[CTcard]

[quote author=OutsiderFan link=topic=17237.msg198351#msg198351 date=1503952441]The PER numbers are shocking.

No they're not. I was pretty sure I could explain them off the top of my head by saying that Costello didn't throw a TD, while Burns and Chryst both did, and they all had similar completion percentages. (Looking at the box score now, this is all true.) I think Burns is probably ahead of Chryst because of his insane TD%, but I don't know how the college formula works exactly, just that it's probably similar to the pro one.
[/quote]

The NCAA PER formula is:
PER = 8.4*yds/att + 330*TDs/att + 100*completions/att - 200*INTs/att

So for our QBs this works out to be:
KC(PER) = 88.6 (yds/att) + 58.3 (cmp/att) +27.5 (TDs) - 0 (INT) = 174.4
RB(PER) = 75.6 (yds/att) + 50.0 (cmp/att) + 82.5 (TDs) - 0 (INT) = 208.1
KJC(PER) = 74.7 (yds/att) + 55.6 (cmp/att) + 0 (TDs) - 0 (INT) = 130.2

So, you are quite correct on the ridiculous TD percentage leading to Burns' high PER.

Personally, I think cmp/att shouldn't be included because it is already inside of yds/att.
I also think TD percentage either shouldn't be included or barely so.

In this case, since nobody threw an INT, a completely useful proxy for efficiency is simply yds/att. On that score we would have Chryst (10.5) > Burns (9.0) =~ Costello (8.9).
All pretty great numbers of course.

[CTcard]

[quote author=OutsiderFan link=topic=17237.msg198351#msg198351 date=1503952441]The PER numbers are shocking.

No they're not. I was pretty sure I could explain them off the top of my head by saying that Costello didn't throw a TD, while Burns and Chryst both did, and they all had similar completion percentages. (Looking at the box score now, this is all true.) I think Burns is probably ahead of Chryst because of his insane TD%, but I don't know how the college formula works exactly, just that it's probably similar to the pro one.
[/quote]

The NCAA PER formula is:
PER = 8.4*yds/att + 330*TDs/att + 100*completions/att - 200*INTs/att

So for our QBs this works out to be:
KC(PER) = 88.6 (yds/att) + 58.3 (cmp/att) +27.5 (TDs) - 0 (INT) = 174.4
RB(PER) = 75.6 (yds/att) + 50.0 (cmp/att) + 82.5 (TDs) - 0 (INT) = 208.1
KJC(PER) = 74.7 (yds/att) + 55.6 (cmp/att) + 0 (TDs) - 0 (INT) = 130.2

So, you are quite correct on the ridiculous TD percentage leading to Burns' high PER.

Personally, I think cmp/att shouldn't be included because it is already inside of yds/att.
I also think TD percentage either shouldn't be included or barely so.

In this case, since nobody threw an INT, a completely useful proxy for efficiency is simply yds/att. On that score we would have Chryst (10.5) > Burns (9.0) =~ Costello (8.9).
All pretty great numbers of course.