jbusse
Posts: 1309
Joined: 9/11/2013
From: Atlanta, GA
Status: offline
|
quote:
ORIGINAL: David Levine quote:
ORIGINAL: jbusse quote:
ORIGINAL: David Levine quote:
ORIGINAL: jbusse quote:
ORIGINAL: David Levine quote:
ORIGINAL: jbusse quote:
ORIGINAL: TJSweens I think the angst and teeth gnashing over the Vikings not hiring Harbaugh is getting as over the top as Brad's objections to him. You guys are really engaging in an incredible amount of speculation. How much he was asking for...did he even bring up money...how well would he have coached a completely different roster with a different organization...he would have assembled a staff of assistants who should all be head coaches themselves... Come on people, you don't know any of this. Kwesi's supposed to be all about analytics. What kind of analytics model favors O'Connell over Harbaugh? Past success is an excellent predictor of future success for NFL coaches. New head coaches are a mixed bag. I think this is more confusion about analytics. Analytics is not looking at boxscores, overall record, number of TDs thrown, etc. That's shit everyone knows and all you need is a newspaper or profootballreference.com to get your "data". Analytics is trying to explain things that aren't so quantifiable. It's not "what happened?" it's "why did it happen?" and "how can I make this happen more regularly?". Make no mistake, I'm not confused about analytics. Analytics doesn't ignore important quantifiable relationships, such as the relationship between past and future head coaching success, even if it is "shit everyone knows". And I'm pretty sure KAM is fully aware of Harbaugh's past record. As such, pretty easy to see that the Vikings did not use analytics to choose O'Connell over Harbaugh based on expected future wins and losses. Going with culture, personality, etc., i.e., hard to measure criteria that would be difficult to correlate with NFL wins, is pretty much anti-analytics. Win/loss record is probably the most "anti-analytics" thing there is. It's not anti-analytics, much less "the most anti-analytics".
|