- Member for
- 5 years 29 weeks
- View recent blog entries
|32 weeks 3 days ago||Mike?||
I always thought I heard "Jake".
|34 weeks 1 hour ago||Speaking of things we still don't get||
Now the Florida game's done, is there time to squeeze out the Ohio State game UFR before this upcoming season?
|35 weeks 4 days ago||Sure, I like playing Notre Dame||
Getting Notre Dame back on the schedule is not so important that it's worth scrapping a series with another good team. I don't like cancellations at all, even if it had been MAC team du jour; sticking a thumb in Arkansas' eye for the sake of resuming some tradition a couple years earlier is just dumb.
|39 weeks 6 days ago||I for one am a fan of Draftageddon.||
And as such I think it's my duty to say so. Haters can remove themselves in a sinister direction.
|42 weeks 1 day ago||I find 9 ideal.||
9 teams means an 8-game round robin; 8 games allows a couple "warm-up"/regional games against local teams from the de facto "lower" G5 conferences as well as a couple OOC P5 teams. 10 or 11 teams is really too many: 9 conference games is inherently imbalanced and I don't much care for it; 10 games would give no wiggle room in OOC scheduling. In any given year you're either lowering playoff/BCS chances by playing P5 teams and risking likely losses... or you're reducing your playoff/BCS resume by *not* playing P5 teams.
If I picked 9 from the current conference, I'd cut it down to Indiana, Iowa, Michigan, Michigan State, Minnesota, Nebraska, Northwestern, OSU, Wisconsin. Other selections might be satisfactory.
But if we're attempting to answer the question, "How do you best schedule 9 games for 14 teams?", the situation is really so kind of ridiculous that you're going to have to make compromises. The current vogue is for divisions, but I prefer to approximate a conference round-robin, and this post was about how to make that work.
The essential problem is that in any given year every team will "miss" 4 conference opponents. This is a small enough number to rotate not less frequently than every other year: the question is how to accomplish that. The scheme in my OP rotates all but one opponent, but other schemes could be devised. For instance:
A team's schedule consists of (a) permanent (historical (Mich-OSU), geographical (Iowa-Neb), or conference-artificial (MSU-PSU)) rivalry, (b) the four teams not played the previous year, (c) four remaining "most similar" teams from the previous season.
Pros: should give every team a plausibly competitive schedule each season; follows familiar reasoning from the professional league; for Michigan, liable to protect games with historical rivals (e.g. MSU, PSU) other than Ohio State as they're these days usually pretty good teams.
Cons: schedules won't be balanced (Purdue is liable to end up with easier schedule that Nebraska, e.g.); risks missing rivalries if teams have very different seasons; scheduling is kind of a pain; schedule is not entirely predictable year to year.
B, Several Protected Games
A team's schedule consists of some (e.g. five) teams played every year and other (e.g. four) teams on a rotating basis. Michigan might play e.g. OSU, MSU, PSU, Minnesota, and Wisconsin every year and the other teams rotate. Similar to divisions, but (a) each team's five don't completely overlap (e.g. Minnesota might play MSU, Michigan, Wisconsin, Iowa, Nebraska) and (b) there are no divisions. The con is mostly that the groups are a nuisance of a combinatorial problem to work out with overlapping vs. non-overlapping opponents. Also the schedules won't be perfectly balanced.
Other schemes could probably worked out as well.
|42 weeks 1 day ago||6 years not necessary||
Over three years, each team you play twice you play once at home and once on the road.
|42 weeks 4 days ago||The goal, which I don't think||
The goal, which I don't think I stated explicitly, is basically to get the next best thing to a round-robin. So for the conference championship you'd have the usual tie-breakers, perhaps: best conference record -> head-to-head result -> record vs. common opponents -> overall record -> point differential -> whatever.
I favor ditching the divisions (certainly) and the championship game, though I recognize that last makes this plan completely unrealistic in the money + eyeballs computation of college football.
|46 weeks 3 days ago||Most of the comments may trend negative||
But I appreciate this series every year. Saves me trying to do my own research on every other B1G team.
|1 year 3 days ago||I don't understand the appeal of North Dakota in the conference.||
Never mind the public relations baggage (which is significant and from what I can tell is deserved), what's the commonality? They're practically as far away again as Minnesota, and nobody, as far as I can tell, likes them.
Do we really want them around just to have them on the scheduleregularly on the assumption we'll win most of the time?
Also Brian made a pretty convincing argument that the 7-team conference is a good idea. If we're adding anybody else, it should be because a current conference member picks up the sport.
Yeah, yeah, they travel well and they've got fans. Maybe that's paramount in people's minds after the B1G tournament disaster. But the first gif to hand, even for Brian trying to make the case, is one of mutual bird-flipping. What about this suggests conference mates?
|1 year 19 weeks ago||My thought on the PI thing||
With the rise of Sparty-esque "hold and interfere subtly all the time and take your lumps a couple times a game when you get caught defense", I'd expect the proportion of bad calls to actually rise, because you end up with refs knowing there's going to be shenanigans there sometimes and sort of guessing when it's actually happening. Fairly or unfairly, our D has got the reputation this year of being that sort of defense, so I think we're picking up a lot of those WTF flags. Would be interesting to get a completely disinterested party (say, a UCF fan - they've got nothing to play for and no real relationship to M) to go through game tape with a super-detailed approach and find out (a) what's being called that shouldn't and (b) if/where the defense is getting away with stuff.
|1 year 20 weeks ago||For this once||
I want to give it to the other team and award it to the Jerrysota flag run-out GIF. Because Kill is a pretty good dude who doesn't deserve this nonsense, and hey, we won, we can be magnamonious about it.
(But I'd have voted for Chesson #1 under normal conditions.)
|1 year 21 weeks ago||FWIW||
Minnesota fans I've talked to have seemed to think the offense and especially Leidner have improved dramatically since Northwestern. How much of that is simply not playing Northwestern's apparently mostly legit defense (and therefore, how much not-better they may look against Michigan's very legit defense) I'm not sure.
|1 year 22 weeks ago||Yeah, but the explanation doesn't wash.||
It adds to the appearance of subjectivity when the explanation of selection is based on "the things I saw and thought I saw" (so, Seth's using his feelings/judgment) and "the things people were talking about.... the plays people bitched about" (implied: "people" means "Michigan fans"). Either one or both together mean we can expect a "Michigan bias" in the selection of plays reviewed even if his analysis of the plays is 100% correct (which I'd be prepared to grant, at least for the sake of argument).
And thus even if Seth's "Sparty homer" check (and which is it, is that a reliable check or should we never take Spartan fans seriously?) agreed with his judgment, the selections still questionable.
And even if the selection were fine, and even if the analysis is 100% correct, the EV is being abused by simply summing these abstracted numbers. It could just be a too-simplistic model (as Seth mentions at the end when he discusses why the last play shouldn't be included), but the caveats, I feel, aren't big enough.
|1 year 22 weeks ago||I'm more thinking of, okay, why these 21 plays?||
You're essentially excerpting some relatively small percentage and saying, "Yeah, this really is everything that mattered". Also I'm not sure where your %called numbers are coming from: that could distort the results pretty badly one way or another and it sounds like they're pretty subjective. And finally, piling up those EVs without regard to time seems implausible.
Part of its comparison to my own feelings, where I didn't feel during the game that the reffing was or ended up particularly biased: sloppy, and I didn't like the way the "targeting" play was called, but that was the only thing that stood out to me.
And finally, complaining about the refs seems like a lame way to try to get over a gut-punch of a loss, anyway. Especially when we've got a clear colossal screw-up all of our own without trying to find anything else to blame it on.
|1 year 22 weeks ago||I'm sort of curious what a similar column would look like...||
if done by a Sparty homer. Your number is so high the only reasonable assumption is that there's a methodology problem somewhere. My suspicion is that that problem is "pissed about the loss and not really being objective".
|1 year 25 weeks ago||I don't know what to make of Maryland's QBs.||
I didn't know they'd brought in Garman - he was a starter (!) for a while for Oklahoma State, and seemed to be one of those inconsistent-but-dynamic types. I'd have thought that made him better than Rowe almost automatically, but what do I know? I'd expect Saturday to feature a lot of handoffs, regardless.
|1 year 26 weeks ago||A meh day from the offense.||
I suppose if I were going to try to put a positive spin on this, I would say that instead of deciding to run up the score (or whatever) by gashing UNLV with outside runs until they did something, the staff said, "Eh, we're winning", and brought out some of the formations/plays they want to have and ran them "live fire"? But if that was the thinking, then it's slightly worrying that they didn't succeed much.
|1 year 34 weeks ago||I'm always conflicted.||
On the one hand, "illegal man downfield" is one of the dumbest rules in sports. On the other hand, it's applied wildly inconsistently, and consistent calling would be better than that.
|1 year 35 weeks ago||What would be helpful||
Can we get a link to a spreadsheet arranged by position - a "depth chart" as it were? So that we can see at a glance e.g. that Ace doesn't have a QB yet (I assume he's assuming that he can pick up QB Indiana in the later rounds) or how the lines are filling out, without having to check through the list to see what position everybody is?
|1 year 35 weeks ago||I'm happy to see this come back.||
After the backlash last year I wasn't sure the proprietors would run it again. Though there's nothing else to talk about at the moment.
|1 year 37 weeks ago||"Maize" isn't really a helpful name to try to determine color.||
Actual maize (corn, the stuff you eat) a more or less pastel yellow (unless dried), which, ick. I don't think anybody's advocating for that right now.
I think the yellow-yellow ("highlighter" or adidas look) actually looks better objectively, but the orange-yellow is far more distinctive and looks just fine.
I really don't much care about the exact shade of yellow. "Maize" (yellow or yellow-ish) and (dark) "blue" is hard to mess up too badly no matter the exact shades; it's the helmet that's distinctive.
|2 years 2 weeks ago||I ...don't hate them?||
The shorts could be better. But the top is great.
|2 years 3 weeks ago||Is there a part 3 coming?||
I ask because there's one apparent glaring fault highlighted by the statistics, which is that Kaep's YPP fell every year under Harbaugh. The NFL-writer-narrative this year was that Harbaugh was "breaking" Kaepernick by insisting on [exact problem not clear], but that seems unlikely given his track record; my instinct is to look for a decline/other problems in the surrounding "pieces" of the team, but (a) I don't know how to go about analyzing that, (b) anyway I barely watched the 49ers this past season, and (c) obviously no position exists in a vacuum.
One concerning interpretation might be that NFL defenses are starting to figure out Harbaugh's "system", if it's fair to say he has a system, and under that interpretation we want to ask whether college guys will do the same. (As a sort of flip side of this argument I'd point to MSU's emerging defensive issues this past season.) I'm not sure whether this adds weight to the argument or is a reason for dismissing it, but Seattle, St. Louis, and Arizona are all at least characterized as "defensive" teams right now.
Of course "struggles vs. really good defenses" isn't exactly a valid criticism of any offensive scheme - no one's claiming Oregon's offense is suddenly in trouble. But I'm curious what you think here. Especially since we're playing MSU and Ohio State every year!
|2 years 19 weeks ago||Well, sure.||
That Iowa and Tennessee (say) were playing in last year instead of Directional State and U of State-Nowheresville is a problem. But that's execution of format, not directly the fault of the format (except that it apparently was spit out by the computers giving the overal #1 the theoretical weakest schedule? I'm not entirely sure how it should work mathematically; common sense says, if you've got four play-ins, they should be for the 4 16 spots.)
|2 years 19 weeks ago||Eh||
68 makes no sense, but 65 made even less sense even from the "publicity stunt" point of view.
Okay, refine that a little bit. If the tournament were really seeded 1-64, then a play-in for the 64 spot, and the 64 spot only, is still silly but is justifiable. As it is, the Big Dance is (except for the "overall #1 spot) basically 4 tournaments seeded 1-16, in both the public mind and reality, so it makes more sense to have either 4 play-in games, or none. (I prefer none.)
|2 years 21 weeks ago||My line of thought||
After the Notre Dame game, my operative theory became that Nussmeier had come in this Spring, looked at what he had, and said to himself, "Self, this is going to take a while," and so went into the year coaching "for next year". Hoke is Brandon's guy, Brandon didn't look nearly so shaky then; it seemed plausible that, say, an 8-4 year with obvious "progress" (whatever that means) would buy Hoke at least 2015. Especially with a win over MSU or OSU.
Michigan has looked fine against bad teams. That's a minimum competence and wasn't there last year (see: UConn). Against good teams - even decent teams - it's been a struggle. The offense to me this year looks like it knows what it's trying to do and can't - missed holes, dropped passes, wild throws, poor blocking, goofy turnovers, the works. The big red flag exception is the Minnesota game, of course, when ... I don't have an explanation. But this is an at least slightly better thing than last year's team which often looked like it didn't know what it was trying to do even when it was working. (And then there were the Notre Dame and OSU games, which... I also don't have an explanation for, but in a good way.)
I have no idea how to try to sift out who's responsible for what problems at this point. I look at the team and my gut feel is that with another year Nussmeier's offense would be at least back to decent even with a new QB, just from familiarity, repetition, and some experience on the line. It was garbage time, sure, but (at least with the rose-colored glasses) the mere fact of a touchdown against MSU is a tiny positive progress.
But I don't think we can keep Hoke, as he's shown no ability to improve a team - the opposite, rather: we need to get rid of him. So it's new coach, new coordinators, new QB, new everything next year, and that's also going to be a pain.
|2 years 22 weeks ago||Okay, thanks.||
Looking at it more carefully, I think where I was confused was that I was assuming the Z was on the LoS - I never realized the "exactly seven" rule.
So if I'm reading it right now, the LT is eligible (except for his number) by being on the end of the line, and the X ditto on the other "end" (which is really out past the hash), the Z is eligible because he's in the "backfield" slightly off the LoS (despite being out wide (until he comes in motion)), the Y (as lined up) would never be eligible regardless of motion, and the "center" of the line for counting purposes is actually the RG?
|2 years 22 weeks ago||So this is where my ignorance shows.||
Not having played football myself, I never can keep receiving eligibility straight. But I thought it had something to do with distance (by bodies, not space?) from the ball - anyway, I can't figure out why on that first play diagrammed the Y is ineligible but the LT is (apparently, by color) eligible. I thought it should be the other way around, but I trust Seth to not have screwed that up. What am I missing?
|2 years 29 weeks ago||Point of Order||
That's Vad Lee, JMU and former Georgia Tech QB. He didn't look so hot at GT, either; I'm putting that percentage chance Gardner's better at well over 70%.
|2 years 29 weeks ago||Little-noted artistic detail||
In "Run blocking!", the streamer on the FG pole blows out marking right where #4 just ran.
It's the details that make things great.