The Postgame (which is apparently a sports blog run by Yahoo) has a decent article on the non-sensical risk-aversion of football coaches especially in the NFL. The article is somewhat marred by its single minded focus on Chip Kelly, but it's nice to see people finally realize some of the glaringly obvious stupidity in conventional football play calling.
Article on Chip Kelly and risk-averse football coaches
Although I'm not sure how Chip Kelly is clearly the best coach in college football if Saban is its best "recruiter/defensive mind/man molder".
If Saban is a better recruiter, better "molder of men" and better defensive mind, it is very very hard to still call Kelly the clear best coach in the country. Conceding those attributes to Saban basically means that Kelly has to be such a good offensive mind that it overcomes his shortcoming in those other phases of the game. While I think Kelly might be the best offensive coach in the country, he's definitely not orders of magnitude better on offense than Saban is on defense.
As to your point, I do think Kelly is a nationally elite coach and that he's done a great job at Oregon (at least before the NCAA comes calling any day now), I just don't think anyone has a body of work to challenge Saban as the best coach in the game. If we're looking for a coach that gets a team to play well above what rectuiting rankings suggest, I'm not sure we can rule Peterson out as the best coach in the country (and it opens the doors for Pat Fitzgerald type guys). They're not what they were three years ago, but Boise might still be the most overachieving team in the country.
This is already too long, but I also think Alabama plays above their talent level, which is why they're so scary. USC, Texas, maybe Florida are as loaded as Alabama is going off recruiting hype, yet Alabama would probably still be favored by double digits on the road.
1) Alabama does bring in back-to-back top five classes but so do the other teams I listed. The big difference seems to be that Saban maximizes that talent while USC and Texas haven't been able to these past few seasons.
2) I think Alabama is better on defense than Oregon on offense. The converse of your point is true as well; the Pac isn't known for producing good defenses.
3) Yeah, I have no illusions that everyone in Tuscaloosa is an angel, but Kelly has some pretty significant infractions that aren't really contested. I wouldn't be surprised to see him forced to the NFL after this season.
Oregon's recruiting classes have been #13, 9 and 16. He's not exactly deling with dregs. The Oregons and Washingtons of the world don't need to be fertile recruiting grounds because there's more talent in California than USC can take. Much like Texas supplies whoever is around them. And USC has had a cloud over it or awhile. They're not good. Are they getting Bama talent? No. But then they've so far not beaten the LSUs or Auburns of the world, either. But don't act like they're getting MSU talent, or even Wisconsin talent. Which is why they beat up on the Wisconsins of the world. And the talent increase started right when the rumors of pay offs started.
When you try to use muscles you don't exercise often.
Definitely interesting. I kinda want to dig into the math myself & play around with whatever it is they're using. I just want to look at its insides & see how it works. For science.
On Advanced NFL Stats anyway, there is a fourth down calculator which is based on league baselines - it will put out win probability on the down and expected points based on user input. It would be very interesting to do something similar at the NCAA (or even conference) level and perhaps even the team level in the case of Oregon to see the variation. It would almost make more sense to go by conference if we wanted to be at least somewhat granular, as it is possible to get conference baselines.
Interesting article. I could definitely see Kelly's approach to the game, if he were to take it to the NFL someday, making for some interesting matchups. I did like the example of the introduction of the Wildcat being introduced in Miami - that did make a huge difference and it became a thing. As conservative in its play as the league tends to be, it latches onto successful styles on both sides of the ball fairly quickly.
The issue I see with going for it on fourth down is that many coaches play not to lose their job as opposed to playing the odds. What I saw in high school was crappy teams like mine were more likely to go for it on 4th down and the like. We knew we were likely going to get our ass beat and as such did everything we could to play the odds and hope for the odds falling our way enough times to pull out the upset. In turn though that created a mindset of "Only crappy teams take risks like going for it on 4th." It will take awhile for the fans mindset to replace that idea with "Smart coaches play the odds and show aggression when the odds benefit them."
The other thing that is kind of interesting to think about, lets say it is fourth down and punt it. While the punt was not in your favor, if you know the other coach is also risk averse and you are confident in your defense, there is in a sense a reward for the punt. In that assuming you give up no plays and manage to punt it at least 30 days, you have three sets of down to force a punt of your own. That can be better than giving them good field posistion and perhaps causing the opposing coach to be more aggressive as they smell blood. Basically if the other team has a damn good offense, there is something to be said for not messing with the conservative mindset of the opposing coach and just playing along. At least until you're down by a few scores and need the points. If the opposing coach is content to run and throw short to medium passes, it might be better to punt than get him thinking about aggression and deep strikes.
The odd thing the pure odds calculator sometimes misses the advantage gained by grinding time off the clock. There are lots of situations where a score is nice in that it pretty much ices the game, but at the same time taking two minutes off the clock also ices the game. At that point there is nothing wrong with punting and taking the odds of forcing the other team into a long drive, as opposed to failing on 4th and giving a short field. At the end of the day, you always want to score, but sometimes time off the clock is almost as nice as crossing the goal line. If the other team needs two scores to tie and you can set the clock up so baring some freak occurence they only have time for one score, that's a win.
The idea that trying to get a good punter is "accepting failure" or however he put it. Even the math obviously says that going for it in every single situation isn't the right thing to do. So when you shouldn't, field position matters. You want that team to be put far enough back that the reward for going for it on 4th down is outweighed right back, and that they're not starting on your side of the field (or close) where the odds shift to their favor. So making sure you have a good punter isn't conceding failure; it's increasing your probabilities (by decreasing the oppositions).
a good nfl punter has a 48 yd avg. a poor punter is < 42. they avg about 5 punts per game. that's 30 yds over the course of the game.
a good nfl RB has a 5.0 ypc. a poor RB is < 4.0. give each guy 25 carries, that equates to a 25 yd difference per game.
maybe having a good punter is undervalued.
It isn't too different in the NCAA when it comes to punting. Taking the top 90 punters in division one, they average 5 punts per game and 41.9 yards per punt. The longest average punt is 48,79 yards and the shortest in the table is 34.39 yards, so you could be gaining or losing approximately 35 yards per game versus the average in the punting game depending on where your player is at in terms of skill, strength and production. I would say it is significant.
Vs. a guard over a really good guard, dependent on 4 other guys? Or a receiver who plays every down, but touches the ball 6-8 times a game? By the 3rd round you're not deciding between a sure franchise QB, the next rushing champ, or a punter.
I love Hoke's risk-taking, as well, making many 4th calls to go for it. But, man, I'd love it if he took this approach. That's MANBALL (not technically-in my mind-yes..but MATHBALL in the eyes of the author).
You really wanna take the ball out of Gibbons hands, or I mean foot?
How can we clearly say anyone's the best coach in college football? There are logical arguments at this point to be made for a number of guys, including:
starts considering how to respond. And the most successful coaches are not always those who have the new ideas but those who see their value and carry them into settings where big-time success is possible (RichRod was working out the kinks at Tulane, etcetera).
That's easy: Saban.
The Miami Dolphins strongly disagree...
IMO, Bill Snider is the man. Just look what he has done at KSU. KSU is a school in the equivalent of nowhere, and yet he has coached them to a football power house twice
Last I checked the Dolphins were still in the NFL...
a coach having one good year shouldn't really be considered (possibly) the best coach in college football. What have you seen from BK that would make you include him in this list other than his teams performance this year? If I'm being honest I agree with maybe 4 of the coaches you put there. If one year was the criterea for your "best coach" than I would have had coach Hoke on there last year, but I don't think we are talking about coach of the year here, we're talking about the best coach overall.
The guy did also win two D2 national championships, turn around a perrenial loser in CMU, and take Cincinnati to an undefeated season and their first BCS bowl ever. I wouldn't call that "one good season"
The much shakier names, IMO, are Sumlin, whose best team got destroyed by Southern Miss, and Strong, who is coming off two 7-6 years and hasn't played a ranked opponent yet this year.
Kevin Sumlin, because he had one good year at Houston and has never even won a conference championship? He seems promising, but the number of coaches who have a great year at a smaller school then go on to do nothing when they get "promoted" is endless.
Charlie Strong? Who before going undefeated (so far) this season was 14-13? Don't get me wrong, another guy who's probably a pretty good coach....but not in the "best coach in the land" category.
And maybe Bill Snyder deserves to be up there....but how many of the pundits hailing him now remember that this isn't exactly remarkable for them, and they won 11 games 6 times from 97-03, then won 4, 5, 6 and 7. Was he always a genius, or did he forget and then get it back?
I'd sooner put Les Miles in there, because for all the fun that's made of him, he's been in the title game twice. And say what you will about their performances in bowls, but Stoops has gotten the Sooners there 4 times.
Now, you have a whole other list if you start grading on a curve for how dirty each institution is, but that's another story.
Just a minor correction, Snyder only had 2 losing seasons before he left the first time. He went 4-7 and 5-6 in '04 and '05 before retiring
But he came back in 2009 and won 6 and 7 (not losing; but I never said anything about losing seasons). It's worth noting because he had to rebuild to get back to 11 wins and such. But then how much of the decline was because of pre-retirement struggles is impossible to quantify.
After the tackle, Arizona linebackers rejoiced with passionate fist pumps and jumped on top of each other like frogs mating.
No mention of what an offensive genius Troy Calhoun is and how he's gonna take over the NFL then? I'm kidding, but it speaks to the writers point better than Chip Kelly does. Think he just has a hard on for him. Calhouns been running the option and going for it on 4th down every 4th down for 6 years now. Plus he was already an offensive coordinator in the NFL.
But it's written with such a fanatical fervour, it undercuts its own credibility. Things like the mentioned "best coach in college football." Or when Chip Kelly joins the League it'll start a revolution....except that his Miami example didn't take, hardly any teams use it other than a rare change up, and the coach that got all those wins was canned and is the OC for the Jets horrid offense now. And baseball which has been dealing with these theories a lot longers still has lots of resistance to it.
Not to say that they're not wrong to resist; just that if the League isn't all falling in line behind what Belichick did, who has had as much success in the League as any coach, they're not all going to start falling in lock step behind what some new college coach is doing.
Good article, and I agree with the premise, but it's strange that an article praising probability calculations doesn't actually include any examples of probability. The one mention of a percentage was "99%", referring to the odds that a QB sneak would have worked, which the author clearly pulled out of his ass. If this author thinks those who don't include probability in their analyses are idiots, then he's just called himself an idiot.
especially on the QB sneak comment. but he also says how dumb it was for the chargers not to go for it on 4th and 1 at the 49 early in the 2Q, considering the chargers had one of the best offenses. he then later states that the chargers didn't score again until the 4th quarter, which kind of goes against the "best offense" comment.
83.57238% of individuals that read your comment agree with you
I think Kelly is definitely one of the most creative football minds out there, but this article didn't really capture it. It focuses way too much on his 4th-down decisions when it's his offense in general that sets him apart. He's managed to develop an offense that is very complex in its downfield passing game (moreso than most other spread offenses) yet is apparently simple enough for a first-year starter to grasp right off the bat.
Much depends on the time and place. For example Dantonio's fake punt call in the 4th quarter against UM was genius. Jerry Kill's fake field goal was an asinine call. Dantonio's call was made at a moment of time when people least expect it(late in a close game) wirh a manageable down-and-distance. Kill's call was made at a ludicrous time with an unreasonable down and distance. facing his squad. You don't fake a field goal on 4th and 17, down 7 in the third quarter. Not when your defense is weak and your offense uncertain.
A good coach understands that such risk-laden calls ought to only be made under certain situations. To try trickeration at random moments--while superficially sensible--is not a wise course to take. You risk not only it blowing up in your face(a risk with any such plays) but dramatically altering the the course of a game in your opponent's favor. (See Vincent Smith's pass attempt versus ND)
Knowing when to pull the rabbit out of the hat is a fine art. The douche in EL is very good at this, but most coaches are not so able.
Dantonio made that call on 4th and 9. I thought it was a dumb call that got bailed out by good execution.
Agreed, it worked but it was not a smart call from a risk-reward perspective (a bad call can work sometimes)
But if Minnesota's fake FG had been just a little bit better executed it would have been a TD. The play was brilliant. Nobody was paying attention to the receiver at all when the ball was snapped. It's not just knowing when to use a trick play, it's also having a well conceived play ready to go, and having the players execute it.
Saw a similar article that pointed out the average difference in field position between a conventional kickoff and an unsuccessful onside kick in high school football is about 15 yards.Why would you ever kick deep?
Using stats and analysis to drive decisions instead of traditions from the 1950s? Revolutionary.
I would expect UM, Stanford and other schools with excellence in quants and game theory to expect their coaches to show rational choices. Maybe Borges is secretly developing a huge model with the Engineering School that requires "conservative inputs" before he unleashes his explosive offense next year.