Member for

9 years 10 months
Points
7943.00

Recent Comments

Date Title Body
10/31/2018 - 8:23pm Depends on whether there's a…

Depends on whether there's a better game she'd rather be watching. If Michigan is playing Rutgers and the Cocktail Party is on or something, she'll get pretty annoyed.  

08/21/2018 - 4:52pm How new is "new"?  The Hold…

How new is "new"?  The Hold Steady's best stuff from the mid-2000s has a similar vibe, although the lead singer's voice is rawer, but it's throwback rock with a piano backing.  Stuck Between Stations is probably their best song.  

08/21/2018 - 4:34pm New Mitski album, which is…

New Mitski album, which is probably going to be my AOTY.  

04/26/2018 - 3:04pm Barkley

Seth: I'm of the belief he's worth a top five pick. Barry Sanders-level guy.

The thing is, Barry Sanders himself shouldn't go in the top 5 in the modern NFL. The replacement value of a truly elite RB is just not that high these days, and the salaries the top guys get paid bears that out. I don't think I'd even take Marshall Faulk (who added a ton of value in the passing game) in the top 5.

03/22/2018 - 2:14pm Michigan can absolutely lose

but man, I wanna play poker against some of y'all very, very badly.  There's some serious struggling with sample sizes going on.  Absent an injury to a key player or a specific schematic issue like Syracuse's zone, the best explanation for "why did team X have a great/horrible 2 games?" is almost always "variance."  The rest is just constructing narratives around it.  

03/12/2018 - 2:12am This

is just question begging.  There's absolutely no reason the goal can't be to pick the teams that have the best shot at winning the tourney.  That's exactly what I think the goal should be.  Sort by Kenpom et al., admit the top 32.  

03/11/2018 - 11:14pm Believe it or not

Multiple people who have looked at it have found basically no correlation between how a team performs in its last 10 games and how it will do in the tourney:

http://blog.bracketvoodoo.com/post/158321282577/momentum-is-a-myth

http://blog.minitab.com/blog/the-statistics-game/basketball-statistics-…

 

03/10/2018 - 1:47pm Contrarian take

But I'm rooting for UK. I think the committee is likely to hose Michigan with a 4 seed because it'll hate the big 10 as a whole, and I really, really don't want to get Kentucky in a 4-5. I'd like to see them win the SEC tourney and bump up to a 4.

03/09/2018 - 11:07pm They actually

Went 2-3 after he went down (including a win over Alabama). 2-4 if you include the game where he got hurt. Which is a pretty thin reed for the Committee to hang it's hat on. Particularly when we're talking about a guy averaging like 7 and 5, not Kevin Durant.

03/09/2018 - 11:03pm Here's one analysis

http://blog.minitab.com/blog/the-statistics-game/basketball-statistics-…

Basically, it's a version of the hot hand myth, and there's no obvious correlation. 10 games is a smaller sample than an entire season, so will likely be less predictive and more subject to variance.

03/09/2018 - 10:53pm Your argument wasn't

The Committee will seed Auburn lower because it'll decide the injury makes it a fundamentally different team. It was "the Committee will see that they've gone 1-5 down the stretch."

Which is explicitly, expressly not how they do things. And Anfernee Mclemore is not Kenyon Martin.

03/09/2018 - 10:37pm Please see #8

https://www.ncaa.com/news/basketball-men/bracketiq/2018-02-15/march-mad…

The committee no longer uses last 10 as a selection criterion.

03/09/2018 - 10:05pm For the millionth time

The Committee does not look at record down the stretch. Which is good, because there's no correlation between performance in your last 10 games and performance in the tourney.

The Committee looks at overall resume, regardless of when wins and losses came.

02/28/2018 - 1:52pm My issue there

is that every year, the Committee insists it doesn't look solely at RPI, that this time it really is going to consider Kenpom and Sagarin and BPI and various other things.

And then pretty much every year, their unexpectedly high seeds and unexpectedly low seeds line up almost exactly with pure RPI (including last year where Maryland got a ridiculously high seed because of it).  

I'll believe the Committee aren't just a bunch of RPI-bots when they actually show it in practice.  

02/28/2018 - 1:38pm They're 25th on Kenpom

and a 7 seed on the Matrix.  Michigan has 2 wins over teams better than that.  

02/28/2018 - 1:28pm The Committee cares about bad losses

It also cares about good wins.  

People around here consistently overestimate the quality of Michigan's wins.  They only really have 4 to write home about (at MSU, OSU, at Texas, at Maryland).  By contrast, Oklahoma, a 17-12 team that's increasingly in danger of missing the whole damn tournament, has 8 quality wins (USC on a neutral court, at Wichita State, sweep of TCU, Texas Tech, Kansas, Baylor, Kansas State). 

Michigan scores very well by advanced metrics and has no truly bad losses.  But the lack of wins absolutely limits its ceiling, and the people saying they can get up to a 4 just by beating Nebraska are not being realistic.  

02/28/2018 - 1:22pm They all get 98% of the teams

The Committee only throws one or two curveballs a year in that regard.  He consistently rates as one of the least accurate brackets on the Matrix in terms of overall seeding accuracy.  He's not good at this.  At all.  

02/02/2018 - 12:26pm Oh no

What will people do without one of the worst bars on the planet?

Seriously people, just go to the Jug or Ashley's.  

01/30/2018 - 4:23pm Counterpoint

Jemele Hill is good.

01/15/2018 - 1:36pm He's probably using per-possession

rather than per game.  Which makes sense for all the usual tempo free reasons.

01/09/2018 - 11:51pm This is like the slide replay plaguing baseball

there is a moment in basically every slide into a bag where the runner's foot comes of the bag for an instant before his leg hits.  If you slow it down to a frame-by-frame you can call the guy out according to the letter of the rule.  But at that point you're just punishing a player for unavoidable physics.  

Unless a defender rips completely through the ball, the ball is virtually always going to be in the ballhandler's hand for an instant or two after the defender knocks it away.  It's basically unavoidable.  You're punishing a guy because frame by frame replay reveals a physical reality we had all just previously agreed to ignore and could never catch in real time.  

It's beyond me how that accomplishes anything useful for the game or is in the spirit of the rule.  

01/05/2018 - 2:57pm So Raiders

Since you apparently have $100m to throw around on a guy who hasn't coached in almost a decade, you gonna give some of that taxpayer subsidy money back, or . . .?

09/27/2017 - 3:02am This isn't the issue

it's not about people reading "fewer articles per unit time."  It's about advertisers having no reliable way to track the effectiveness of video ads.  The reason everyone is pivoting to video is that traditional clickthrough web ads are like the most transparent form of advertising out there.  Advertisers know exactly how effective the ads are because of the clickthroughs, and the answer is: incredibly ineffective.  

There's no reason to think video ads are more effective, but people who do this say they're far more difficult to track and analyze.  Facebook recently had to adjust its estimates of video engagement massively downward because of a "mistake" it made in the calculations.  "Pivot to video" is just another way of saying "trick advertisers for a while," and I don't see any way it's sustainable in the long term.  In the meantime, it's pissing off its readers, who want good written content (very few people like video, in part because people do a huge chunk of their internet surfing at work).  

07/31/2017 - 12:17am Worth Nothing

"Arguably the Cubs #1 prospect" doesn't meant what it meant three years ago.  He's a back end of the top 100 guy (midseason MLB rankings had him at 92).  Obviously not bad, but people shouldn't think he's Kris Bryant just because he's the Cubs' top guy.  

06/22/2017 - 11:20am Which is why

the one-and-done rule is bad and anti-labor and needs to be eliminated immediately.  

06/08/2017 - 4:31pm Beating a woman

is not a mistake.  It is an intentional crime.  

 

And the NFL's handling of the Ray Rice situation was so egregiously terrible that it caused me and several others I know to stop watching the NFL completely.  That Stoops was marginally less shitty in his own handling of the situation does not mean his handling of it was appropriate or acceptable.  

Edit: also, many, many people (including Orson) raked Stoops over the coals at the time.  To claim this is backdated outrage simply isn't true.  

06/02/2017 - 10:37am Carly Rae Jepsen

Is actually our greatest modern pop star.

03/10/2017 - 3:58pm And

Even though everyone has apparently forgotten it, Michigan had hilariously bad opponent three point luck for the first 8 games of the Big 10 season.  Opponents shot like 55%.  

It's not really a coincidence that the defense improved when threes started going in at a more normal rate (though as Ace has noted, Michigan's been a lot better at preventing opponents from taking 3s in the first place).  

02/16/2017 - 11:28pm 3 Point Defense

There was a big debate around here after the Illinois loss about whether Michigan's opponents were shooting 54% from 3 over a 6 game stretch because Michigan's D was that bad or because of variance.  Hope people remember the last few games (where Michigan's opponents can't buy a 3; Wisconsin was 3-16 tonight) the next time we go on an unlucky stretch.  

02/01/2017 - 8:16pm The dropoff is 2 guys

and one of them is a 3 star kicker.  Their only other 3 star is a low ranked WR out of Texas.  

02/01/2017 - 7:42pm WAR Graphs For Recruiting!

I love this.  Also, I know a lot of it is a class size thing, but the fact that OSU doesn't even have the #1 overall class is just insane given what they put together.  

01/25/2017 - 2:20pm Greenstein is a tool

And he's obviously wrong that it's uncommon.  But that doesn't mean it's not bad.  

01/25/2017 - 2:00pm I mean

the post I'm responding to said that the shoes would be distracting because they're pink.  I genuinely don't understand that at all.  And no, I wouldn't feel weird at all wearing a pink suit one day a year for breast cancer.  Why would I?  

01/25/2017 - 1:50pm Is there

something wrong with bright pink?  Again, I don't see why bright pink is supposed to be any worse than any other color.  

01/25/2017 - 1:46pm Serious Question

Would you be offended if the shoes were a color other than pink (i.e., if the official breast cancer color were, like, slate gray)?  If not, I have . . . concerns.  

01/23/2017 - 3:13pm This is Wheatley's replacement

Sam has been saying for over a week that we would fill the Wheatley opening with a non-rb coach and have the rbs coached by someone on the existing staff.

01/19/2017 - 2:55pm Because

He is 247's #1 player.  He's just not the composite #1.  

01/19/2017 - 1:14pm Yes

And some of that is because they've ranked Michigan guys lower, but a lot of it is about genuine process complaints.  Of the 4 sites, they're the most likely to do an initial evaluation/ranking and then never update it, even when a very low ranked guy starts getting interest from big time schools incommensurate with the initial rating.  

01/16/2017 - 2:39am Carlos Brown

and it's not close.  Michigan people weren't even tracking his announcement because he was supposedly down to Georgia and one other school that wasn't Michigan.  SEC fans had to come on the Rivals board to tell Michigan fans he'd even committed.  

01/14/2017 - 8:36pm And what I'm saying

Is that, according to every advanced stats guy, teams literally do not have control over how well an opponent shoots from three (to the point that they have more control over how well a team shoots from the free throw line--something they have absolutely no control over). What they do have control over is the number of attempts they give up, which we obviously suck at.

And, yes, good teams can go through stretches where opponent shoot a ridiculous percentage from 3, because variance is massive, and a total bitch.

01/14/2017 - 8:28pm The entire season

Is not a big enough sample to normalize some stats. Hell, the entire baseball season isn't big enough for some stats.

01/14/2017 - 8:10pm Do people realize

How laughably small a sample 85 shots is? The variance within that many attempts is massive.

01/14/2017 - 7:41pm Except Advanced Stats Say

That teams have tons of control over how many 3 pointers their opponents shoot, but almost none over how many they make.  So the high percentages are very likely the result of running bad.  

01/07/2017 - 4:55pm To be fair

People like Chris Brown (Smart Football) basically say that  Saban's pattern matching defensive schemes are far more sophisticated than anything anyone else in college is running.  Recruiting is probably the biggest factor in his success, but the guy's a genuinely brilliant defensive coach.  

01/07/2017 - 4:37pm It's not a crapshoot though

The reason Alabama is going to win its 17th national championship in 9 years is because they have the number 1 recruiting class every single year.  Same for USC in the mid-2000s.  And it's one of the reasons OSU keeps beating Michigan too.  Recruiting rankings are highly correlated with college success.  

There are really good reasons not to follow recruiting, and they get stronger every year.  But the argument for following it is pretty simple: it's a pretty good indicator of how strong your team is going to be.  

01/05/2017 - 3:06pm Because

Messing with people is really fun?  

01/04/2017 - 1:40pm Very Much Starting to Think

that Harris isn't as adverse to playing recruiting games as we've been led to believe (not that there's anything wrong with that).  Given how confident both Alabama and Michigan commits/coaching staffs are, seems pretty likely that at least one school is getting fed bad info.

12/22/2016 - 2:33pm Yes

Thinking we have our own share of dipshits clearly means I'm actually this particular dipshit.  

12/22/2016 - 1:59pm You say that

like there's some massive difference between the two sets of commenters.

12/15/2016 - 1:43pm I don't necessarily disagree

But there are enough rumors/takes from enough different people (Farrell, Bartow, Georgia people) at this point to raise real doubts.  Either Wilson is putting up one hell of a smokescreen, or there's genuine reason to think he'll pick Georgia.