Blake Baker- By the Numbers

Blake Baker- By the Numbers

Lance Roffers
Lance Roffers

Comments (48)

@Lance Roffers , can you elaborate a bit more on how you derived the std. dev? Was "per-play" the unit of measure throughout? Or did you look at something else when deriving std.dev of "performances"?

Also, what is your source data?

Thanks in advance.

"I calculated how each offense fared against the La Tech defense on a per-play basis and then calculated how each opponent fared in their games against other Group-of-5 opponents and then calculated the standard deviation of their performances over the course of the season."
 
Thanks for the excellent breakdown. Putting what seem like intangibles into numbers is an art.
 
Empirical evidence is very comforting... Lance you are a beast!
 
@Lance Roffers , can you elaborate a bit more on how you derived the std. dev? Was "per-play" the unit of measure throughout? Or did you look at something else when deriving std.dev of "performances"?

Also, what is your source data?

Thanks in advance.

"I calculated how each offense fared against the La Tech defense on a per-play basis and then calculated how each opponent fared in their games against other Group-of-5 opponents and then calculated the standard deviation of their performances over the course of the season."

Sure. The standard deviation of their yards per play data from each of the games they played against other group-of-5 teams. The point of that was two-fold: 1. It tells me the variation in performance for a team. Such as, are their numbers perhaps skewed by big performances in a couple of games, or maybe they had a game without their QB that brought their overall numbers down, but their norm was something much better. 2. I was then able to decipher how many standard deviations away from the mean the defense of La Tech was able to hold them above/below.

Source data is game logs for each game/opponent of La Tech. I tend to use CollegeFootballReference, but there are many options out there if you're wanting to check my math.
 
Advertisement
Sure. The standard deviation of their yards per play data from each of the games they played against other group-of-5 teams. The point of that was two-fold: 1. It tells me the variation in performance for a team. Such as, are their numbers perhaps skewed by big performances in a couple of games, or maybe they had a game without their QB that brought their overall numbers down, but their norm was something much better. 2. I was then able to decipher how many standard deviations away from the mean the defense of La Tech was able to hold them above/below.

Source data is game logs for each game/opponent of La Tech. I tend to use CollegeFootballReference, but there are many options out there if you're wanting to check my math.

Cool. Thanks Lance. Just wanted to clarify my understanding on the std.dev.

And I'm tinkering with Python and looking for some useful data I can play around with that is actually interesting to me.
 
Cool. Thanks Lance. Just wanted to clarify my understanding on the std.dev.

And I'm tinkering with Python and looking for some useful data I can play around with that is actually interesting to me.

Python is great for data scraping. I recommend beautifulsoup if you need a storage library.
 
FYI nice 4 minute youtube video with the press asking him questions after LT spring scrimmage. I really like what I see.
 
Advertisement
Back
Top