront. You may remember the February 19 loss to Indiana last year just four days after reports of Kelvin Sampson's cellphone habits started to resurface.
  • Along the same lines, how upsetting will it be if some of Connecticut's players were found to be ineligible and they have to vacate this victory? Such a conclusion is unlikely to be reached, but I would bet it would be painful if it was discovered you were eliminated from the tournament by cheaters.

Wednesday, March 25, 2009

Kelvin Calhoun?

Looks like UConn's in trouble for some recruiting violations including excessive phone calls.

Given what happened to the Hoosiers last year when this happened to them, I like Purdue's chances tomorrow a lot more than I did a couple hours ago.

Friday, March 20, 2009

The Unbalanced Schedule

One of the unfortunate parts of the Big Ten basketball schedule is that it is not a true home and home round robin. Such a schedule is rare; only the Pac 10 among the six major conferences can pull it off. The Big Ten, since expanding the schedule to 18 games last year, has taken a step in the right direction to balance the conference schedule. Unfortunately, there are still two games that are not played, and in a year where the conference standings are close, those two games can play a big role in deciding the standings. This, however, was not one of those years.

Still, using the log5 method developed by Ken Pomeroy, let's take a look at the missing opponents on each schedule and figure out who had the toughest schedule in the conference. The chart below will list each team with the team they did not play at home and the team they did not play on the road. We'll crunch some numbers after looking at the teams.

Team H-Miss A-Miss
Michigan State Michigan Wisconsin
Illinois Northwestern Iowa
Purdue Minnesota Indiana
Wisconsin Michigan State Ohio State
Ohio State Wisconsin Penn State
Penn State Ohio State Northwestern
Michigan Indiana Michigan State
Minnesota Iowa Purdue
Northwestern Penn State Illinois
Iowa Illinois Minnesota
Indiana Purdue Michigan

The most common trueism about the Big Ten this year is that it was incredibly balanced with everybody except Indiana having a chance to knock you off on a given night. Given this fact, the teams who didn't face Indiana twice were at an extreme disadvantage. Does that mean Purdue and Michigan had the toughest schedules? Maybe. Below is the average expected winning percentage (using offensive and defensive efficiency and adjusting for home/away) for each set of extra opponents before factoring in the team, with the fancy conditional formatting included as a bonus.

Michigan 0.629
Purdue 0.635
Illinois 0.763
Minnesota 0.803
Northwestern 0.828
Penn State 0.844
Ohio State 0.844
Michigan State 0.865
Iowa 0.892
Indiana 0.904
Wisconsin 0.915

To further clarify what these numbers mean, Michigan's remaining opponents, against average competition, would win 629 out of 1000 games. Wisconsin's would win 915 out of 1000. The teams on top got an easier schedule, since they didn't get to play the easier teams, while the teams toward the bottom got a tougher schedule.

Our gut reaction to the schedule was correct. Purdue and Michigan, by virtue of not playing Indiana, had the toughest schedules. Schedule strength is only half of the equation. It doesn't really matter, for example, how tough Indiana's remaining schedule would have been, as they probably would have lost those games anyways. Similarly, even though Purdue and Michigan had easy remaining schedules, Purdue probably has a better chance of winning both of their remaining games. Below are the chances that each team goes 2-0, 1-1, and 0-2 against their remaining opponents.

Team 2-0 1-1 0-2
Purdue 0.77 0.218 0.012
Illinois 0.578 0.371 0.05
Michigan State 0.421 0.499 0.079
Ohio State 0.319 0.496 0.185
Wisconsin 0.191 0.5 0.309
Michigan 0.151 0.806 0.043
Penn State 0.151 0.494 0.356
Minnesota 0.143 0.711 0.146
Northwestern 0.112 0.627 0.261
Iowa 0.055 0.376 0.569
Indiana 0.003 0.112 0.885

Chances are that Purdue would have won both of their remaining games had there been a full schedule. For that reason, you could say the Boilermakers would have had a legitimate gripe for the regular season conference crown had they finished within a game of the Spartans.

One more quick excercise: I am going to take the favorite of each unplayed game and add them into the standings, to get the "True" Big Ten Standings.

Team W L TrW TrL
Michigan State 15 3 17 3
Purdue 11 7 13 7
Illinois 11 7 13 7
Ohio State 10 8 12 8
Michigan 9 9 10 10
Minnesota 9 9 10 10
Penn State 10 8 10 10
Wisconsin 10 8 10 10
Northwestern 8 10 9 11
Iowa 5 13 5 15
Indiana 1 17 1 19

You may notice that the numbers above disagree slightly with these. Michigan State was statistically more likely to go 1-1 against the remaining competition, but went 2-0 here. This is likely because they were a very very slight favorite over Wisconsin, so the odds of going 2-0 were very reduced despite being favored in both games.

You can see that Ohio State, among the conference's middle teams, got a raw deal from the schedule in addition to Purdue and Illinois. Is anyone at all surprised to see that Penn State had an easy schedule? That team was the definition of "flash in the pan."

When the conference schedule does not allow for a complete round robin, the standings can be a bit skewed. While the Big Ten is unlikely to produce a schedule champion, the same cannot be said of other conferences like the Big East. Remember this when Louisville gets knocked off in the tournament.

Thursday, March 19, 2009

Free Bobby Buckets!

No, I am not advocating that Bobby Riddell get the Lion's share of playing time. However, as a role player, there is one time of the game when Mr. Buckets would come in handy: the end.

When down, his three point shooting ability would be nice to help come back efforts. When up, his free throw ability would come in handy when trying to ice the game. The fortunate part about the pace of play at the end of games is that Riddell can be brought in and out at will on offense and defense, given the numerous timeouts and fouls that stop play. His defensive inabilities and propensity to foul can allow the Boilermakers to sub him out of the game. Now, who should be playing in his place during these situations? The answer is easy.

One of the more frustrating parts of this season for me has been the play of Lewis Jackson, especially on the offensive end. His poor decisions have often led to poor shots, turnovers, and empty possessions. He's an easy target for the opponent to foul during a close lead (54% from the line), and has often made horrible decisions on the offensive end and should not be trusted with the ball in a clutch situation. Purdue would be much better served if Bobby Riddell was playing in Jackson's stead down the stretch of close games on the offensive end.

Tuesday, March 17, 2009

On the Tournament and Big Ten Selection

In seed order, I wanted to take a look at the relative outlooks for the Big Ten teams in the tournament and in some cases how close they were to other seeds.

Michigan State, 2 Midwest
In the short term, I am going to go on record as saying Michigan State is pretty safe this weekend. Robert Morris isn't awful, but they have much of a chance of putting up a big fight. USC is probably their biggest roadblock to the Sweet 16.

As far as their seeding goes, it's hard to tell from the bracket how close they were to a 1 seed. While they got the geographically favorable Midwest region assigned to them, it's possible another team was the #5 overall seed and they would not have been able to balance the regions if the best 1 and 2 seeds were in the same bracket. Oklahoma, assigned to the south, may have been the top 2 seed in the committee's eyes, especially since they received consideration for a 1.

Purdue, 5 West
It worries me a bit that some people who's opinions have been proven wrong time and time again are hyping Purdue as a sleeper. However, this is a team that, when fully healthy, is much better than a 5. The presumptive second round match-up with Washington in Portland is potentially problematic, as they would likely have something of a home court advantage. Connecticut, the lowest seeded 1 seed and missing guard Jerome Dyson, likely awaits the winner. If there is a vulnerable 1, it's Connecticut. I'm cautiously optimistic.

Illinois, 5 South
The loss of Chester Frazier is somewhat overstated. He is solid defensively and his loss definitely hurt the Illinois some on Saturday. However, he turns the ball over way too frequently and we saw them run a perfectly efficient offense against Michigan without him Friday. In short, I don't think he'll be missed too much this week. The big difference between Illinois' draw and Purdue's is in the first round. Western Kentucky poses a bigger upset threat than Northern Iowa. Also, while Purdue is a slight favorite against Washington, Illinois comes in as a definite underdog to Gonzaga, while also playing an essential road game if it gets that far.

Ohio State, 8 Midwest
A bittersweet draw for the Buckeyes. On one hand, they probably deserved at least a 7 seed, so they have every right to feel slighted. On the other hand, they get a virtual home series this weekend in Dayton. They will start as favorites against an overrated Siena team. Should they advance, they will get a prime shot at an upset against top seeded Louisville. While people will point to the fact that Louisville is also within driving distance, you need to consider the fact that the whole state of Ohio is obnoxiously obsessed with their beloved Buckeyes. There will be a partial crowd in Dayton, and that could help Ohio State do the unthinkable.

Michigan, 10 South
Given how much the committee looked at quality wins when considering bubble teams, Michigan can thank their wins over Duke, UCLA, and Purdue. Ed Dechellis, in particular, should take note of John Beilein's non-conference scheduling. Their wins over Duke and UCLA and their close loss at Connecticut, combined with Beilein's history of NCAA success make Michigan a trendy cinderella pick. Combine that with the fact that I felt Oklahoma and Clemson were likely upset victims before the bracket was announced, and I think Michigan has a good chance to surprise some folks.

Minnesota, 10 East
Probably seeded below Michigan because of the head to head sweep, Minnesota probably has the worst chance of any Big Ten team to reach the Sweet 16. The first round presents an interesting match-up with a Team in Texas that is so inconsistent from game to game. If they get past them, I'm not sure Minnesota can handle presumptive second round opponent, Duke.

Wisconsin, 12 East
A common misconception about Wisconsin is that they were the last team in, because of their seed. I'm not so sure this is true. Let's take a walk in the committee's shoes, assuming that Wisconsin was a true 10 seed, ranked below Minnesota, Michigan State, and Maryland.
  • 10 Midwest, Minneapolis: Wisconsin cannot be placed here, because it would set up a second round rematch with Michigan State. Conference foes cannot face eachother until the Elite 8.
  • 11 Midwest, Minneapolis: The most geographically favorable 11 seed for Wisconsin, they cannot be placed here because of Michigan State's presense in the same half of the bracket.
  • 11 West, Kansas City: An 11 seed in the west would have matched the Badgers with Marquette. No regular season rematches are allowed in the first two rounds of the tournament.
  • 11 South, Kansas City: Michigan's presence as a 10 seed here means Wisconsin can't be placed here.
  • 11 East, Greensboro: Minnesota is the 10 here. Wisconsin cannot be here.
When all is said and done, Wisconsin clearly got bumped down to the 12 line. The same can be said for Arizona, as there were Pac 10 teams preventing them from getting an 11 seed. There is simply no way that Wisconsin was seeded behind the likes of VCU and Temple. I would go so far as to say that Dayton could have been the lowest seeded at large team.

While you can argue that Wisconsin gets hurt the most from this, the true loser could be Florida State, who now does not get to play a true 12 seed in their first round match-up. In fact, Wisconsin could be the favorite, not only in their first round game, but also in their entire "Pod." It's a true toss-up between Florida State, Wisconsin, and Xavier.

Penn State and Northwestern, NIT
It would be safe to say that Penn State was one of the teams bumped out by USC, Mississippi State, Cleveland State, and Temple stealing bids. Certainly, their fate this year will be a lesson for a team when they think about scheduling New Jersey Institute of Technology.

Monday, March 16, 2009

Why Purdue Won the Big Ten Tournament

The Boilermakers put on an amazing clinic this weekend, yet their most significant offensive accomplishment was not mentioned as much as it should have. Whether this is because of the fear of an offensive jynx or because making shots happens to be more sexy, I don't know. Something that has hurt Purdue repeatedly in their losses did not rear its ugly head this past weekend. They went on two separate 14 minute streaks without committing this particular offensive atrocity and largely kept them to a minimum over the course of the week. This is something that happened 16 times in the losses to Michigan and Penn State and 12 times in the losses to Michigan State and Ohio State, yet only happened 15 times in the three games this weekend. It should be pretty clear what I am talking about:

Turnovers

When offenses are judged, they are frequently judged on how well they are shooting on a particular night. This, obviously, is only part of the puzzle. When a team doesn't turn the ball over, they will get more opportunities to score and significantly reduce the opponent's chances to get fast break points. The first point was illustrated early in one of this weekend's games (I believe it was the Illinois game) where Purdue had eight more shots than the opponent ten minutes into the game. The second point is especially crucial for Purdue, who relies on one of the best half court defenses in the country.

Now, how impressive is 15 turnovers in three games? To answer this question, I'm going to start by changing the question. Using games as a denominator in sports, like basketball, where the number of chances is dependent on style of play is going to lead to some misleading results. For example, a team that runs up and down the floor and takes the first open shot is going to have more possessions, and therefore more points, and in this discussion more turnovers. Instead, we are going to look at turnovers per possession to reduce the dependency on pace of play. A great reference for tempo-free data is Ken Pomeroy's website.

The Boilermakers had 185 possessions this weekend and only 15 turnovers, good for 8.1%, meaning they got a shot off in an amazing 91.9% of possessions. To show how amazing that is, Purdue is at 17.4% on the year, which is among the best 25 rates in the country. Had they kept it to 8.1% all year, they would have easily beaten Houston's national leading rate of 13.7%.

I know this may sound like a biased thing to say, but if Purdue continues to get shots on 92% of their possessions, they will have a hard time losing in the upcoming tournament.

Intro

For those that haven't read my baseball blog, I wanted to give a little bit of an introduction as to what to expect here.

Who Am I?
A 23 year-old Purdue graduate who avidly follows their sports teams and likes to think I have a knowledge of why they win or lose.

Why start a blog?
Because I've had a bunch of random thoughts about college sports in general and wanted a place to post them. I'll delve into some statistical analysis and pure randomness. I'm not a reporter, just somebody with an interest in sports who likes following them in my free time.

I also just projected the bracket with greater accuracy than all but one outlet listed here. I wanted to partake in that contest next year, and needed a place to post a bracket projection.

Which Sports?
Unfortunately, probably mostly football and men's basketball. While I will be the first to say that volleyball was my favorite sport in school, it would be unrealistic to provide much more analysis than most news outlets on sports other than the Big Two because they are very difficult to follow outside of West Lafayette.

What to Expect?
  • Statistical Analysis
  • Pot Shots at idiotic analysts (I'm looking at you, Digger Phelps)
  • Praise for announcers I actually like (ie. Bill Raftery, Gus Johnson, Erin Andrews)
  • Random pre-game, in-game, and post-game thoughts
  • Thoughts on other teams in the Big Ten and around the country
  • Random pictures of Ed Hightower from here:

I will start with a look back at why Purdue won the Big Ten Tournament, and then take a trek around the bracket, looking for some possible surprise teams.