<body><script type="text/javascript"> function setAttributeOnload(object, attribute, val) { if(window.addEventListener) { window.addEventListener('load', function(){ object[attribute] = val; }, false); } else { window.attachEvent('onload', function(){ object[attribute] = val; }); } } </script> <div id="navbar-iframe-container"></div> <script type="text/javascript" src="https://apis.google.com/js/plusone.js"></script> <script type="text/javascript"> gapi.load("gapi.iframes:gapi.iframes.style.bubble", function() { if (gapi.iframes && gapi.iframes.getContext) { gapi.iframes.getContext().openChild({ url: 'https://www.blogger.com/navbar.g?targetBlogID\758865224\46blogName\75Hawkeye+Hoops\46publishMode\75PUBLISH_MODE_BLOGSPOT\46navbarType\75BLACK\46layoutType\75CLASSIC\46searchRoot\75http://hawkeyehoops.blogspot.com/search\46blogLocale\75en_US\46v\0752\46homepageUrl\75http://hawkeyehoops.blogspot.com/\46vt\0751162862754422728162', where: document.getElementById("navbar-iframe-container"), id: "navbar-iframe" }); } }); </script>
Friday, April 08, 2005
 
Teaching An Old Dog Some New Tricks
After reading an insightful article (by Studes of Baseball Graphs) about representing data with pictures over at the Baseball Analysts, I decided I better update my routine. I spend a lot of time sifting through numbers, creating tables, etc, so I'm pretty comfortable interpreting data contained in a table, and that's become the standard medium for presenting my information. I hadn't really thought about it, but that's probably not always the best way to go about it.

As an alternative, I'll try and spruce things up with a spiffy graph from time to time. The following picture is probably easier to grasp than just listing PPP numbers for each team in a table. (Note the eerie similarity of this graph to the ones Studes used in the article.)



Interpreting the Graph
Offensive efficiency runs the horizontal axis, so the further a team's dot is to the right, the better their offense was this year. Likewise, the higher the dot, the better the defense. The diagonal lines represent expected winning percentages for a team with a given combination of offensive and defensive efficiencies. For example, a team that scores as many points as it allows (offensive rating = defensive rating) would be expected to finish with a .500 record.

Note - I used the following formula for expected win % -

Expected Win % = Offensive Rating^10 / (Offensive Rating^10 + Defensive Rating^10)

If you're not familiar with the formula, Ken Pomeroy penned a good piece explaining how it works.

Observations
Illinois and Michigan State were really head and shoulders above the rest of the league this year. Not only were they the two best offenses, but their defenses were both among the top three.

Despite their 7-9 league record, Iowa played like a slightly-better-than .500 team. They fell short of their expected win % because they lost so many close games. On a side note, Iowa's expected win % during their 12 conference games without Pierce (9 regular season + 3 BTT) was a solid .560, which works out to a 9-7 record over 16 games. Since they bring everyone back next year, there's plenty of reason for optimism for next year's conference season. I'll have plenty more on this soon.

Minnesota's success was largely attributable to their exceptional defense, as their offense was one of the worst in the Big Ten. Minnesota was the opposite of Iowa in that they exceeded their expected win %, as they won several close games. You can usually expect those kind of breaks to even out in the long run, which bodes well for Iowa next year.

Wisconsin was much closer to middle of the pack than they were to MSU and Illinois. Given the number of players returning to Iowa, Indiana, and Ohio State, and the number of seniors leaving Wisconsin, it's conceivable that the Badgers could finish as low as fifth or sixth next year (though I doubt Bo Ryan will let that happen). Right now you're probably saying "Wisconsin not in MSU's league? Doesn't this guy remember Wisconsin making the Elite Eight?" While their tournament run was admirable, don't forget that the only single-digit seed the Badgers faced was North Carolina.

Remember how good Illinois was this year? Historic....season for the ages...etc, etc, right? That's approximately how bad Penn State was this year. With 931 points scored and 1181 points allowed, their expected win % was a laughable 0.085. Illinois's expected win % was 0.906, so you could make a good case that Penn State was worse than Illinois was good this year. Good luck without Aaron Johnson next year.

Purdue was another team that fell short of their expected win %. They actually had a better points margin than Northwestern, despite finishing three games behind the 'Cats. I think you can attribute most of that gap to NW's success in close games, most notably the comebacks against Iowa and Minnesota.

Feedback Requested
So, what's your take? Do you like the graph, or do you prefer that kind of information neatly tucked into the rows and columns of a table? Comments and emails are always appreciated.
Comments:
The graph is sweet and was a great surprise.
 
I ilke the graph... keep up the good work.
 
I like the graph and I LOVE statistics. I used to be a high school math teacher and used this kind of stuff a lot when I coached. But it all boils down to Bob Knight's (in)famous quote: "There's only one stat that matters: W. All the rest of the stats are for losers."
 
Thanks for the comments.

billh - Love the quote. Bobby's team was a perfect counterexample to what I'm trying to show with these graphs and stats like Efficiency Margin, as his team went 10-6 in the Big XII and made the Sweet Sixteen, despite being outscored in their conference games. In many situations, wins are all that really matter.
 
Post a Comment

<< Home

Powered by Blogger