A Semi-Objective Look At The 2012 Draft

[Sapper]

(Disclaimer, this was written on Sunday, but I posted it today to let my script run overnight to test the numbers 1000's of times for to .00001 accuracy. I rounder to a 10th of a degree however).

Okay, so with time on my hands on a lazy Sunday I decided to do a little statistics. Normally I would be on Gin and Tonic number 3 on the balcony enjoying a post-draft Sunday Funday, but today I remain sober until the sun sets (and by sober I mean beer #3).

Anyway, I am using the grades here: http://www.nfl.com/d...racker/by-round

I don't know if one person, 2 people or 30 people came up with them. But I take the site as the most neutral that actually compiled ratings. A better way to do it would be to take the ratings from multiple sites and average them out, but I'm not that nerdy ;). Before you say, you have too much time on your hands, I wrote a script in 10 minutes that traverses through each page based on team and player average, arrays them and averages them. This list in itself was written as follows:

Average grade of each player draft divided by number of team's picks. Or % = A/N. Also of note, since the lowest gauge they use is 20-49, I am counting all players without a ranking as 19.

This isn't a great gauge, I will admit (the taints had very few picks). But it was an easy algorithm to write. The script needs handicapping (like value per round, number of picks adjusted, player projections per round vs. where they actually went, and the inclusion of undrafted players). Teams with higher/more picks are at a pretty good advantage. I am not going to go that far, because that would take more thinking than I am willing to do today. But if you know the programming language Python, feel free to message me and I will send you the algorithm and you can adjust it any way you like for something more accurate.

Carolina ends up getting hurt by drafting a punter and Alexander who was given his grade of 60 when they thought he had a heart condition. The number in parenthesis after Carolina's grade is excluding the punter we took and putting Alexander at a 75, where he could be around now that we know he was misdiagnosed.

Anyway, here are the results, top to bottom.

1. Miami: 76.9

2. Cincinatti: 76.8

3. Arizona: 72.5

4. Green Bay: 72.3

5. Cleveland: 71.1 (tie)

6. Houston: 71.1 (tie)

7. San Diego: 70.2

8. Carolina: 70.1 (76.0)

9. Pittsburgh: 69.8

10. New York Giants: 69.3

11. Tampa Bay: 69.0

12. Minnesota: 68.8

13. Kansas City: 68.2

14. Detroit: 68.0

15. Tennessee: 67.9

16. Philadelphia: 67.7

17. Baltimore: 67.4

18. San Francisco: 67.0 (tie)

19. St. Lous: 67.0 (tie)

20. Denver: 66.4

21. New York Jets: 66.3

22. Buffalo: 66.2

23. Dallas: 65.0

24. Indianapolis: 64.5

25. Chicago: 62.7

26. Jacksonville: 62.1

27. Atlanta: 61.5

28. New England: 58.2

29. Seattle: 57.6

30. New Orleans: 56.6

31. Washington: 56.3

32. Oakland: 53.6

Closing notes, Indy and Washington seem low for getting the 2 best players in the draft. But this encompasses the whole draft. Shocker who came in last (rolls eyes)

(edited for spelling errors)

Also, let's say we keep Alexander's 60, and remove the punter. We would have a 73.6. Re-upped Alexander's grade to 75 and left the punter? 72.2

Give me thoughts/adjustments/critiques/what have you. I would love to write a amazingly adjusted algorithm for this draft when I have the extra time. Thanks!

[megadeth078]

[Gallagher]

I think a more accurate statistical measurement would be the player's draft grade weighted by where they were taken, so the end stat would not just show how good the players we drafted are but rather, how good they are given where we drafted them (ie their relative vs. absolute value).

For example, (draft grade) x ((Total Number drafted - Position drafted) / total number drafted). So Kuechly would be 91*((253-9)/253) + [Amini] 83.7*((253-40)/253) + ...... + [DJ Campbell] 53*((253-216)/253)

This would put a heavier weighting on the earlier rounds, meaning teams would be judged more harshly on what they do in the early rounds and thus would be biased against teams like the Saints* or Ravens that traded out of the first. You could also reverse the weighting to place more value on later draft picks but, that would rank teams by who found late round value, a different measurement entirely.

[Sapper]

I think a more accurate statistical measurement would be the player's draft grade weighted by where they were taken, so the end stat would not just show how good the players we drafted are but rather, how good they are given where we drafted them (ie their relative vs. absolute value).

For example, (draft grade) x ((Total Number drafted - Position drafted) / total number drafted). So Kuechly would be 91*((253-9)/253) + [Amini] 83.7*((253-40)/253) + ...... + [DJ Campbell] 53*((253-216)/253)

This would put a heavier weighting on the earlier rounds, meaning teams would be judged more harshly on what they do in the early rounds and thus would be biased against teams like the Saints* or Ravens that traded out of the first. You could also reverse the weighting to place more value on later draft picks but, that would rank teams by who found late round value, a different measurement entirely.

A great suggestion, if I get the time later I will alter it and let it run overnight. More suggestions welcome!

[Pimpdaddy]

....you should have gotten drunk....

[Sapper]

....you should have gotten drunk....

Well, I still did ;)

[mav1234]

interesting, but I don't think it is worth spending a lot of time figuring out who had the best draft before any players have played a single down. neat idea still, I like things like this even if they are not worth much.

I would probably do something like taking the average score of the 5 picks below that draft pick and subtract it from that particular pick in order to generate a score per pick, giving you a general idea of if the experts thought that pick was a good one given its rough placement.

for instance we'd get 2.56 for Kuechly, -1.81 for Silatolu, -10.92 for Alexander (given improper grade, if we bump him to 72 he gets a 1.08), 7.52 for Adams, etc... compared to the Dolphins that had a -2.24 in the 1st, 3.62 in the 2nd, etc... I dunno, experts, meh.

[bababoey]

You guys make my head hurt. And I do uncertainly calucations for a living.

[Sapper]

interesting, but I don't think it is worth spending a lot of time figuring out who had the best draft before any players have played a single down. neat idea still, I like things like this even if they are not worth much.

I would probably do something like taking the average score of the 5 picks below that draft pick and subtract it from that particular pick in order to generate a score per pick, giving you a general idea of if the experts thought that pick was a good one given its rough placement.

for instance we'd get 2.56 for Kuechly, -1.81 for Silatolu, -10.92 for Alexander (given improper grade, if we bump him to 72 he gets a 1.08), 7.52 for Adams, etc... compared to the Dolphins that had a -2.24 in the 1st, 3.62 in the 2nd, etc... I dunno, experts, meh.

Definitely will look into that too.

[Sapper]

interesting, but I don't think it is worth spending a lot of time figuring out who had the best draft before any players have played a single down. neat idea still, I like things like this even if they are not worth much.

I would probably do something like taking the average score of the 5 picks below that draft pick and subtract it from that particular pick in order to generate a score per pick, giving you a general idea of if the experts thought that pick was a good one given its rough placement.

for instance we'd get 2.56 for Kuechly, -1.81 for Silatolu, -10.92 for Alexander (given improper grade, if we bump him to 72 he gets a 1.08), 7.52 for Adams, etc... compared to the Dolphins that had a -2.24 in the 1st, 3.62 in the 2nd, etc... I dunno, experts, meh.

Any chance you know of a site where all players have projections vs where they actually went?

[Argus Plexus]

[Sapper]

Hah...just ranking the draft by a script that I thought up in a few minutes strictly based on ratings/# of picks. Obviously the best way to rank this draft is to come back in 4 years and see who benefitted the most. A more accurate analysis soon to come based on suggestions.

[SOJA]

I am very poor with statistics. However, if you want a fun biology fact: there are more bacteria cells in your colon that human cells in your entire body!

The more you know!

[Captain Obvious]

It's like nerdvana. This is cool :)