Re: Calling Rusty and his Minions.....
posted at 3/6/2013 11:29 PM EST
In response to seattlepat70's comment:
In response to 42AND46's comment:
In response to seattlepat70's comment:
In response to Low-FB-IQ's comment:
I am not going to read all that mainly because I think Kerry is a doofus who sits behind his compter trying to predict real football by making up and modifying his statistical methodolgies until he gets the results he wants for his site and future articles.
The last time I looked up at the score board at the end of a game the team who scored more points won the game.
...and that's not to say that the defense can't or should not be improved upon.
According to Kerry, at the end of your post, if the Pats simply improve the D and the Offense still goes out and scores 13 points it's Lombardi time. Idiotic.
Here are the only stats that matter...
Offense - The SUPERBOWL years
Defense - The SUPERBOWL years
So basically two out of the three years the Pats won the superbowl the defense was worse than they were all season long in the biggest game but the offense saved them by not completely falling flat on their face.
...and if someone wants to use the argument that the Defense did not generate any turnovers in 2011 I can understand that and accept it as long as you also then accept that it means the offense is just smoke and mirrors and there success and point scoring ability is completely dependent on the defense giving them extra touches to figure it out because they are not efficient or consistent on their own merit.
...and if you accept that later point then you should be wishing the Patriots spend more money in FA and in draft collateral helping the offense over the defense. You can't argue the point both ways. IMHO.
I don't know if this Kerry is an idiot, but his work presented above is deinfitely idiotic.
His analysis fails on a very basic level -- i.e., using the wrong evidence to support his point. He used so much data to support his analysis yet all that data/ and fancy stats are from regular season (I know because he showed league rankings on every one of them) and even other playoff games; therefore are irrelevant to how the SB games themselves actually trasnpired.
At most, he could use those data and fancy stats as benchmarks for evaluating performance in the in the actual SB games - kinda what you did except that you did it only on the simplest, yet ultimate indicator of productive output - the score.
For about 6 months now, I have not commented on this perpetual debate of who's to blame for the two SB losses. I had a hard time resisting this time, to make the point that quant analysis per se is not to blame here. To blame here is how this person is applying quant analysis.
Frankly, in my work, I have seen a lot of people similar to Kerry - people who think they are good analysts, just because they know how to caclulate a desriptive stat like average or standard deviation, or perhaps calculate a correlation. There is more to analysis than that.
this may be the most nothing ever said using the most amount of intelligent sounding words and phrases
(stands and applauds)
perhaps because it went over your head. the article actually fits what you describe.
go ahead and elaborate on your point.
Sorry Seattle, but I agree.
There's nothing wrong with this analysis, in fact it's very concise.
You state it's wrong or he uses the wrong facts???? Where?
He is using it to predict SB winners based on RS stats. What else would you use?
Score? Average score is not a good indicator. It means they blew out some teams and struggled against better ones. A team that has a 34ppg can just as likely score 19 as they can 49. The 34ppg is the average of the two or 16 as it stands.
The metric used to evaluate O's and D's have been around since the dawn of time. He does use a more accurate measure of passer rating and defensive passer rating to include more relevant factors but that's about it.
Again, they are the same metrics used by the NFl with a little tweaking.
It is all qualified by the winning percentages shown in each category.
This is a result of using the same measure of success or failure for each team for a period of over 60 years.
It makes perfect sense to try and evaluate a victor if you evaluate both sides of the team, as both are contingent on the other.
A more balanced team is usually going to trump a team with only one strength, at least it is shown to be true 80% of the time. Of course there are other variables that can't be predicted like human error, game planning, a lucky play or an injury. There's no way to predict that but those factors are not as prevalent over the course of of a season, compared as to how well the team performed as a whole.
The difference (what he is proving) in defensive and offensive play as a whole is major and that is all he is saying. A bad Defense can negate a good offense and you really have to look no further than any game any Sunday to see that. If you have a D that can't hold it's own, it tends to make Offenses one dimensional and predictable and easier to defend. When you are behind, you pass. If you no lead is safe, you pass. Even after a 6 minute time eating drive, if you don't score, you still gotta give the ball back to the D and the other team scores and eats away at your lead. SB46.
The Pats Offense does this to teams frequently and it's really the only reason the D has success at all. Sorry but it's true.
It's been proven over and over and over that when the O has problems, and that really is to be expected occasionally, that the D goes limp.... the picks disappear, the fumble recoveries disappear, the rz stops disappear, they can't get of the field, and they lose.
There is a reason no team has ever won a SB with a DPR of over 82.
That's because it puts your O in a position to have to better that against a much better D.
Impossible, unless you are very, very, lucky. So far, no team has been that lucky.