Phil Steele: Michigan's % of Tackles Returning & % Yards Returning (2017)
Phil Steele posts some interesting analysis today, 19 June on the upcoming 2017 football season:
% of Returning Tackles for 2017: Michigan is ranked 128th out of 130 with only 36.5% of all tackles returning
% of Returning Yards for 2017: Michigan is ranked 73rd out of 130 with 62.1% of all yards returning.
But Michigan returns both Don Brown and Jim Harbaugh, so there's that.
Because I'm in a CFB betting pool each year and am a stats nerd... I always look at Steele's data (returning players offense and defense, returning yards, etc.) when I'm making my picks, but only for the first two or three weeks. They are quite predictive of game outcomes (ATS, at least), but really only for those first two or three weeks. After that, they are essentially meaningless, or so I've found. If anything, we need to worry about Florida, Cincy, and Air Force. And since the last two games we will be prohibitive favorites, we really only need to worry about Florida, which we already are...
he is very heavy data and back up things with facts, not some tool like Finebaum etc. It's tough to predict football, and he's actually made some pretty good predictions in the past, though like everyone, some very bad ones too.
The issue with Steele is that he decides in his head what is important first, and then goes looking for that, rather than sorting through and figuring out what is predictive, and using those measures.
% of tackles returning sure seems like it would be predictive, but Steele shows none of the actual legwork behind figuring out how much, instead just haphazardly showing some random examples, the first of which he brings up doesn't even support his theory. This is straight up just shitty science that may guess right some times, even a lot of times, but demonstrates nothing that should be considered reliable.
money with his picks but last year his ATS picks were horrendous
He sucks. I stopped reading him last year when he ranked Michigan's DL 10th in the country. He also had all these stats about how Ohio State was the least experienced team in the country and how they lost all this production then picked them to win the conference because, uh, he didn't say. He also ranked the B1G as the worst power 5 conference in the country, only see to UM, Wisco, PSU, and OSU all finish in the top 10. He doesn't even have a rudimentary understanding of statisitics and correlation vs causation.
I am pretty sure a fair portion of this blog will fire Steele in the annual August tradition of breaking down the more detailed rankings as they slowly trickle out. Like many things around here, you can almost set your watch by it.
Please let us know your findings.
I find your ideas fascinating. Please provide the information needed to subscribe to you bi-monthly newsletter...
Man, that site needs an update.
haha
never said this one didn't!
Steele is an Ohio State fan who lives in Cleveland and his bias always shows through this time of year. Athlon passed his magazine years ago and the best preseason mag in the business.
Wow if this guy is an osu fan I really have to laugh at his b.s.. Some teams actually upgrade at some positions year to year depending on recruiting talent. Michigan's defense will have to grow some as the year progresses, but the offense should come out firing on all cylinders. Speight should be jamming the ball down opposing defenses and breaking their will consistently or he can watch Peters do it. The talent is there, so there are no excuses with a Harbaugh coached team. By years end the defense will be playing elite football, so once again O lets see what you got. Since I believe we will get outstanding production out of the QB position one way or the other this offense will really shine. I would'nt bullshit wolverine fans, so buckle up because we are in for a great ride despite this assclown's assumptions.
If you return all your starters from a crappy offense, how is that a good thing? I have a hard time believing that Purdue's offense is in better shape because their returning more skill position players.
That means. I think it means less when you have elite young talent. Gary alone will elevate this group to a pretty high level. And some of these young studs are going to start to shine this year and next. Gonna be fun to watch!
GO BLUE!
If these were stats that predicted improvement for Michigan, the comments here would be much different.
Percentages of these things represent experience, and Steele is using them to predict whether a team will take a step forward or back. Is he saying Michigan's defense or team will be #128 because they're #128 in returning tackles? No. He's saying it suggests that Michigan's defense is going to take a hit, and I think we all agree. I don't think anyone here can honestly expect a defensive improvement after we lost Lewis, Wormley, Charlton, Gedeon, Stribling, Thomas, Hill, etc.
Is Michigan going to take a step back? A lot of people I see are saying Michigan will be 9-3 or 8-4. We're not better at any one position group, and we're probably worse at DL, DB, WR, and TE.
Should we be? I guess. Based on the spring game, I don't know that we are.
Not sure about better, but one important difference is depth at QB. Last year it turned out there was nobody, really, behind Speight. O'Korn did his job and won against Indiana, but Speight played injured when he wouldn't have needed to if there was a second QB on his level. Think Iowa especially.
This year, if Speight gets dinged, Peters comes in. That's a significant difference, in my book.
So, we've got no shame to blame?