Tuesday, August 30, 2005

Hart to Hart

Brian over at MGoBlog checks in to dismiss the notion that Michael Hart "disappeared" during the OSU and Michigan games because the Michigan O-line couldn't hack it. He chalks it up to OSU getting out to a big lead and forcing Michigan to throw.

"Bizarre, since Hart went for 99 against Iowa's fabulous front seven and broke 200 against Purdue's equally stout D. Michigan lost in Columbus because it yielded two huge, time consuming 90+ yard scoring drives and a Ginn punt return touchdown. Hart was kept under control early, but by the time Michigan got the ball back on offense later he was a nonfactor because of the score."

I humbly disagree. Here's why: he's wrong.

First of all, outright yardage (the measure used in the article) probably wasn't the best way to measure Hart's disappearance. Yards per carry is. Here are Hart's YPC stats by game, with the games singled out by Brian in bold.

Miami, OH: 6.7
Notre Dame: 3.4
San Diego State: 4.8
Iowa: 3.8
Indiana: 4.0
Minnesota: 4.6
Illinois: 5.8
Purdue: 6.2
Michigan State: 6.8
Northwestern: 6.6
Ohio State: 3.4
Texas: 4.0

First of all, we can probably discard the first two games, since Hart played only sparingly.

Now, it looks to me like Hart was not nearly as effective in the final two games (as the article suggested) as he was during the previous, and somewhat improbable-- as in "unsustainable"-- four games. If someone wants to dispute that conclusion based on the numbers above, feel free.

It also seems to me that Hart really didn't play a spectacular game against Iowa. Decent? Sure. But 3.8 yards per carry is not outstanding by any means. Frankly, it's pretty damn average.

Really, only one of those four games lies significantly outside the realm of expectation that's created by the others.

It looks to me like Hart did exactly what one of the commenters in that post accused Ted Ginn of doing-- fattening up against crappy teams. I'll follow his example and take out Hart's performances against Michigan State's defense, Illinois' defense and Northwestern's defense (non-bowl teams, all) and the Purdue game looks like a downright fluke.

You tell me which number doesn't belong: 4.8, 4.0, 4.6, 6.2, 3.4, 4.0. Really, I don't think you can say that Michigan consistently ripped off big running games against good defenses. Hart had a good game against Purdue. Hart did not have a very good game (especially relative to the rest of the season) against Iowa, Ohio State and Texas. This would seem to suggest that the original statment "Hart disappeared against OSU and Texas" is significantly more accurate than Brian would have you believe. If you want to make it more accurate and change it to "Hart became a very average back against most good defenses," then I think the facts would support that as well.

Now, to the idea that OSU's long drives and quick scoring took Hart out of the game because they forced Michigan to throw. Yes, that played a role later in the game, but it wasn't exactly 28-0 in the first half, either.

Michigan had eight drives and 38 plays under its belt before OSU opened a lead of more than seven points in the third quarter. (Info from here.)

Even after that, Michigan got the ball back with 9:56 left in the third, down only 13 points (hardly panic time) and ran another six plays before OSU stretched it to 34-14.

So that gives UM a total of nine drives and 44 offensive plays before OSU really took the rushing game away.

So what did Hart do with all those chances?
First drive (down 7-0)
5 yard rush
1 yard rush
Total: 2 rushes out of 9 plays (22%)

Second drive (tied at 7)
11 yard rush
10 yard rush
1 yard rush
3 rushes, 6 plays (50%)

Third drive (up 14-7)
6 yard rush
1 rush, 3 plays (33%)

Fourth drive (tied at 14)
4 yard rush
1 rush, 3 plays (33%)

Fifth drive (tied at 14)
Rush for no gain
1 yard rush
2 rushes, 3 plays (67%)

Sixth drive (down 17-14)
6 yard rush
1 rush, 5 plays (20%)

Seventh drive (down 20-14)
1 yard rush
1 rush, 3 plays (33%)

Eighth drive (down 20-14)
2 yard loss
1 rush, 6 plays (17%)

This is where we should see some significant change in Hart's usage, according to the "we fell behind" theory (he singles out Ginn's punt return, which happened here).

Ninth drive (down 27-14)
Rush for no gain
5 yard rush
2 rush, 6 plays (33%)

Tenth drive (down 34-14)
2 yard rush
2 yard rush
2 rush, 6 plays (33%)

Eleventh drive (down 34-14)
3 yard rush
1 rush, 4 plays (25%)

Twelfth drive (down 34-21)
This drive started with 7:51 left in the game
No carries
0 rush, 4 plays (0%)

Thirteenth drive (down 37-21)
5 yard rush
1 rush, 12 plays (9%)

Fourteenth drive (down 37-21)
No carries
0 rush, 4 plays (0%)
End of game

When the teams were separated by a touchdown or less, Hart carried the ball on the following percentage of plays: 22, 50, 33, 33, 67, 20, 33, 17.

Once OSU took a lead of more than a touchdown, it was: 33, 33, 25. That's hardly inconsistent with the previous part of the game.

Only with eight minutes left in the game did Michigan truly abandon the run: 0, 9, 0.

So if you want to base your argument on the fact that Hart only got to run for 52 minutes of the game, so be it.

I would suggest that if in fact, Hart's carries were limited at some point, it probably had a lot more to do with the fact that he wasn't getting anywhere on the ground, rather than the score of the game.

By quarter...
1st: 6 carries, 34 yards (5.7 per)
2nd: 4 carries, 11 yards (2.75 per)
3rd: 6 carries, 8 yards (1.25 per)
4th: 2 carries, 8 yards (4.0 per)

Going into the fourth quarter, why would you keep giving the ball to a guy with 19 yards on his last 10 carries? (Unless you were up by three touchdowns and just needed to kill the clock)

I also don't see anything in the data to suggest that Hart was getting stronger or running better against the OSU defense as the game wore on.

Frankly, I think all the stuff about OSU's long drives is a bunch of crap. Everything I see here suggests that unless Michigan ran it 50 times and theoretically wore down the defense at some point, that Hart could have carried it on every play during the fourth quarter (or the second, frankly) and still not ended up with good numbers.

Brian also points out that OSU sacked Chad Henne only once. On this point, he's exactly right. Of course, the article didn't say OSU sacked Henne a bunch of times.

"Chad Henne got knocked down time after time and simply did not have time to throw on many plays."

Guess what? Chad Henne did get knocked down a bunch of times, and ended up hurrying a lot of his throws. It even got to the point where even the TV announcers mentioned it. In this case, Brian is right. But the article is not wrong.

As for Garret Rivas, dude missed four extra points last year. He (as pointed out in the article) twice missed two field goals in a game. He's good, but far from automatic.


At 12:51 PM, ny1995 said...

Two responses here.

1. 4 yards/carry in your worst games is not disappearing. I think it's fairly normal for good backs to get 4 yards/carry against the best defenses they play. Heck, that would have been a huge improvement for a guy like Benson or Sproles. Hart was more consistent than Perry was when Perry won the Doak Walker in 2003.

I don't think it's at all unusual to drop about a yard/carry off your average against the best teams.

2. Rivas missed a bunch of extra points early when we were using Matt Gutierez as the holder. Two in the first game, I believe. We later learned Gutierez had enough shoulder damage that he needed surgery. When they replaced him (with Richard, IIRC) Rivas hardly missed again all year. I think there may have been timing issues because of Gutierez's limiting movement.

At 1:32 PM, Tom said...

Rushing yards by BCS-conference backs against OSU last year (limited to guys who carried it 10 times):
McLendon, NCSU: 6.3
Noah Herron, NW: 3.4
Anthony Davis, UW: 4.3
Brownlee, Iowa: 3.5
Green-Ellis, Ind.: 1.5
Tony Hunt, PSU: 4.2
Michael Robinson, PSU: 2.9
Robinson (excluding sacks): 4.3
Jason Teague, MSU: 6.3
Brandon Jones, Pur: 3.7
Michael Hart, UM: 3.4

Looks to me like Hart's performance against OSU was pretty damn average for a Big Ten back. Not "above average". Not "average for a good back." Just average.

When you compare it to the previous month (averaging 6-plus per carry), that seems like a pretty good Houdini act in the biggest game of the regular season.

At 1:43 PM, ny1995 said...

But Ohio State is one game. The other two are not of the same pattern, I don't think. Ohio State had an *okay* run defense, and Hart had a subpar game, so call it disappearing. The whole team disappeared.

But Iowa had one of the best run defenses in the country and he ran for 99 yards against them. Texas had a top 20 rushing defense and he ran for 4 yards/carry, albeit on (for him) limited touches.

As an OSU fan, you may put extra emphasis on performance against OSU, but to Michael Hart that was one bad game in 10 starts.

He had a good game against a great Iowa run defense and great game against a good Purdue defense and a very good game against a Minnesota defense that (at least statistically) was about as stout against the run as OSU's.


Post a Comment

<< Home

order online online online