A look at stat priorities for Defensive Linemen.

[7hawk77]

[div align=\\\"center\\\"]A look at stat priorities for Defensive Linemen[/div]

Hey everybody,

Today I'm going to be going over the defensive linemen stats and what to prioritize. I'll also be doing some minor guesswork and analytics on the data once I present it.

To do this, I'll be using data for the 14 game season. I'm only using players that played all 14 games. I think there is only one player that didn't play all games so I wanted it to be more reflective of players starting out and updating along the entire season to smooth out any statistical outliers.

Below is the data for the defensive linemen tackles/sacks etc.








Now lets also look at player stats.





Now that we have both of those, lets look at the correlation between them.

First lets look at tackling. Which stats correlated best to making tackles?

First let me talk about the unimportant stats. Correlation does not mean causation. Pass blocking run blocking, arm strength all have positive correlations? Why is that? My guess is that the players that have higher stats of these have purchased all of the equipment which means they probably have bumped up their TPE more than other players as well as invested more time advancing their character. Because of this they are outperforming their peers and yielding an irrelevant stat correlation.
Anyways, After disregarding irrelevant stats, we get the following stat priority to make tackles.

Strength > Speed > Tackling > Endurance > Intelligence > Hands > Agility

This seems about right to me but strength being so high is kind of surprising. I imagined that it would be something like Speed>Tackling>Strength>Endurance.
My biggest guess for this is pancakes. Offensive linemen pick up pancakes but defensive linemen don't get a negative stat saying that they were pancaked. My guess is that without sufficient strength, the player will get pancaked and be unable to really contribute at all to the play. Either that or you just need strength to bring the player down.

Next lets look at tackle for loss correlations.



So this stat priority is:

Speed > Agility > Tackling > Hands > Strength > Endurance > Intelligence

Again this is interesting. Strength is quite a bit lower and hands looks like a big outlier. I honestly don't know how hands beat out strength. Speed, agility and tackling make perfect sense on tackling for loss, but I don't have a great explanation on why hands would be more useful than strength. Maybe this is just an outlier. There is a drop off between tackling and hands and then strength and endurance. The trend looks like the top 3 stats on a priority list are very important and then things start dropping off.


Next lets look at Forced Fumbles.


I'm not that interested in fumble recoveries because I feel like it's more random than forcing fumbles. Also forcing a fumble seems to make the fumble recovery more likely.

Stat priority:
Strength > Tackling > Speed> Endurance > Intelligence > Agility > Hands

This makes sense. You get to them fast and hard. Strength tackling and speed are all very important which again shows that top 3 coefficient.


Finally, lets look at everyones favorite. Sacks.




Stat priorty

Strength > Speed > Tackling > Endurance > Intelligence >Hands > Agility

Now this is pretty surprising. Agility is very far down as far as sacks are concerned. Also Strength and Speed are so much more valuable than any other stat. I'm starting to think that tackling counters agility, and really there is only 1 mobile quarterback and who knows what will happen to him with the bridges he burned. Because of this, I think that tackling isn't as important for sacks until quarterbacks start investing in agility.


I already wrote a piece about sacks and more specifically team sacks so if you are looking for more information on that, go read about it here: http://nsfl.jcink.net/index.php?showtopic=1363

Anyways, If you want to model your defensive linemen to be really good at a certain aspect of the game, feel free to use these stat priorites at your own risk.

One final note that I'd like to add. I currently have the excel sheet set up where I can easily do the same analysis with update stats. In the future I'm going to see how the correlation coefficients shift and speculate on why that is happening.

If you have any questions about the above, feel free to ask.

Code:

727 words