NFL retirement numbers
Posted: Wed Jan 28, 2026 1:08 pm
For curiosity's sake, I pieced together NFL retirement numbers going back to players that started in the league in 1995. Numbers should at least be close. Since our players don't age, these percentages won't make sense for us, but if the point of adjusting the physical percentages was to mimic the NFL regarding how many seasoned veterans are in the league, this should give us a target. And once the number of veterans are close, there should be a way to calculate the percentages to use for physicals to keep us in the ballpark.
If the league wants, these numbers could be used to come up with a total player arc so that the draft, aging, and retirement all work together to mimic these numbers. But that would mean making roughly 3 out of every teams' draft picks bad enough that they'd be out of the league in two seasons. That might not make the late rounds of the draft very exciting.
If there isn't interest to replicate the full arc, at minimum, I'd love to see the league use these numbers to determine the free agent pool size and number of retirees each season. Based on these numbers, NFL teams average 10 rookies per season (9710 players/32 teams/30 seasons = 323 rookies/season and 10.1 rookies/team) and just as many players leave the NFL every season. If we use the same number of rookies per team, 180 rookies should make it into rosters and the free agent pool every draft. And that means before the draft, exactly 180 players should be purged from the FA pool, including the latest retirees. Any additional players from the draft pool beyond the commish's formula for the top 180 should be purged - the PNFL wasn't interested. I know some coaches want more choice in the FA pool, but these players aren't good enough to be playing for your team - that is why they are in the free agent pool. They are now teachers, have a goat farm, are working at Home Depot, or are NPCs in GTA5. For example, it makes zero sense for any free agent to be good enough to plug-and-play into your roster halfway thru the season.
season retired_n at_risk_n retire_pct
1 2591 9710 27%
2 1479 7119 21%
3 1101 5640 20%
4 828 4539 18%
5 705 3711 19%
6 572 3006 19%
7 530 2434 22%
8 483 1904 25%
9 409 1421 29%
10 348 1012 34%
11 238 664 36%
12 171 426 40%
13 97 255 38%
14 69 158 44%
15 38 89 43%
16 23 51 45%
17 11 28 39%
18 10 17 59%
19 2 7 29%
20 2 5 40%
21 1 3 33%
22 0 2 0%
23 1 2 50%
24 1 1 100%
30 seasons of data 1995-2024
If the league wants, these numbers could be used to come up with a total player arc so that the draft, aging, and retirement all work together to mimic these numbers. But that would mean making roughly 3 out of every teams' draft picks bad enough that they'd be out of the league in two seasons. That might not make the late rounds of the draft very exciting.
If there isn't interest to replicate the full arc, at minimum, I'd love to see the league use these numbers to determine the free agent pool size and number of retirees each season. Based on these numbers, NFL teams average 10 rookies per season (9710 players/32 teams/30 seasons = 323 rookies/season and 10.1 rookies/team) and just as many players leave the NFL every season. If we use the same number of rookies per team, 180 rookies should make it into rosters and the free agent pool every draft. And that means before the draft, exactly 180 players should be purged from the FA pool, including the latest retirees. Any additional players from the draft pool beyond the commish's formula for the top 180 should be purged - the PNFL wasn't interested. I know some coaches want more choice in the FA pool, but these players aren't good enough to be playing for your team - that is why they are in the free agent pool. They are now teachers, have a goat farm, are working at Home Depot, or are NPCs in GTA5. For example, it makes zero sense for any free agent to be good enough to plug-and-play into your roster halfway thru the season.
season retired_n at_risk_n retire_pct
1 2591 9710 27%
2 1479 7119 21%
3 1101 5640 20%
4 828 4539 18%
5 705 3711 19%
6 572 3006 19%
7 530 2434 22%
8 483 1904 25%
9 409 1421 29%
10 348 1012 34%
11 238 664 36%
12 171 426 40%
13 97 255 38%
14 69 158 44%
15 38 89 43%
16 23 51 45%
17 11 28 39%
18 10 17 59%
19 2 7 29%
20 2 5 40%
21 1 3 33%
22 0 2 0%
23 1 2 50%
24 1 1 100%
30 seasons of data 1995-2024