Unverified Voracity Summons Chris Hansen

Submitted by Brian on March 7th, 2011 at 4:13 PM

So there's this.


There you go.

Screening. Very cool article from Mike Rothstein on the increasing use of ball screens and pick-and-roll in college basketball going all the way back to the days when LaVall Jordan was helping run it at Butler. It comes complete with pithy epigram:

The ball screen forces defenses to choose where they want to recover.

John Beilein has started using it frequently, getting Jordan Morgan a wide array of dunks and others various open shots—I wonder if that's Jordan's influence? Here is where we compare and contrast Beilein's program reboot after last season with Rodriguez's defensive flailing. [comparison] [sadness/frustration] [basketball team swept state] [woo]

Literally less than nothing. I was away when SI came out with a story about college football criminals heavy on the research and light on the context. The blogosphere duly blew it up. I'm with Braves & Birds in that I'd rather have a big media organization doing research instead of, you know, not doing it, but I'm also with Orson when he rips it. Two main takeaways:

  1. Journalists are terrible with numbers. It's appalling. I bet there isn't a journalism program in the country that requires a statistics course. They are the equivalent of dog groomers once you bring out a decimal point.
  2. Journalists will not stand for doing a lot of research and declaring "nothing to see here."

SI found nothing but still made the monkey dance:

Of those seven percent, "nearly 60 percent…were guilty or paid some penalty". If we assume "nearly 60 percent" means 57% (shockingly, the actual numbers and survey methods aren’t given), then 4% of players on top 25 football teams have been actually convicted of, or plead guilty to, a crime.

The number of average college students with the same criminal record? According to this article from Corvallis, Oregon’s Daily Barometer, 3.45%. That’s right: Your typical college football player is one-half of one percent more likely to have a criminal conviction. To put that in perspective, a team of 85 players has half a person more convicted criminals on it than a sample of 85 students drawn randomly. Hide yo kids, hide yo wife.

"Nothing" is actually generous. Consider that the kids on college football teams are disproportionately male (duh), black (45% as of 2006), and poor (presumably, right?) and that male, black, and/or poor groups tend to have more criminal activity. SI really discovered that putting someone on a college football team is a good way to keep them out of trouble. Which, duh. You're giving them something to lose.

Braves & Birds criticizes a lack of "solutions" in the SI problem, but how do you solve the opposite of a problem? (Other than hire Greg Robinson.)

BONUS: Remember the Free Press going ape that Michigan didn't do a juvenile background check on Demar Dorsey? Yeah

…when the nut graf of the piece mentions that only two out of 25 programs conduct background checks on their incoming recruits, there's two instances of serious slippage here. First, programs probably don't do them out of negligence and cost, not because they know that juvenile records searches are sketchy business at best. Second, they assume this means anything when they also write this in the middle of the piece:

Nor did SI and CBS News have access to juvenile arrest records for roughly 80 percent of the players in the study.

The issue of background checks for most recruits in most states is dead before you finish the first page of the article.

BTW, Feldman's latest features a bunch of quotes($) from coaches and administrators citing the same problems bloggers did.

The way it had to end. MSU's hockey team did get swept in Fairbanks, ending Rick Comley's career, but it wasn't easy. Both games went to overtime. On Friday Michigan State had a potential game-winner ruled out and suffered a seemingly controversial UAs game winner. This caused an epic fit of bitching on MSU player twitter feeds—Derek Grant hashtagged "awful," "embarrassing," and "disgraceful" in a single tweet—that suggested Comley had complained to his players about the call in the locker room. The disgraceful event: the MSU net lifted up momentarily but was settled on its moorings before the shot was taken.

MSU's season ended the next night with another overtime goal, and thus ends Rick Comley's career. That's karma. This is something beyond it:

Michigan State hockey head coach Rick Comley reportedly was involved in a physical confrontation Friday night in the Carlson Center with Alaska Nanooks fans Robert Downes, a Fairbanks Superior Court judge, and his daughter, attorney Amy Tallerico.  …

Downes, during a telephone interview Saturday, said he talked to Comley after the game. “It was a comment on his complaining about every goal that was scored,” Downes said.

The confrontation reportedly turned physical and Tallerico allegedly was struck. Speaking Saturday night, Tallerico said they exchanged shoves. Her father said she filed a complaint with the CCHA.

I'm not inclined to believe a random fan who dispenses frontier justice over Comley—never been anything but stonefaced in my experience—but for Comley to get into a confrontation with a fan in the last weekend of his career is a weird echo of the Kampfer incident that was the beginning of his end. May it haunt his dreams.

Meanwhile. Other than State getting swept it was a bad week for Michigan on the TUC cliff. OSU and NMU both lost, ending their seasons. Michigan's 5-1 record against them is now gone. Compounding matters, NMU's loss against BGSU sends the Falcons to Yost for a second-round series that can't do much to help Michigan. Sweeping gets them .001 for their RPI.

mfan_in_ohio broke down the comparisons in a diary bumped yesterday, but a brief recap:

  • Michigan is still the last one-seed but lost a comparison against UNO. That will be tough to get back unless Bemidji State starts winning games.
  • Denver lost over the weekend, keeping them behind M. Michigan can probably stay in front of them by doing at least as well as they do but pulling BGSU complicates things. Denver has a much better opponent this weekend and could pass Michigan in RPI if they win the WCHA.
  • Any chance of stealing the BC comparison is gone after the Eagles swept UNH.
  • Miami will be dangerously close to passing M if they sweep this weekend but since one or the other will have to lose it's kind of a moot point.
  • Ferris is safe as a TUC.
  • Lake State can become a TUC by beating ND.

In simple terms, if Michigan wins the CCHA they will very probably be the last one-seed. If they don't they'll be a two.

More dudes. A local newspaper article on 2013 commit Tyler Motte lists offer-type substances:

Motte committed recently to the University of Michigan, choosing the Wolverines over Miami (Ohio), Ferris State, Western Michigan, Ohio State and Michigan State.

It's even more difficult to sort fiction from reality when it comes to college hockey offers since their recruiting cycle is so accelerated, but Miami was Motte's "second choice" so that's probably legit. They're a good team to snatch a recruit from. Knock on wood, but Motte does not sound like he'll give the OHL half a thought. Backing that up: his older brother is ticketed for Ferris.

Michigan continued its run of getting commits from kids who will hit campus after the Mayan apocalypse with 2013's Alex Talcott, a teammate of Alex Kile on Honeybaked's U18 midget major team. He had an 0-10-10 line at the recent Select 15 camp and was the seventh-best forward there according to USHR. All they said was "good hands," though. Michigan Hockey Net has a full googlestalk of Talcott waiting for you; FWIW, The Scouting News claims he's an NTDP "cinch."

This is a bit convoluted. But Simmons's latest column on the NFL is a compelling takedown of the sort of shortsighted thinking that plagues NFL owners specifically and, more generally, anyone who is obsessed with getting the highest Financial Oligarch Pacman score at the expense of the future. That people like Daniel Snyder and Dan Gilbert can own incredibly expensive sports franchises is a condemnation of the whole system. If those comic-sans-deploying, Mark-Shapiro-hiring idiots can make billions of dollars just so they can prove their ineptness in games with a score the idea this is a meritocracy is fanciful, isn't it?

Etc.: Yost introduces $38 "all you can eat" seats. Seriously. Red Berenson will be honored by the Blues today. All Big Ten teams from UMHoops; Morris second, Hardaway third, Morgan and Hardaway all-frosh. Kellen Russell wins a Big Ten championship in wrestling. Even tackles can be too tall.



March 7th, 2011 at 4:38 PM ^

"Journalists are terrible with numbers. It's appalling. I bet there isn't a journalism program in the country that requires a statistics course. They are the equivalent of dog groomers once you bring out a decimal point."

Never mind that we are also talking a statistical sample of 100 teams x 200 players or 20,000 poeple out of 232,000,000 adults or a 0.0086% sample. That is if the journalist tracked the record of all current div 1 players - I bet he only looked at a few 100.




March 7th, 2011 at 7:42 PM ^

In fact, 20,000 people is way, way more than you need ... keeping in mind that there will be a margin of error (which is kind of the point of sampling: you know you're not going to get everyone, so you get a randomly-selected segment and get similar results).

There are plenty of calculators out there (just google "sample size calculator"), but it'll make more sense to pick up the theory behind it first. (It's Wikipedia, but it's close enough to get you started.)

Anyway, at 95% confidence (pretty standard) with +/- 4% accuracy, you need only interview 600 people out of a population of 232 million. To halve your margin of error, you need 2401.

So in an alternate universe where glossy-magazine journalists do the hard work, they could actually have something of a decent sample. (It's not quite the same, though, as football players obviously aren't selected randomly from the population.)


March 7th, 2011 at 4:41 PM ^

Why the hate? He isn't an NFL owner. Him ripping LeBron with a silly font does not put him on the same level as Dan Snyder. As an owner he is generally more quiet and works within the NBA mandated rules.


March 7th, 2011 at 4:43 PM ^

I saw the picture of Denard and Hoke and thought, "YES!  There is some football content coming in this post.  Maybe an article quoting Hoke saying that Denard was 100% definitely going to win the Heisman when he runs for 2000 yards and throws for another 4000 in the super-secret new pro-style, west coast, option, zone-read, spread-and-shred, power I MANBALL offense." 

Alas, it was not to be.  Still love the hockey content, but man don't get my hopes up like that again.

Michigan Arrogance

March 7th, 2011 at 4:46 PM ^

I'm very interested in how that MSU coaching search will go, especially from a timing persepective. it's been a month or so since Comley announced his retirement & their early exit means they may have to wait until other coaches (eg, Seth Appert) finish their seasons.... several weeks possibly.


March 7th, 2011 at 5:07 PM ^

Unfortunately I think one of the reasons that journalists are so bad with statistics is that the average person neither understands nor cares about statistics.  I have a friend who is utterly convinced that driving is safer than flying - simply because an airplane crash is more spectacularly terrible than a car crash - in spite of all statistical evidence to the contrary.


March 7th, 2011 at 9:15 PM ^

Pretend you read an article in consumers reports that says the new Chevy Truck is great and that outbid a sample of 10,000 trucks it was the most reliable out there. You might be persuaded to buy one.
<br>Then, pretend a friends says, "I wouldn't buy one ever. I had one and I had to take that to the shop 10 times in the two years I owned it."
<br>Your friends story will have a bigger impact on your perceptions and your decision making process than the statistics.
<br>This is true for the journalists and the readers out there. Every one of us falls victim to this.


March 7th, 2011 at 7:42 PM ^

You critique the journalist's use of statistics but then go on to quote this: 


"The number of average college students with the same criminal record?According to this article from Corvallis, Oregon’s Daily Barometer, 3.45%. That’s right: Your typical college football player is one-half of one percent more likely to have a criminal conviction."


No! The proportion of convicted criminal football players is one half of a percentage point higher than the proportion of convicted criminals in the full college kids population. This suggests that the likelihood that a football player is have  a criminal conviction is about 14% higher than the likelihood that an average college kid will have one (0.5/3.45*100).  Given the huge sample size, I'm gonna go right ahead and say that's a statistically significant difference.


Not that I want to reinforce the stereotype, but if we're going to bitch about statistics we might as well get them right ourselves. 


Edit: What you write just after about the socio-economic status of football players makes sense though. We should really be comparing apples to apples. 


March 7th, 2011 at 7:48 PM ^

Yeah, I'd really be more interested in seeing conviction rates of specific crimes in football players as compared to similar groups of non-athletes. (That OSU (NTOSU) article mentioned that 60% of the convictions among the regular student population were for driving offenses. I'm not interesting in speeding and such ... I'd rather see violent crimes listed.)

My feeling is that the percentage of athletes with a conviction for a violent crime is much lower than the comparable non-athlete group, and if so, I'd rather see stories like "Troubled Kids Really Do Use Football As An Escape" rather than SI's "hur hur football hur hur crime".