John Heuser postulates in the AA News that Michigan’s defense is one of its worst ever. OK, he didn’t exactly use those terms, but he made his argument in such a way that it’s the easiest conclusion to draw from the article, even for Doctor Saturday.

Heuser raises several valid points, but overall, there’s a little margin for error in his analysis. Most egregiously, I think there isn’t nearly enough emphasis on the offense’s role in the entire debacle. The unit seems to be mostly thrown in as an afterthought:

While a strong defense can help a team’s offense by forcing turnovers, stopping third-down plays and preserving field position, an offense can also support the defense. The Wolverines’ attack is doing little to make things easier on Michigan’s defenders. Michigan is the lowest-scoring, most turnover-prone team in the Big Ten.

They aren’t just not making it easier for the Michigan defense, but this Michigan offense is so epically bad that it’s tangibly hurting the defense’s chances for success.

First off, Heuser doesn’t even mention that, regardless of field position, the Michigan offense has scored 16 points this season for the opposing team. The field position hasn’t helped either, as several opponent scoring drives started well into Michigan territory. And although time of possession isn’t really a cause of team success so much as a result, being on the field a lot can really tire out a defense, leading to an inability to stop opposing offenses. Michigan is near the bottom of the NCAA in that metric, with the defense averaging more than 33 minutes of time on the field in each game.

So let’s look at each Michigan game, and see how many points the offense “helped” the defense give up, and how many the defense helped the offense score. A legend for you, since the first time I showed this to Paul he was confused as hell:

• “Defense -3” means that the defense is being blamed for a score when the opposing offense got the ball in field goal range. This is not entirely the fault of the defense, since the opposing team could have theoretically kicked (and presumably made) a field goal on 1st down. If the opposing offense scored a touchdown, it is still a -3 for the defense, since they could have conceivably held the other team to a field goal. The last four points are on them.
• “Defense +3” occurs when the opposing team drives the ball on the Michigan defense but misses a field goal. This is 3 points that the defense should be accountable but didn’t show up in the final score.
• “Defense -7” is when the Michigan offense gets scored on. The defense’s reputation shouldn’t suffer if Notre Dame or Toledo can take a Steven Threet mistake back for 7 points.
• “Defense +3” means that the defense has provided the ball to Michigan’s offense in field goal range. The defense earned 3 of the points scored on the drive. Even if Michigan ends up scoring a touchdown, the offense really only earned 4 of the points since, again, they could conceivably have kicked a field goal on first down.
• “Defense +7” is scored when Johnny Thompson takes an interception all the way back, and the defense scores those 7 points all on its own, without help from the offense.

Red = opponent points and Green = Michigan points. As a quick note, special teams plays do not count for this one way or the other, except in terms of scoring plays and field position. I guess that theoretically means a touchdown with no extra point should count for less against the defense, but whatever. “minus” means that the defense isn’t responsible for those points, even though they are being blamed, and “plus” means they should be held responsible for points, although they aren’t being blamed for them. “Plus” means that the defense is not getting credit for points, even though they should be.

Of course, you could do this unscientific study with other Michigan teams as well, and they may have scored more for their offense (2006 comes to mind), but I doubt the offense has ever hurt Michigan’s defensive unit as much as it is this year.

Utah
Defense on the field for 35:54
Michigan TD: 26-yard drive after a Utah special teams fumble.
Utah TD: 75-yard drive.
Michigan FG: 50-yard drive after long KR + Utah personal foul.
Utah FG: 66-yard drive.
Utah FG: 47-yard drive.
Utah FG: 58-yard drive.
Utah TD: Drive started in FG range after INT. Defense -3 points.
Utah FG: Four-yard drive started in FG range. Defense -3 points.
Michigan TD: 33-yard drive due to good defense and blocked punt. Defense +3 points.
MIchigan TD: 31-yard drive following Utah fumble. Defense +3 points.

So, in the first game alone, the offense netted Utah 6 points (and the Utes earned the last 4 on their touchdown drive against the Michigan defense), while getting Michigan three points (and Michigan earned the last 7 by scoring touchdowns on each, but failing on a 2-point conversion on the second).

Miami
Defense on the field for 35:13
Michigan TD: 77-yard drive. Note: this is Michigan’s first touchdown “drive” of the year.
Michigan FG: 4-yard drive starts on the Miami 35 after a RedHawks fumble. Defense +3 points.
Miami FG: 72-yard drive.
Miami FG: 60-yard drive.
Michigan TD: 87-yard drive.

The Michigan defense is culpable for all 6 Miami points, but also earned 3 points of their own by forcing a fumble.

Notre Dame
Defense on the field for 27:48
Notre Dame TD: 11-yard drive after a special teams fumble. Defense -3 points.
Notre Dame TD: 14-yard drive after a special-teams fumble. Defense -3 points.
Notre Dame TD: 63-yard 1-play drive after Michigan fails on 4th down.
Michigan TD: 75-yard drive.
Michigan FG: 67-yard drive after an interception.
Notre Dame TD: 87-yard drive.
Michigan TD: 60-yard drive after an ND kick goes out of bounds.
Notre Dame TD: 35-yard fumble return. Defense -7 points.

Michigan’s offense is pseudo-competent for one game, but the special teams destroys any chance of victory, and the offense gives up a gift at the end.

Wisconsin
Defense on the field for 36:04
Wisconsin FG: missed after a big KO return starts them in FG range.
Wisconsin FG: 23-yard drive starts in FG range. Defense -3 points.
Wisconsin FG: 43-yard drive.
Wisconsin TD: 60-yard drive after a Michigan fumble.
Wisconsin FG: 3-yard drive after a Michigan fumble. Defense -3 points.
Wisconsin FG: 16-yard drive following and interception of Threet.
Michigan TD: 8
0-yard drive.
Michigan TD: 85-yard drive.
Michigan TD: John Thompson interception return. Threet sacked on 2-pt attempt. Defense +6.
Michigan TD: 77-yard drive.
Wisconsin TD: 64-yard drive.

Michigan’s defense was put in bad positions on 3 Wisconsin field goals (one of which the Badgers missed), and scored a touchdown of its own.

Illinois
Defense on the field for 33:06
Michigan TD: 48-yard drive following big punt return(!) by Greg Mathews(!!).
Illinois FG: 38-yard drive.
Michigan TD: 61-yard drive.
Illinois TD: 67-yard drive.
Illinois TD: 84-yard drive.
Illinois TD: 77-yard drive.
Illinois FG Missed: 48-yard drive. Defense +3 points.
Illinois TD: 48-yard drive following a Michigan fumble. Defense -3 points.
Michigan TD: 73-yard drive.
Illinois TD: 68-yard drive.
Illinois TD: 16-yard drive following a Michigan special teams fumble. Defense -3 points.

The defense was put in two tough spots by the offense, giving up 6 points (3 each on two TDs) that they can’t be held responsible for. However, Illinois bailed them out on 3 more points that they should have been responsible for.

Toledo
Defense on the field for 33:08
Michigan goalline interception thrown after a 28-yard drive: Defense +3 points.
Toledo TD: Said interception returned for a touchdown. Defense -7 points.
Michigan TD: 55-yard drive.
Michigan FG: 45-yard drive.
Toledo FG: Missed after a 75-yard drive. Defense +3 points.
Toledo FG: 69-yard drive.
Toledo FG: 9-yard drive after Michigan turns the ball over in FG range. Defense -3 points.
Michigan FG: Missed after a 58-yard drive.

This game has the largest swing of any so far. The defense earned the offense 3 points they didn’t capitalize on, and wasn’t responsible for Toledo’s touchdown or one of their field goals. However, they also should have given up 3 more points, but didn’t thanks to a miss by Toledo’s kicker. If the defense had their way, the score in this game would have been at least 16-6, with Michigan jumping out to an early 13-point lead, which may have caused the Rockets to fold early in the game. Regardless, they had trouble getting the Rockets off the field all game, whether they turned most of those drives into points or not.

Penn State
Defense on the field for 30:19
Michigan TD: 86-yard drive.
Michigan FG: 45-yard drive (defense got the ball to the offense at their own 45).
Penn State TD: 52-yard drive after a good kickoff return.
Michigan TD: 78-yard drive.
Penn State FG: Missed after a 48-yard drive. Defense +3 points.
Penn State TD: 74-yard drive.
Penn State FG: 61-yard drive.
Penn State Safety: Sheridammit. Defense -2 points.
Penn State TD: 50-yard drive after Zoltan’s free kick goes out of bounds.
Penn State FG: 27-yard drive starts in FG range. Defense -3 points.
Penn State TD: 19-yard drive after a Threet fumble. Defense -3 points.
Penn State FG: 56-yard drive.
Penn State TD: 80-yard drive.

So, yeah. Penn State actually had tons of success against the Michigan defense. Only a net of 5 points that the defense shouldn’t be held accountable for, including 3 that Penn State didn’t even get due to a missed FG. On points actually scored, it’s defense -7.

Michigan State
Defense on the field for 35:18
Michigan State TD: 83-yard drive.
Michigan TD: 18-yard drive following a Stevie Brown fumble recovery. Defense +3 points.
Michigan State FG: Missed after a 48-yard drive. Defense +3 points.
Michigan State FG: Missed after drive starts in FG range.
Michigan State TD: 64-yard drive.
Michigan TD: 65-yard drive.
Michigan TD: 66-yard drive.
Michigan State FG: Missed after a 55-yard drive. Defense +3 points.
Michigan State TD: 63-yard drive.
Michigan State TD: 82-yard drive.
Michigan State TD: 40-yard drive after interception. Defense -3 points.

The defense broke even in this game, and also allowed a ton of long drives. Taking quality of opposing offense into account, it might be their worst game of the year.

Add ’em up
The score of Michigan’s games (aggregate) so far this year:
Opponents: 230
Michigan: 151

Taking out all the net changes that the defense accounted for (or adding those that they should be held accountable for), the score would be:
Opponents: 193
Michigan: 130

Without being harmed by offense and special teams, the defense would be giving up a more-respectable (but still not very good) 24.13 points per game.

So the defense has still earned lots of the flak it’s getting. However, with all the time they’re on the field, their delta in points for each team is Michigan +16. That means the defense’s net points scored for Michigan that they weren’t responsible for (on defense) or were responsible for (by scoring themselves) is 16 points, and that’s even using a fairly harsh metric.

Which outcomes would have been affected? Going strictly by points, and disregarding any swings in momentum, only the Toledo game would have had a different outcome, though the Notre Dame game would have been within one touchdown, easy range against the Irish defense. If you take momentum into account, the Utah and Notre Dame games could have easily had a different outcome, and the Penn State game could have been significantly more competitive. Even still, how much happier would Michigan fans be with a 4-4 record, including wins over Utah and Toledo?

Maybe the defense is being judged just a little too harshly for losing games for Michigan. Still, the big play is too available, thanks to sketchy safeties and linebackers. This defense needs to improve if Michigan is going to have any success at all over their last four games.

Posted under Football

One of the great allures of the bowl season is being able to see evenly-matched teams duke it out on a neutral site. However, is that perception or reality? Of course, the teams aren’t always evenly matched (see USC v. Illinois), nor is the site ever truly neutral (see USC v. Illinois). Fans of the SEC are always quick to cite their good record in bowl games (and any other stat that perpetuates the OMG SEC AWESOMEZ myth). What they are failing to mention is that the SEC has a de facto home game for many of their games. What other conferences are favored? Let’s take a look.

2008 bowl season winning percentage by conference:

Average Distance from Bowl Site (in miles as the crow flies) tells us, on average, how far each conference team had to travel to get to their game. However, this doesn’t tell us the whole story. If two teams are both very far from the bowl in which they will be playing, there is no real home field advantage ceded for either one. To give a better idea of home-field advantage, it would be better to look at how much location may favor one team over the other. Delta distance measures how much closer to bowl site a team is over their opponent. The conference numbers are sums of all of these (in miles as the crow flies). As you can see, the PAC-10 and SEC are heavily favored by this, as is the Mountain West conference. Perhaps not coincidentally, these have the highest win percentages of any conference. Of note is the fact that every single SEC team had a positive Delta distance. That is, they were all closer to their bowl site than the opponent.

Of course, the sample sizes are very small, so it is hard to draw any definite conclusions, but it is apparent that some conferences are favored rather heavily by the locations of their bowls. If anyone would like this data to work with the numbers a bit more, drop your contact information in the comments.

Posted under Analysis

…a post about a team that wasn’t even close to getting there!

Michigan’s 2007-08 season was a disappointment by almost any standards. Not only did Michigan not make the postseason, but the Wolverines were well below .500, and lost to opponents both superior and inferior. A lack of senior leadership (Ron Coleman was the team’s only graduating player) and overall depth (Michigan had only 9ish scholarship players after several Wolverines left the program) were key factors leading to the poor outcome for the team. A lack of players who were capable of running John Beilein’s offensive system, which is heavy on three pointers, was also a stumbling block.

The future, however, is much brighter. Michigan loses only Ronald Coleman as a scholarship player, and has three incoming freshmen who are better fits for the Beilein offense than the players left behind by Tommy Amaker. Ben Cronin, a big man who can shoot, and a pair of shooting guards with long range capabilities in Stu Douglass and Zach Novak spell a brighter future for the team. Additionally, Arizona transfer Laval Lucas-Perry will bring additional talent. Losing Coleman is not a big hit. Though he was recruited as a shooter, he could not hit from the outside consistently enough for Beilein’s system.

Roster Analysis – 2008-09

With 13 scholarships allowed for the team, Michigan will either be able to take one more freshman next year, or bank the schollie for the following recruiting class and give it to preferred walkon Eric Puls. Please note that the roster on MGoBlue is terrible, so this may not be entirely accurate.

Minutes Played: Top 4 return. (Harris, Sims, Udoh, Grady).
Points Scored: Top 6 return. (Harris, Sims, Udoh, Grady, Gibson, Wright).
Steals: Top 5 return. (Harris, Sims, Udoh, Lee, Grady).
Blocks: Top 6 return. (Udoh, Sims, Gibson, Harris, Wright, Shepherd).
Assists: Top 3 return. (Grady, Harris, Lee).

The main scary part is assist-to-turnover ratio. While the top two (Kelvin Grady and C.J. Lee) return, Coleman was third, and at 1.25 the only other player with more than one assist per turnover. Fans can take heart in the fact that two of the top three Michigan players in terms of turnovers (Manny Harris and Kelvin Grady) were just freshman, and it stands to reason that they have nowhere to go but up.

Posted under Basketball

If you really want to know about tempo-free statistics, you should really head over to Basketball Prospectus. Anything and everything you would ever want to know about the statistical analysis of basketball is there. Since the Big Ten Wonk left his blogspot home and moved to BP, there hasn’t really been a focus on the statistics of Big Ten Basketball lately.

I’m just starting to get into this and I haven’t really gone too deep into the statistics. For those of you who like charts (and/or AJAX filled Web2.0 applications) I compiled some basic tempo free stats online. You can see them here. The conference sheet is sorted by Possessions and the U of M sheet is sorted by effective field goal percentage.

If you want a quick crash course on tempo-free statistics, I recommend checking out the tutorial on Big Ten Wonk.

From looking at Michigan’s stats compared to the rest of the conference, I’m cautiously optimistic. One of the biggest complaints I had from the Amaker era was the frequency of turnovers, which killed any chance of Michigan winning. Michigan this year is number 3 in the conference in Turnovers per Possession (TOPP on the chart). Another complaint is that the games were boring. Michigan versus Northwestern did not even pretend to be a basketball game as both teams pointlessly dribbled or passed the ball around the perimeter until the shot clock ran down. Michigan is 3rd in the conference in possessions behind only Purdue and Illinois. This does not mean they are good, but it’s more a matter of personal preference. I like a smooth, flowing game; the change is like going from Mike DeBord to Rich Rodriguez.

Where Michigan is crashing and burning is shooting the ball. They have the worst effective field goal percentage (eFG%) in conference. eFG% even appropriately weights the benefits of 3pt shots and it still doesn’t bring them up. Michigan is shooting about 46% when the median is 50%. There could be a silver lining on this horrible showing. Michigan is 4-1 when their eFG% is greater than 50%. It isn’t ridiculous to think that once Beilein gets his players in and really gets the freshman and sophomore class shooting well this team will contend.

Beilein has stopped a lot of the bad the Amaker left him. He just needs to get his players playing his system and things should look up. If Michigan shoots well, they can beat some good teams (see Ohio State). I think most of the pain from this season is coming from square pegs meeting round holes.

I’ll try to dig a little deeper through the week and look up some defensive numbers. The spreadsheet is saved on GoogleDocs, so if anyone wants to put together the team table for their (or any) team, let me know and you can be added as a collaborator.

Posted under Basketball