Rushing Success Rate for The Game
Came up with a metric that helped me understand rushing success and wrote a diary about it a few days ago.
The basic idea is - did this play keep us on schedule to move the sticks/score a TD?
For example, a rush for 4 yards on 1st and 10 would be a success because if you replicated that result for each subsequent play on that set of downs the team would convert the first down by an additional two yards on the 3rd and 2 play.
Meanwhile, a rush for 4 yards on 2nd and 10 would be a failure because if you replicated that result for each subsequent play on that set of downs the team would be faced with a 4th and 2. But the following play on 4th and 2 that went for 4 yards would be considered a success.
The results from Saturday are:
Team - 31 Successes, 7 Failures, .82% Success Rate
Haskins - 21 Successes, 6 Failures, .78% Success Rate
Edwards - 1 Success, 0 Failures, 1.00% Success Rate
Henning - 1 Success, 0 Failures, 1.00% Success Rate
McCarthy - 2 Successes, 0 Failures, 1.00% Success Rate
Corum - 5 Successes, 1 Failure, .83% Success Rate
McNamara - 1 Success, 0 Failures, 1.00% Success Rate
November 29th, 2021 at 12:04 AM ^
I like this as an efficiency metric. Thanks.
November 29th, 2021 at 12:17 AM ^
so basically Haskins was our WORST runner.
I actually like your post and metric.... but Haskins wasn't our worst back, so....
November 29th, 2021 at 12:32 AM ^
I would look at that disparity in a different light. I would say that our workhorse back was extremely efficient throughout the game, and the other players' successes are truly reflective of Gattis's skill as a game-planner/play-caller.
The analogy that springs to mind is fastballs vs. off-speed pitches in baseball. UM's fastball (Haskins) was extremely effective. And that set up our off-speed pitches (the other runners) extremely well.
This metric might be used to more fairly compare two backs that get somewhat equal carries. Here, I think it gives better insight into the success of UM's overall run game.
November 29th, 2021 at 12:35 AM ^
Don't misinterpret this statistic by considering the efficiency but not the volume -- Haskins had 21 successful rushes while nobody else had more than 5. :)
November 29th, 2021 at 12:46 AM ^
think about a baseball hitter hitting .400 for a month vs a pinch hitter hitting .500 in 6 at bats. Hassan is the .400 hitter, and the rest didn’t get near enough at bats to be eligible for comparison.
November 29th, 2021 at 7:10 AM ^
Haskins had 28 carries
next highest was 6 carries
November 29th, 2021 at 2:09 PM ^
lol guys I was being 100% sarcastic. my bad should have included /s
In fact imo I think H2 is in the pantheon of UM RB's that include Hart, Wheatley, and Biakabutuka.
November 29th, 2021 at 12:52 AM ^
Interesting stat. Here's another: Haskins ran 28 times. 16 were for first downs; 5 for touchdowns. 75% success rate in moving the sticks or scoring. Hard to imagine any other back (let's set the minimum at 12 carries) would be in that zone.
November 29th, 2021 at 1:17 AM ^
Totally agree. An amazing performance.
I also removed the run at the end of the second quarter because I don't believe it was intended as a "successful" run. Just FYI why I have 27 and you have 28 rushing attempts.
November 29th, 2021 at 7:30 AM ^
That makes me feel better. I thought for a minute there we were going to have to fire Haskins right after we fire Harbaugh.
November 29th, 2021 at 9:32 AM ^
11 first downs and 5 touchdowns.
November 29th, 2021 at 12:55 AM ^
I appreciate what you're trying to do but I think this (like a lot of sports statistics) doesn't really paint an accurate picture of what happened. This attributes the success or failure to the runner when running plays have more to do with the OL.
The magic of Haskins and Corum can only be measured once they encounter the D¸ but even that becomes difficult. For example, if a DT breaks through the line and Haskins dodges him in the backfield, it's all on him. If the OL makes decent blocks but not great push, we could start attributing the success to Haskins once he gets past the linemen. If the OL plows through the DL and then smashes the linebackers out of the way and then the receivers block the secondary perfectly. . .well, pretty much any running back will have success.
Also, while gaining 4 yards on 2nd and 10 may be a failure according to your metric, it's much easier to complete a 3rd and 6 than a 3rd and 10. And maybe that 4 yard gain on 2nd and 10 will set up a counter that will give us a big chunk of yards or a touchdown.
Statistics can be useful but an over-determination to make sports statistics-driven leads to a lot of meaningless numbers that inspire bad decisions like the Mike D'antoni Houston Rockets (cutting out 2-point jump shots because they are statistically the worst shots without considering how defenses will adjust to your predictability). It seems to me that you've presented some pretty meaningless numbers.
November 29th, 2021 at 1:12 AM ^
Agree with many of your points. Please see my diary about the Maryland game to see why this might not be a meaningless statistic. Many people said that UM's running game failed that game. Haskins had less than 100 yards but had a 70% RSR that game. His effort more than likely facilitated much of UM's offensive success. This metric was born because of the disparity between the base stats and UM's obvious success on the ground and how we might understand that disparity better.
Also, I would look at this as more of a measure of the run game as a whole. We want to understand how well we ran the ball that game and understanding how each runner did, even if of secondary importance, is relevant. You can't divorce runningback play from OL play but we all know that each running back is not equal. As stated above, this metric should only be used as a comparative statistic between ball carriers when they have a similar amount of carries.
November 29th, 2021 at 1:44 AM ^
Heard Bud Elliott talking about how shocked he was the degree to which Michigan "physically whipped" Ohio State. Cited success rate and leverage rate as the kind of stuff you hardly ever see in a conference game (and more for like Alabama-New Mexico State type of matchups). Michigan was basically never off schedule or behind the chains
November 29th, 2021 at 6:07 AM ^
I think you've misplaced a decimal point in your percentage rates. eg 100% not 1.00%
November 29th, 2021 at 6:23 AM ^
Haskin's Success Rate at wearing down the defense? 100%! By the middle of the 4th quarter, OSU's entire defense wanted nothing to do with him.
November 29th, 2021 at 6:23 AM ^
I think this is already a metric but not sure where you can find it. I think Andy Staples talked about it on his podcast after the game and I think they said it was something like 67%(?) which was an absurdly high number for OSU to give up and tells the tale of why we whooped their defense
November 29th, 2021 at 7:11 AM ^
Cool stats
scoreboard is all that really matters