Im finished with the data part of the project.
Some things confirm what we already knew, although I did not expect such a gap between QB and other positions. You'll also notice the slope starts going up for all positions and then drops down again. While I havent tried to play with the data to explain this yet. My off the top of my head crack at it is the rookie wage scale... Which looks like this: (sorry for the ugly chart lol)
Thats a very steep slope and as you'll see in the EV chart, the EV starts to stabilize around the same place where the salaries slope stabilizes.
A few things to consider, the elite category was harder to separate from the good for a few positions. All OL positions, ILB and interiorDL are part of this group. Whats interesting is that also showed up in the salaries, OL for example didnt have a big gap between elite salaries and good salaries, that even extended to jag players. And this is not driven by my analysis, I used quantiles to define group vet salaries...
What I get from this is, it is just as hard for NFL people to evaluate those group as it is from us. The lack of quantifiable data makes it hard for everyone to separate them into tiers.
Also notice that a couple of positions are actually undraftable at the very top of the draft.
**I sorry I couldnt separate DBs into CB and safety categories, my datasource did a bad job of labeling them and while I did have some correctly labeled, the vast majority were labeled as DB. This could have been corrected, but it would have been time consuming as **** as different sites name players differently. Just decided to put them all together. So for the DBs, just assume CBs have a bit more EV and take off a couple of points for safeties.**
Here's the chart: (the interactive one is linked underneath)
https://public.tableau.com/views/NF...ille2?:display_count=y&:origin=viz_share_link
Some things confirm what we already knew, although I did not expect such a gap between QB and other positions. You'll also notice the slope starts going up for all positions and then drops down again. While I havent tried to play with the data to explain this yet. My off the top of my head crack at it is the rookie wage scale... Which looks like this: (sorry for the ugly chart lol)
Thats a very steep slope and as you'll see in the EV chart, the EV starts to stabilize around the same place where the salaries slope stabilizes.
A few things to consider, the elite category was harder to separate from the good for a few positions. All OL positions, ILB and interiorDL are part of this group. Whats interesting is that also showed up in the salaries, OL for example didnt have a big gap between elite salaries and good salaries, that even extended to jag players. And this is not driven by my analysis, I used quantiles to define group vet salaries...
What I get from this is, it is just as hard for NFL people to evaluate those group as it is from us. The lack of quantifiable data makes it hard for everyone to separate them into tiers.
Also notice that a couple of positions are actually undraftable at the very top of the draft.
**I sorry I couldnt separate DBs into CB and safety categories, my datasource did a bad job of labeling them and while I did have some correctly labeled, the vast majority were labeled as DB. This could have been corrected, but it would have been time consuming as **** as different sites name players differently. Just decided to put them all together. So for the DBs, just assume CBs have a bit more EV and take off a couple of points for safeties.**
Here's the chart: (the interactive one is linked underneath)
https://public.tableau.com/views/NF...ille2?:display_count=y&:origin=viz_share_link