Show HN: StatMilk - Football Stats Visualized | Hacker News Reader

Show HN: StatMilk - Football Stats Visualized

statmilk.com

85 pointsdigitalmarks47 comments

Very impressive. Definitely the best data presentation dashboard I've seen. The depth of information (even if the source of information is so centralized) plus the diversity and clarity in presentation, not to mention the overall cleanness of the whole thing, really is very cool.

Hah, read the title as "football, the sort you play with a round ball" and got rather excited. Only minor disappontment ensued when I realized you were talking about "handegg". ;)

Still, I love this kinda stuff and have been pining for more creative, dynamic presentations of futbol stats. I wish anyone could explain to me why every darn mainstream coverage site has the same, boring chart of the same, boring dozen stats. Sagarin does some fun math, but his presentation is uninspired at best.

For all the uses of highly-dynamic HTML5/JS that are distracting and unnecesary, presentation of stats is just a homerun! </zing>

Football, a sport played on foot.

What about handball?

Beautiful! Incredibly clean and intuitive. A great model for those of us building statistical interfaces. A+ and bookmarked.

Small thing: The sorting parameters for "Margin of Victory" and "Turnover Margin" are a bit off. I believe sorting alphabetically rather than numerically?

Looking forward to other leagues/sports?

Absolutely! iPhone app is in the final stages as well.

The interface is excellent. If you had NBA statistics, I would be willing to pay for this.

That looks nice. You could improve things a lot by using even a simple model to account for opponent quality, home field advantage, etc.: a simple fixed-effects regression model with team-level effects would be one approach and would take only a few lines of R code.

For an example, look at: http://www.pseudotrue.com/details which is my blog, so this is a pretty shameless plug but I don't really know of a different reference.

The point of using model isn't because you believe the model is true or expect it to necessarily be accurate, but because it can give you a better measure than just looking at (say) yards per game.

But your site is cool as is, so good job.

edit: tweaked URL to point to the right page.

This is one of the coolest things I've seen in a long time. Would be cool to have the BCS rankings in there as well.

Awesome work guys, the presentation and design are excellent, top of the line. Some much attention to detail, pixel perfect, I'm really blown away! I'm not into sports at all, but this is inspiring at many levels.

Very nice! Beautiful and very fast for the amount of data that is calculated on the fly. What technologies did you use? Also, did you have to pay license fees to display team and conference logos?

MVC3 + SQL Server + Mongo + custom software for the realtime calculations.

All components (with the exception of the front-end) are highly abstracted enabling us to swap technologies (e.g. SQL Server to MySQL, Mongo to Memcached) quickly.

Thanks for replying. I'd still love to know more about the process of getting permission to show conference and team logos.

Yeah, I'd like to know about the process of showing team logos and the stats. Are you just scraping the data from somewhere, or do you have permission to use any of this data? I'm wanting to build and application similar to this for the NHL (if we have a season... har har...) but the licensing of big teams is scaring me away some.

Always great to see awesome things built on .Net!

I sent this to my sports and statistics obsessed friends. They will definitely love this.

This is awesome. My only concern is that the data looks to be a bit stale, as the Oregon Ducks won last night but none of their stats are up. Mind if I ask where you're pulling the data from?

Thanks. Data is sourced directly from the NCAA. Expect updates every Sunday morning.

are you scraping results into your database from their site? Also, are you familiar with sportsML and sportsdb[1] ?

I ask because I am also working on a sports product for release in November.

Nice work!

[1] - http://www.sportsdb.org/sd

This is beautiful. What charting libraries are you using?

Would love to see a breakdown of how you set this up as well as know what libraries.

We'll have a blog up and I promise to cover this.

Here's the breakdown of charting libraries OP is using

* HighCharts http://www.highcharts.com/

* jQuery Sparklines http://omnipotent.net/jquery.sparkline/

* jQuery Knob http://anthonyterrien.com/knob/

Being a fan of data visualization myself, I find this work inspiring and would like to hear his thoughts on why those libraries over using either d3.js or Raphaël.

This is so ill, I love everything about the presentation. I'm partially glad you don't have NFL stats, otherwise my work productivity would plummet.

Thanks - and I suggest you get a lot accomplished over the next few weeks.

Great data visualization but the stat widget headers are pretty hard to read due to the complete lack of contrast with the background color.

As someone who really appreciates sports stats, I find myself ignoring the numbers to admire the design. Really, really beautiful.

What are you using for the help array overlay?

The margin of victory does not correctly sort.

As well, I'd generally think a 6-0 team should be ranked above a 6-1 team, does not currently appear that way.

Agreed. It should be rank ordered by losses then wins.

Seems like it's using a lexographical sort on any column with numbers and letters (like "42 yds").

Where are you obtaining your stat data from?

It appears the only thing that correctly predicts victory is turnover margin!!

There is a high correlation in the NFL between turnover margin and winning.

really nicely done, would love to see NFL stats

What are you using for the help arrow overlay?

Overlay image + CSS + jquery-cookie

Incredible. I will definitely return to this.

Too, too sexy.

Yeah, I think it's trying to hard in this regard. It's got a nice clean look, but the font can be a little more traditional, and the dark background with glowing buttons makes it look more like a toy than something used for serious analysis.

It seems geared to comparing conferences and not arbitrary matchups. I was expecting to be able to compare any 2 teams. I'm sure you're working on it.

Good job presenting the data though.

Of course you can compare any two teams - click on the logo and select the team(s) you want.

I can only select one at a time. And it's not obvious that the NCAA logo is a button.

How can I see a Matchup between Cal and Stanford, like you have for Miami and Florida St.?

http://statmilk.com/NCAAF/MatchUp/107/674/

Ah, how silly of me. You told me to just click on the logos and I was just clicking on them on the wrong tab. However, you really need to put more time into the design of that feature, and that select "button". I know you have a splash screen explaining "select a team", but there's no affordances for the user who overlooks that instruction and is looking to achieve the same thing I was. A simple mouse over effect would help big time.

BTW, on the "Matchups" page, I was expecting to see more than one. You might want a featured list somewhere on that page to make it easier to load this week's big games.

Addressing both issues this week - thanks for the feedback.

* Looks great, I think everyone is in agreeance on that. I had some questions about the actual data being shown and whether the emphasis should be on other datapoints and aggregations...

So the aggregate conference W-L standings are overall records, right? As the season goes on, most games are played within conferences...so...what's the relevancy of displaying conference by these W-L numbers? If there were no early non-conference games, then the W-L ratio would basically be 50-50 for every conference. I think you should remove that statistic as it's not terribly useful.

* Also, the filters provide a few confusing options. If I select "Winning games", the view stays the same except that everyone (and every conference) has a winning record...I understand programmatically why that is, but I don't think the layperson will.

Great feedback - we're listening...

Reply on news.ycombinator.com