(S26) Ultimus Week

[Troen]

2) Tell us about your draft class. Is there anything that makes it special in comparison to other classes? Where did your fellow draftees land, how are they doing? Did anyone turn out to be an unexpected steal of that draft based on what you know today? Do you think anyone in your draft class will become a hall of famer? If you’re new to the league, how do you think your class will do? Where do you think people will get drafted to?

The S22 draft class should already be well-known to all as the biggest reddit-sourced class, with over 270 members at the time of the draft. It led pretty directly to 4 expansion teams and is now in the position of taking over the top TPE positions in S27 and onward. Really, it may well be the most impactful class since the start of the league.

However, since I've pretty much always enjoyed the quantitative nature of the league, I'm going to quantify some of the impact.
Looking at the positional stats for the ISFL (actually still called the NSFL on that page, heh), the average TPE per position is:
QB: 1013
RB: 766
WR: 738
TE: 660
OL: 598
DE: 646
DT: 758
LB: 688
CB: 768
S: 731
K/P: 657

Here's how many players in the S22 class have more than the average positional TPE:
QB: 5
RB: 5
WR: 7
TE: 5
OL: 5
DE: 6
DT: 5
LB: 8
CB: 8
S: 8
K/P: 3
In total, that's 65 players from S22 who are above-average players for their position. Now to be fair, looking at averages might be slightly misleading since backups could drag the average down a bit, but that's still a clear sign of impact.
Looking at it a different way, the average team has about 23 players on its roster right now (326 players in the ISFL per the team stats page). The S22 class could make up nearly 3 above average teams on its own.

Looking at TPE earning a different way, here's the rounded earning rate of the top TPE earner in each non-regressed class (per the TPE tracker as of 1/10/21):
S20: 165
S21: 183
S22: 185
S23: 190
S24: 196
S25: 199
S26: 201
S27: 217

(Calculated as (highest TPE of top player - 50) / (28 - season) - ie S20 players earned TPE in S19, 20, 21, 22, 23, 24, 25, and 26, so 8 seasons)
This decline over time follows naturally given the rookie tasks, decreasing amounts of TPE
from training camps, and the higher likelihood of missing a TPE earning task for some reason eventually. How do those max earners compare with the 75th and 90th percentile earners for the class?
(Side note: I'm looking at current class information per the tracker, which doesn't include retired players it looks like. So, the stats aren't perfect vs. the initial draft class. I believe but didn't go check that S23 is the class they started auto-retiring 50TPE players with no updates)
S20: 24 players, 11 with under 100 TPE who are ignored as non-earners. 90th percentile (2nd/13 remaining): 160/year. 75th percentile (4/13): 150/year
S21: 76 players, 34 with under 100 TPE who are ignored as non-earners. 90th percentile (5th/42 remaining): 175/year. 75th percentile (11/42): 151/year
S22: 181 players, 85 with under 100 TPE who are ignored as non-earners. 90th percentile (10th/96 remaining): 177/year. 75th percentile (25/96): 163/year
S23: 49 players, 2 with under 100 TPE who are ignored as non-earners. 90th percentile (5th/47 remaining): 182/year. 75th percentile (12/47): 163/year
S24: 49 players, 0 with under 100 TPE who are ignored as non-earners. 90th percentile (5th/49): 188/year. 75th percentile (13/49): 179/year
S25: 86 players, 1 with under 100 TPE who is ignored as a non-earner. 90th percentile (9th/85 remaining): 188/year. 75th percentile (22/85): 183/year
S26: 58 players, 7 with under 100 TPE who are ignored as non-earners. 90th percentile (6th/51 remaining): 191/year. 75th percentile (13/51): 175/year
- Note the tracker currently has 4 GM bots; I ignored those
S27: 70 players, 31 with under 100 TPE who are ignored as non-earners (obviously this overcounts a bunch of active slow earners vs. other classes, but I went for consistency; a more rigorous analysis would probably define non-earner differently per season). 90th percentile (4th/39 remaining): 201/year. 75th percentile (10/39): 188/year

So, looking at the earning rates of the high performers, I actually think that S22 doesn't really stand out. In some ways, this makes a lot of sense - the league only offers so many ways to get more TPE. The percentage of people per year who are near-max earners seems to be consistently over 25%, as well. S22 still stands out as the largest class, but it's only 11 ahead of S25 in terms of number of people who stuck around for at least a couple weeks.
S22 and S25 actually do have a similarity if you look closely at the 75th percentile trend line - S25 actually goes up vs. S26, while S22 holds even with S23. I think this is the strongest signal for those classes being unusual, but since they're also very big classes I can't quite envision the cause - are people joining from reddit more likely to remain active? Do larger classes help activity? I know that a lot of people when I joined formed rivalries and had media poking at each other as well as Twitter wars. Since the total number of people in the league went up as a result of those classes, though, I'd think that there would actually be more things to keep people interested and active over time. As an alternative hypothesis, maybe it's the number and concentration of people doing the same thing? With a lot of people joining discord, figuring out how to do everything, it would make sense that there's a ton of connection-forming. Compare the newly arrived people who are trying to figure everything out and spending a couple hours on the forums and the main discord to people like me who mostly hang out on my team's server and only read the media section of the forum - maybe there's something to be said about making connections with someone from the group.
That does raise another (mostly unrelated) thought - I wonder if the league would ever consider a recruiting post much closer to the start of the league year so that new creates could be on a team almost immediately? I know that for me it was something like 3 weeks from creation to drafting, and it really felt like a whole lot of nothing for a while. I'm guessing that waivers would be a nightmare to deal with if 50 new people all showed up at the same time, but I have to wonder if that would be a better player experience and help people to come back.

So anyway, getting back to S22: My class was clearly impactful to the league, but mostly for size rather than abnormal quality. But, quantity has a quality of its own, y'know. I do expect that the next couple years are going to see a lot of S22 players on championship teams - 8 of Yellowknife's 22 players were from S22, for example (and all had TPE over their positional averages), and it's interesting to see that they also have 5 S25 players on the roster. I expect that the history of the league won't be complete without a selection of the top of the class and understanding how it led to the expansion teams. My personal theory is that the good timing of the reddit post (fairly soon after the Super Bowl so that people were keeping their eyes on football news but when it wasn't drowned out by games) along with getting followed pretty soon after by lockdowns and travel restrictions (making it easier to be/remain active) will make it unlikely that we'll see another single class that reshapes the league so drastically ever again.