01-24-2024, 12:59 PM
(This post was last modified: 01-30-2024, 04:35 PM by lemonoppy. Edited 2 times in total.)
Chaos Project #1
Player Retention
Player Retention
As promised, my retirement from GMing has given me a bit more time and energy to spend on league-related projects other than GM tasks, and I’m very excited to start sharing them with you. Now, if truth be told, this particular one just came to me yesterday, but I think it’ll be pretty interesting and informative nonetheless.
As a GM, my general rule of thumb when it came to drafting and player retention was “Half after one season, and half again after three”, meaning I was only expecting half of a given class to stay active for a full season and only half of what remained would still be active after three, but at a certain point, I started asking myself how accurate this actually was because, in all honesty, I can’t even remember where or when I got this information. It could be something that was passed on to me 7+ years ago in SHL, possibly early in this league, or maybe I just pulled it out of thin air, so I decided to do some digging.
Using our handy dandy TPE tracker, I scraped update data on every player available, both active and retired, but not all of that is necessary for what I’m trying to do here, so I filtered that down to players from classes S24 to S33. Simple reason for those parameters being that the league schedule was expanded to 16 games in S23, so S24 would’ve been the first group impacted by the elongated league season, and on the other end, S33 is the most recent class with no active players remaining, thus giving me 10 classes worth of players with full careers of update data. This sample gives us 800 players exactly, oddly enough, so that’s an average of 80 per class, with most falling somewhere in the 70s.
Now, as you may or may not know, the first update that the tracker pulls for every player is the base 50 TPE we all get upon creation. Of those 800 players, 127 never had a positive update after that initial pull. That’s nearly 16% gone right there, which is not very encouraging. In fact, I’ll take this a step farther before I move forward. When it comes time for the ISFL Draft, all inactive players with 57 TPE or less are automatically removed, so how many of the remaining 673 never had another update greater than 7 TPE? 55. That gives us a grand total of 182, or 22.75%, that we’re losing right off the bat to the 57 TPE rule, so I’m not feeling too bad about my half & half rule to this point.
Here’s where things get a little bit hairy, though, because not every player is created equally. Depending on when people create, we could have players ranging anywhere from 4 weeks to 10 weeks old, or more if there’s a holiday break, so basing my data strictly on the number of tracker updates for each player isn’t the most precise, but it gives you a pretty good idea of what’s happening. For the purpose of this exercise, I’m going to assume that players are being created in the ideal window and equate 1 full season to 8 weekly updates, as well as use the “update greater than 7 TPE” benchmark to constitute some sort of effort to participate.
Using the 8 week mark as our breakpoint for a player’s first full season in the league, we now arrive at the number 295. That’s 295, or 36.88%, of the initial 800 players in this sample who never posted an update greater than 7 TPE, meaning we’ve effectively lost 113 people at some point during their DSFL rookie year who at least tried to get started. Don’t get me wrong, 37% is not good, but it’s certainly better than the half I’d been assuming.
I realize I’ve been a bit too negative here in highlighting the numbers we’re losing, so let’s flip the script a bit. How many of the poor souls that have stuck with us through their ISFL Draft are hanging on for year 2? This is the weird gray area for a lot of people where you’re now in 2 LRs, trying to engage with your ISFL club, while still playing in the DSFL. Good news, we have 434 still with us after tracker update #16. Now, we’re getting closer to that halfway mark, but still above it after our ISFL Draft year.
Onto year 3 we go. Hopefully, everyone’s getting acclimated to their ISFL teams and getting called up, but we’ll still have some slower earners being sent down, and that’s perfectly ok. How are we doing once we get to week #24? We inevitably dip below the 50% mark with 379 remaining, but that took longer than expected, and we’re losing people at a much slower rate. By my “half & half” rule, we should be down to about a quarter now, but instead, we’re just barely under half at about 47%. Boy, am I glad to be wrong, and since I was so far off through the first 3 seasons, I’ll extend this a bit to see how long we can stay above that 25% line.
Now, we get to year 4, where most users that are still earning have maxed out their DSFL builds and hopefully are excited about being called up to the ISFL rather than leaving a 250 TPE corpse to rot as a filler. In this range, we’ve only lost 42 people, or an average of about 4 per season. For a league with 14 teams, that’s not too shabby as long as they’re not all coming from the same team.
Onto year 5, and it’s make or break time for those slow earners still straggling in the DSFL because their eligibility down there has now expired. For the max earners, these guys are really starting to make an impact, getting their lifetime contracts and pushing for that 1,000 TPE milestone. As far as retention, we’re holding pretty steady at this point, only dropping another 36, which is pretty well in line with our last update, leaving us with 301 active users after 5 seasons.
We’re seeing a trend here the last couple seasons as the curve flattens out a bit in the 4-5% range, so I’ll skip ahead a little bit. Let’s talk about everyone but Mooty’s least favorite thing. That’s right, the R word, regression. That starts after year 8, so how are we holding up to this point? No surprise here that people aren’t a fan of giving back TPE they’ve rightfully earned, as we see a little bit of a spike in the numbers, with 59 people making their final updates in the week 57-64 range. For context, my most recent update for my first round of regression on Piggott was tracker update #70, so maybe my estimate is a little early, but you get the point. What this ultimately means is we have about 22% of our initial sample still kicking, trying to fight regression in some shape or form.
One last group I want to highlight is the real fighters, the grinders, that are willing to fight regression all the way to the end. In this 10 season sample, we had a total of 45 players make it to at least their 100th weekly update, and just 10 of those made it past 110. I’ve played 11 ISFL seasons with 2 of my players and the full 13 with another, so I know how brutal it gets by the end, and I tip my cap to the brave few willing to go the distance.
In summary, my rule of thumb on player retention was wrong, but not terribly so as 63% of creates are making it to their draft day instead of my assumption of half, and 75% of those remaining after three seasons as opposed to my “half again”. For those of you still with me, thank you for reading. I know this is a lot to digest and a lot of numbers being tossed around, but I hope you found it as interesting as I did.