[msqueri]

Ok, you guys have gotten this bee in my bonnet and it can't get out without empirical observations. With full acknowledgement of the sbortcuts/flaws/blind spots in this approach, here is a list of every Pac-12 team since 2014 ranked by number of games missed by players in their season-end top 15 for snaps on either offense or defense:

1. 2021 Washington - 61 (12 games) - 5.08
...

Some questions about your chart.

1. What exactly is the number listed for each team? 
You stated, "number of games missed by players in their season-end top 15 for snaps on either offense or defense".
Number of games missed ought to be a whole number. Here something is apparently divided out. 
Is it something like average number of players missing per game? 

2. Where do you find stats for snaps played?  And games missed? (PFF?)
Or is the stat you list itself available somewhere? 

3. It would seem that this might account for minor injuries but not major injuries. i.e. a player with a major injury would miss enough time that he wouldn't end up in the top 15 for the team in snaps played. Does that seem correct?

1. The format "2019 Stanford - 43 (12 games) - 3.58" signifies that the 2019 Stanford team had 43 games missed by players who finished the season top 15 in snaps on offense or defense and, given that the team played 12 games that divided to 3.58 key contributors missing per game the team played.

2. Yes, PFF is my source for all of this.

3. One of the blind spots of this is that it would miss players who missed essentially the whole season with injury. It should account for starters who missed two thirds of the season or less as being a starter for only a third of the season would usually be enough to qualify for top 15 in snaps. This window into games played definitely misses guys who missed so much time as to not be key contributors, but all teams experience that to some degree and most injuries either occur sometime some weeks into the season and/or do not result in the player missing the entire season. My sense is this methodology generally captures things.

There definitely are contributors who this methodology misses, which is why if I were truly crazy I would have picked another methodology with a qualitative element where I assess all preseason depth charts as well to see who might be excluded by this approach. Or maybe run some sort of statistical analysis to figure out the appropriate cutoff for number of snaps rather than "top 15" (for instance, Tremayne and Wilson tied for #16 in snaps last year but were closer to #13 than #18 and obviously were "key contributors.") But I'm not that crazy. Given that these dynamics occur for all teams I think this analysis can provide general/directional insights even as I recognize it's definitely not perfect.

I also think that this approach of counting missed games by top 15 snap players seems to even out in the way it demonstrates the extent to which teams had to cobble together the requisite snaps from a mix of players. The teams that are relatively unscathed by injury have a much higher proportion of key players playing all or nearly all of their games. As an example, the 2014 Stanford team had 14 of the top 15 snap count guys play 13 or 14 games while the 2019 Stanford offense had 10 of the top 15 snap count guys play 11 or 12 games. In other words, this analysis perhaps could have looked at it that way in terms of how many of the top contributors missed negligible time. But my sense is that it would have been two different routes to generally similar destinations in terms of what the data demonstrate.

[msqueri]

Ok, you guys have gotten this bee in my bonnet and it can't get out without empirical observations. With full acknowledgement of the sbortcuts/flaws/blind spots in this approach, here is a list of every Pac-12 team since 2014 ranked by number of games missed by players in their season-end top 15 for snaps on either offense or defense:

1. 2021 Washington - 61 (12 games) - 5.08
...

Some questions about your chart.

1. What exactly is the number listed for each team? 
You stated, "number of games missed by players in their season-end top 15 for snaps on either offense or defense".
Number of games missed ought to be a whole number. Here something is apparently divided out. 
Is it something like average number of players missing per game? 

2. Where do you find stats for snaps played?  And games missed? (PFF?)
Or is the stat you list itself available somewhere? 

3. It would seem that this might account for minor injuries but not major injuries. i.e. a player with a major injury would miss enough time that he wouldn't end up in the top 15 for the team in snaps played. Does that seem correct?

1. The format "2019 Stanford - 43 (12 games) - 3.58" signifies that the 2019 Stanford team had 43 games missed by players who finished the season top 15 in snaps on offense or defense and, given that the team played 12 games that divided to 3.58 key contributors missing per game the team played.

2. Yes, PFF is my source for all of this.

3. One of the blind spots of this is that it would miss players who missed essentially the whole season with injury. It should account for starters who missed two thirds of the season or less as being a starter for only a third of the season would usually be enough to qualify for top 15 in snaps. This window into games played definitely misses guys who missed so much time as to not be key contributors, but all teams experience that to some degree and most injuries either occur sometime some weeks into the season and/or do not result in the player missing the entire season. My sense is this methodology generally captures things.

There definitely are contributors who this methodology misses, which is why if I were truly crazy I would have picked another methodology with a qualitative element where I assess all preseason depth charts as well to see who might be excluded by this approach. Or maybe run some sort of statistical analysis to figure out the appropriate cutoff for number of snaps rather than "top 15" (for instance, Tremayne and Wilson tied for #16 in snaps last year but were closer to #13 than #18 and obviously were "key contributors.") But I'm not that crazy. Given that these dynamics occur for all teams I think this analysis can provide general/directional insights even as I recognize it's definitely not perfect.

I also think that this approach of counting missed games by top 15 snap players seems to even out in the way it demonstrates the extent to which teams had to cobble together the requisite snaps from a mix of players. The teams that are relatively unscathed by injury have a much higher proportion of key players playing all or nearly all of their games. As an example, the 2014 Stanford team had 14 of the top 15 snap count guys play 13 or 14 games while the 2019 Stanford offense had 10 of the top 15 snap count guys play 11 or 12 games. In other words, this analysis perhaps could have looked at it that way in terms of how many of the top contributors missed negligible time. But my sense is that it would have been two different routes to generally similar destinations in terms of what the data demonstrate.