Unverified Voracity Surveys

Submitted by Brian on January 14th, 2009 at 4:15 PM

Site note: A UMHoops/MGoBlog joint CIL is tentatively on for tonight's Illinois game. Tip is at 8:30, game is on BTN, CIL gets going about 15-30 minutes before.

Correction: Dennis Franklin wore #9, not #6 as claimed yesterday, in case you were looking for him in the afro-tastic team picture.

Survey results. The survey hit its maximum very quickly; results can be seen here (for most questions) and here (for the open-ended "what do you hate" question). General impressions:

  • Mobile MGoBlog was the big winner in the "new feature" category and will be implemented ASAP. Better integration with MGoVideo was also popular. A unified ticket search came in third.
  • About 50% are registered (FYI: even if you aren't interested in posting, logged in users can customize how they see the blog. You can turn some blocks on and off, change the way comments appear, etc.)
  • About 10% of people who tried to register never get a response. (If this happens to you, email me.)
  • Most people read the board and diaries, with about half participating on the message board and a small number posting diaries, which is about right, IMO.
  • Page speed was mostly "good."
  • People seem to think the level of self-policing in the comments is about right, but they'd like to see better organization of the user-produced content.
  • Advertising is at a tolerable level.

Sorry if you got locked out; I dislike Wufoo's pricing schemes. (I don't want to sign up for something monthly and have to cancel, but I'd pay ten bucks to have a single unlimited survey.)

Clone wars. UMHoops digs out some Kenpom stats and compares this year's basketball team to the 2005 West Virginia team that introduced the world to Gansey and Pittsnogle, et al. The key chart:

mich-wvu-comparison

The offenses are eerily similar and can quickly be compared: Michigan doesn't shoot as well—though they're not bad—but values the ball more than anyone else in the country; they don't get offensive rebounds or free throws, as they are an extreme POT, which you can see by the percentage of three pointers chucked skyward.

Defensively it's a bit tougher. Michigan looks superior in just about every number up there except turnover percentage, but WVU's defense went up against a lot of good offenses. Michigan not so much.

One thing I did find interesting: Michigan isn't actually that bad on the defensive boards: 33.8 is just about the national average. That's still not good, as an average power conference team with 60+% of its schedule to date against mid- and low-majors should have above-average rebounding. I feel like that sentence was very confusing, but am at a loss to fix it. Rephrase: Michigan's probably a poor defensive rebounding team but not a disastrous one.

A side note: there's been some discussion of Kenpom's grim forecast for Michigan—8-10 in conference and 18-13 overall before the Iowa game, now up to 9-9 and 19-12—and what this says and etc. While I think the Kenpom ratings are worth looking at, keep in mind that they can't account for the absence of Laval Lucas-Perry—currently the team's most efficient offensive player—for about 60% of the season. That's probably worth a game or two (or three!) in Kenpom's projections.

Dylan has an array of interesting observations as well; check his post out.

Elsewhere. The Wall Street Journal takes note of the Big Ten's basketball revival, and does so with heavy deployment of tempo-free statistics.

Is it just me or have mainstream basketball writers taken to advanced stats much more quickly than writers covering any other sport? Baseball writers often take pride in their ignorance. Football broadcasts still propose that 3=7 whenever they mention redzone efficiency. Advanced hockey stats aren't yet important enough to sneer at. Basketball guys, on the other hand, took one look at Kenpom and said "hey, that makes sense." Wonder why that is.

Etc.: Rick Reilly declares beer pong the "next great American pastime," causing reader Jeremy Hekhuis to ask "if reilly is calling something the next great pastime, hasn't its time come and gone?" and causing me to respond "yes."

Comments

Undefeated dre…

January 14th, 2009 at 4:47 PM ^

Good point. The survey is probably representative of MGoBlog's most active, loyal users -- because they were the ones most likely to see the link and respond. So the results are good for appealing to those who are already heavy users of MGoBlog.

But the results may not be so useful on how to get more readers in general, or to build loyalty and participation among semi-serious MGoBlog readers.

wolverinekeith

January 14th, 2009 at 5:00 PM ^

Seriously, Brian - is there anyway to allow users to just stay logged in until I choose to log out (or my cookies delete...but that's up to my browser). I mean, I have Scout & Rivals memberships and even they let me stay logged in.

Thanks.

J. Lichty

January 14th, 2009 at 5:13 PM ^

drea - I hope you are not serious about being frozen out by "more loyal readers" than yourself apparently.

I am sure that Brian will read your brilliant suggestions of how he can generate more traffic which you apparently did not get to share in the poll.

Perhaps Brian could run another poll - How can we get drea to become a more loyal reader. Maybe a free haircut by four at coach and four. Oh, I know, how about a secret password for drea to be able to participate in polls that are monopolized by loyal readers.

I am equally excited to learn about your ideas for generating more traffic. Maybe a free triple blimpie burger with egg for every new reader you direct to Mgoblog. Perhaps a Hopson cam, or how about a way for you to feature posts of readers not bothering to read the mgoboard before asking TomVH for any news about a recruit when he just provided news about that recruit like 15 minutes ago.

Dying to hear your suggestions.

Ernis

January 14th, 2009 at 7:41 PM ^

It is entirely possible that he may have been making an observation about the generalizability of the survey results. If true, it is entirely possible that Lichty overreacted, indeed. These are all possibilities. Then again, we must consider the possibility that Undefeated Drea was, in fact, offering a slew of suggestions for Brian in how to run the site. How, oh how, shall we ever deduce which to be true?

Undefeated dre…

January 15th, 2009 at 12:25 AM ^

Lichty, you're dripping with sarcasm. And drool.

I was only pointing out that having a survey that's open for less than an hour isn't the best way to get a random sample of users. I don't know what Brian wants for the site; hell, it's his site to do what he pleases with.

The only risk of basing too many decisions off the survey he ran is that he may make the site more appealing to those who are already diehard users of the site, while possibly not helping to generate new traffic to the site or to increase readership among those who check MGoBlog maybe once a day or so.

Md23Rewls

January 14th, 2009 at 7:05 PM ^

In my experience, Flip Cup has replaced Beer Pong as the best drinking party game. Several reasons.
1. My parents know what beer pong is and that's uncool.
2. Flip Cup can have almost unlimited players at once, not just four.
3. More active. You have to chug a beer in Flip Cup, not drink slowly before the next shot is hit. Also, you know when both teams in Beer Pong can't hit the last shot and everyone watches pissed off? Doesn't happen in Flip Cup.

spartansweblog

January 14th, 2009 at 7:36 PM ^

"Basketball guys, on the other hand, took one look at Kenpom and said "hey, that makes sense." Wonder why that is."

I think part of it is simply that tempo-free basketball stats are more elegant/intuitive than advanced stats in other sports. It takes quite a bit of effort to figure out what WARP and DVOA are. Points per possession is pretty much self-explanatory (as are all of the "four factors" stats).

Other Andrew

January 14th, 2009 at 9:55 PM ^

I think your sample is plenty big (opening it up to more people wouldn't change the results much at all), but as others have said, it probably skews more towards the people who make comments and generate diaries. Not so much because the survey filled up quickly, but because I would imagine they're at least twice as likely to bother to click over and completed it in the first place. Of course, coupled with other analytics, you can probably figure out the proper factor to shift survey results if you're interested.