Out of the box, combine results = NFL success?

[Scrap]

These are the linear model plots illustrating Sacks per game relative to combine results. (Players who appeared in more than 45 games since 2005)

View attachment 38728
View attachment 38729
View attachment 38730

These models wouldnt be useful for any kind of precise predictions. But you can still see where there's a relationship and where there clearly isnt. The shuttle for exampe is showing minimal relationship when its comes to sack production at the next level...

These show me that explosiveness and leg power is the most important factor.

 

[BigNastyFish]

Science!

Science?

Isn't that bad voodoo in today's climate?

Speaking of climate...

Interesting work.

Tho I prefer the madlebrot look!

 

[NBP81]

OLBs

Size: This is the thing with OLBs, there's a difference between a 3-4 one and a 4-3 one. These guys are labeled as OLBs one way or the other. Some place have started using the 'EDGE' classification, but muy data goes back to 2005, which wasnt even a thing then.

The distribution illustrates that with its jellybean shape. It really looks like there are 2 "averages" for this distribution. Which I assume is WOLB -> ~ 6'1" - #235 and SOLB -> ~ 6'3" - #250. (labeling these 2 groups is pure assumption on my part, but I think its a fair assumption).



40 yard dash:



Vertical:



Bench:



Broad Jump:



3 cone is interesting for the good group... There seems to be a gap where some have very good drills, and some not so much... This is probably another distinction for the strong/weakside guys...



Shuttle clearly matters:



2 most significant ones here are the 40 and the shuttle, the 3 cone is interesting when filtering for scheme... Size also seems to fit the latter category.

 

[13marino13]

Since its combine week, I thought I'd dive into how combine results actually correlate to NFL production. While I'm only getting started on this, there's really not a whole lot that clearly tips the scale one way or another at first glance. Some of it though, does seem to matter.

Before we get into this I posted a poll to gauge interest on the subject. This is quite time consuming and if nobody gives a **** I just wont keep doing those. So if you like that stuff, just click on the position groups that you'd like me to do this sort of analysis on and you'll effectively tell me 1) You're interested in that sort of analysis and 2) What positions you are interested in.

This one's about offensive tackles, I would've started with QB but the honest to god truth is that there was really nothing to talk about in there... The combine for QBs just doesnt tell you anything(Sorry, I dont have access to reliable hand size measurements data).

I didnt go with my previous way of doing things in this one... Instead of getting [Elite, Good, Jag, Bust] categories, we'll be going with over 75th percentile vs. the rest. You can effectively view these groups as good vs. bad... I'd rather have less groups with more sample size for this kind of thing, is another way to put it...

Tackles

Size matters?

Here are the charts illustrating size for both groups. The good group looks alot more focused around ~ 6'4" 315 while the other group is kind of all over the place, which is to be expected.

Good:

View attachment 38666

Not so good:

View attachment 38667

Bench press matters?

Average is about the same for both parties... Though the distribution is skewed towards more is better... Nothing under ~18

View attachment 38677

3 cones? meh... Although theres a pretty clear threshold... Absolutely nothing over ~8.25 here for the good ones...

View attachment 38678

Shuttle matters...

View attachment 38679

Example of something that doesnt matter... Vertical...

View attachment 38680

Everything remaining looks like the last one, trying to draw conclusions on these wouldnt be wise. Also keep in mind I dont have data on wingspan, arm lenght and hand size so I cant comment on those.

Hope you guys find this useful.

Thanks for taking the time, awesome posts brother!