Professional Football Researchers Association Forum
PFRA is a nonprofit organization dedicated to the history of professional football. Formed in 1979, PFRA members include many of the game's foremost historians and writers.
For players who return kicks and punts, there is no good way to compare...I mean there is, it can be done as a percent of the total. Here I took 20 yards away from the KR average and then also divided by 2....then did a harmonic mean for both, and then did dame for KOR TDs and PR TDs...
Next step is what is a reasonable minimum, Williams and Sayers land on top... but they had to numbers on one of their stats. Also, need to give some toal for zero TDs...so there can at least be a mean figured...
here is a rough chart but lots of names below these....
JohnTurney wrote:I took 20 yards away from the KR average and then also divided by 2....then did a harmonic mean for both, and then did dame for KOR TDs and PR TDs...
John, Can you tell me in words what this formula is doing? I looked up what the harmonic mean was, but I'm still a little lost.
Without understanding and just looking at the results, did the player have to return both kicks and punts? In other words, if a player did only one of the two and was really good at it, would they be able to get a high ranking? Finally, I was expecting to see Hugh McElhenny, but I guess as stats weren't as great as I thought.
JohnTurney wrote:I took 20 yards away from the KR average and then also divided by 2....then did a harmonic mean for both, and then did dame for KOR TDs and PR TDs...
John, Can you tell me in words what this formula is doing? I looked up what the harmonic mean was, but I'm still a little lost.
Without understanding and just looking at the results, did the player have to return both kicks and punts? In other words, if a player did only one of the two and was really good at it, would they be able to get a high ranking? Finally, I was expecting to see Hugh McElhenny, but I guess as stats weren't as great as I thought.
Yes, it's for kicks who did both only. Harmonic mean is just a certain kind of average and it averages kick return average and punt return average. But kick return average is modified, I did it two ways but have not decided which I like better. One takes a KR avg and subtracts 20 (amoung of a touchback, i.e.e you get 20"free yards". I may end up using 15. the other is KR ave divided by 2. It attempts to make KR and PR averages closer. Kickoffs havethe advantage of if it is in end zone you an kneel down and get 20 yards. It's a work in progress I wanted to share
Thanks John. I see what you are trying to do now. Minus 20 sounds right for a kickoff into the end zone and is probably a little high for a shorter kick off. I'll think about it and post back if anything comes to mind.
I'm not criticizing the methodology, by any means, but a method that's often seemed helpful to me is to express a player or team stat as a ratio of the league average.
For example, in 1965 Gale Sayers averaged 14.9 yards on punt returns and 31.4 yards on kickoff returns. The NFL averages were 7.8 and 23.2, respectively.
Sayers would get a score of 1.63 (or 163, if you prefer big numbers) based on the following calculations:
rhickok1109 wrote:I'm not criticizing the methodology, by any means, but a method that's often seemed helpful to me is to express a player or team stat as a ratio of the league average.
For example, in 1965 Gale Sayers averaged 14.9 yards on punt returns and 31.4 yards on kickoff returns. The NFL averages were 7.8 and 23.2, respectively.
Sayers would get a score of 1.63 (or 163, if you prefer big numbers) based on the following calculations:
14.9/7.8 = 1.91
31.4/23.2 = 1.35
1.91 + 1.35 = 3.26
3.26/2 = 1.63
That's all fine, have used that in other ways, but what I did here was done in a few minutes to prepare for the MMQB draft. What you are talking about would have taken a lot of time. Also, that's been used for QBs. I have toyed with something as a percentage of the leader
If NFL leader had a 30 avg and player A is the leader and Player B had a 20 yard avg
Player 30 would be 100%
Player B would be 67%
