Its astonishing that you are still fukking clueless... and yet arrogantly declaring yourself as winning thw argument. Your math skills are atrocious. You simply dont undersrand that the size of the sample of 3 and 2 stars (being 50,000 times larger) matters. In fact, not only is it statistically impossible, it is statistically a given that every NFL team is made up of moatly 2 and 3 stars. You are literally the only person on the board who doesnt get it.
There is really no hope at this point bc weve unsuccessfully tried to explain it to you in every way possible.
It reminds me of you insisting that Butters was doing a great job even in Sept and coming up with your bizarre stats that proved we were really doing well on offense and defense in the SEC.
You should apolgize to Daffy bc you are far more daffy than Daffy.
Noticing the typos, misspelled words, grammar, etc., as well as the inaccurate and intemperate nature of your post, I infer that you were probably drunk. Nonetheless, you have left the post standing so I will treat it as if you were sober. Insisting that I don't know the extent of2/3 stars is ignorant. And, "yes", I did post as if I had won the argument because I knew I had before I did the numbers. The conclusion that the rosters were too different to admit of random coincidence was obvious to anyone who dealt with statistics. Nevertheless, here is the statistical analysis.
The starting lineups break down as follows: 12 high school elites (four and five stars) on the Eagles with ten who were not H. S. elites. For the Patriots, there were five high school elites and seventeen who were not. In all, then, there were 17 elites and 27 non-elites. The question is whether this distribution resulted from random chance, or was it an artifact of coaching style and evaluation. One poster argued that talent issues had washed out over the college career and were no longer valid. If that is true, we may expect a random result. Swonkey argued that I was too stupid to understand the fact that there were more 3 and fewer stars than there were 4/5 stars (elites). This despite the fact that, on the data provided, 39% of the starters were elite and 61% (i.e., "most") were not elite. Those percentages are relevant and are fully accounted far in the following analysis. (How stupid would I have to be not to have noticed that?) If this distribution turns out not to be random, then it does hold implications for the wars over the value of recruiting service evaluations. Perhaps, it does not change the major features: almost everyone agrees that stars matter in the large (for a roster) but do not determine the success of an individual. Even casual fans must have noticed that championships are almost always won by teams with excellent recruiting classes. It may help us to understand questions at the margin: Is it possible to be a successful coach without stressing the talent aspect? (Some of you may think that Scott Frost answered that question last year with a perfect season, but despite notable wins, he did not play a Power 5 schedule.)
Since there are two possible outcomes of the variable, elite or non-elite, we use the binomial distribution. We use the distribution with a probability of 0.3864 for getting an elite and a probability of 0.6136 for not getting an elite, probabilities obtained from the data. We use the distribution with replacement of the sample. This will not significantly affect the results since the rosters are drawn from the very large population of potential NFL players and not just the two rosters which have provided the estimated probabilities. The probability that Belichick winds up with only five elite players is 9%. But this is not the correct answer. We would be yet more amazed if he had four or fewer elites. Thus, a fair comparison includes adding the probability for four (3.5%) and the probability for only three elites (1%), the probability for two elites (0.2%), as well as the negligible probabilities for one elite and for zero elites. So the total probability is 14%. There is less than one chance in seven that the Patriots roster is accidental and not the result of deliberate coaching and evaluation decisions. Lest you not be impressed, Swonkey, you call your number and cast a die. I will give you $5 when you are correct and you give me $5 when you aren't. At the end of the day I would own Swonkey and he would have enjoyed better odds (1 in 6, instead of 1 in 7). There is one final technical detail. Since we were provided only two rosters, had we never seen as much as 30 seconds of a Belichick press conference, we might wonder whether it is the Eagles coach who is the deviant. Having watched Belichick make fools of the press, everyone with half a brain knows who the deviant is. But just in case you lack more than half a brain, consider the additional evidence provided in another thread and dismissed by the innumerate: the Patriots rank first in the NFL in number of college graduates (1 chance in 32) and first in the NFL in number of ex-team captains (1 chance in 32). these are clear differences in approach to talent and evaluation. The probability that a team would accidentally have the most grads, the most captains, and only 5 elites in the starting lineup is (1/32)x(1/32)x(1/7) = 1/(7,168). However, Swonkey et al have less than 1 chance in 7,168 because the data from the two teams have suppressed a degree of freedom.
None of this means we don't want elite players. I am a bit of a recruitnik and hope we roll in the 4 and 5 stars. It does mean that when a coach passes on an elite talent, as it appears Mullen has on Xavier, that it is dangerous to condemn him before you see how his coaching and evaluation of players works out. And I still believe some of you Recruitniks would prefer the number one recruiting class to the National Championship.
