Yakuza Rich
Well-Known Member
- Messages
- 18,043
- Reaction score
- 12,385
I gave you the benefit of the doubt that you were a statistician but this paragraph proves you're just a wannabe that has no clue what he is talking about.
What you just described isn't a statistical analysis, it is a cause and effect. If a pro golfer hits a ball off a tee at 100 MPH and I hit it off the tee at 50 MPH, with all other factors being equal, then yeah his ball will go farther than mine (duh!). That's cause and effect, not a statistical correlation.
The fact that you don't know the difference shows how clueless you are about this subject.
This is the last I will say on this subject now that you have proved your obscene ignorance. I know your childlike mind demands you have the last word so go ahead and have it.
Oh BTW, the #1 rule of statistical analysis: Correlation does not imply causation.
The issue is that you don't understand what that phrase means (so few non-statisticians really do).
In the golf example, I can absolutely prove a correlation. I can simply record ball speeds and distance traveled. I can then run a mathematical correlation and use a simple linear regression to project how far a golfer will hit the ball if they reach a certain ball speed. It's flat-out silly for you to claim that I can't prove a mathematical correlation.
In fact, I did a similar study of handicaps and club head speed this past year. The correlation coefficient was 0.9 which indicates a strong mathematical correlation. In fact, here's the linear regression formula that projects USGA handicap based on club head speed:
(Club Head Speed - 106.486783804431) / -1.38899923605806 = Projected USGA Handicap
Correlation is a symmetrical relationship.
So in the case of club head speed and handicap, you may run into a golfer that has significantly lower club head speed, but has a much lower handicap and vice versa. But when testing over 300 golfers from the Tour player to the 25 handicap, the sample size is large enough and it draws a very strong mathematical correlation of +0.93. So with a high degree of confidence we can say that the faster the club speed of a golfer the more likely they will have a lower handicap.
The same can apply with Sharp's statistical study. Obviously, the Patriots could still fumble a deflated ball just as there are going to be teams that will not fumble much with a properly inflated ball.
But, we can see a symmetrical relationship between their fumble rate and pre-2007 and post-2006. That coincides with the rule change that was proposed by Tom Brady. And that was the point of Sharp's article and backed up by Burke.
It was the critics of Sharp's work that jumped on the idea of deflating a football cause spikes in the Pats' fumble rate. The core of the problem is that we don't even know if deflated footballs will actually cause less fumbles because we've never tested out that theory. I just wouldn't ignore the stats that coincide right with that timeline.
YR