(rubs temples)
Mar. 1st, 2023 09:23 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
goaltender
why is it impossible to do any kind of significant analysis with existing ao3 scrapers for hrpf without suffering because the archive is 60% locked! been thinking about hrpf ships again and how it's so hard to do overviews of """popular""" ships the way you can with like f1 where there are only 20 active drivers + a few straggling retirees to care about because the usual quick sidebar filtering really is not sufficient in uncovering trends and blorbos... like i want to argue that hockey fandom is actually active-ish in discovering new ship narratives and making edits and posts for them on tumblr/twitter, or like... idk, there are SOME concerted efforts from fandom to identify >This is a potential ship/pairing for this team!< etc., but that identification just doesn't really translate to a large quantity of fic because there are too many combinations spread out across "fandom" (maybe trying to call it one fandom is a disservice to the Hockey Community as a whole too but that's for another day).
like the problem with hockey really is that 1) the Old Established Ships (sidg, kzr) remain near the top after all this time, which when you think about it is kind of a reflection of "live die retire with this team" LegacyCulture that hockey is especially prone to, and then also 2) brother forget about PLAYERS we've got 32 teams in the league! and assuming 23 roster spots that's 253 possible combinations per team and 8096 across the league *without* any cross-shipping which of course is a huge aspect you can't overlook. like sure i'd say at least 70%? of roster spots in a season will never have fic written about them but there still lies some value of ambiguity in the possibilities there. basically what i'm getting at is that maybe you have to set your expectations SUPER low to engage in hrpf analysis and that it's necessary to reframe something like 10 fics in a year for one ship as being actually kind of good or as being one of the top pairings for a team's fandom and not bottom of the barrel niche "rarepair material"....... in this day and age. and i know like "what are all the ships in a fandom" is very basic data collection but just because of sidebar filter limitations + nature of archive locking these "small-ish" pairs are hard to get pure numbers for unless you go through the trouble of scraping collective tag metadata yourself.
like this was from my old post on girlrock but it's a whole different beast to be finding lifetime ships that have a high enough threshold of fic you can easily sort out and then like... ships that people might kind of care about in niche subpockets of fandom over a specific timeframe but that haven't experienced enough longevity for you to really speak on yet. like i remember being mildly surprised that fandom characters like jh86 were still so far off the map lol
anyway wrote literally the dumbest python script ever so i could just extract relationship tags specifically from all the hrpf fic published since jan 01, 2022, aka the last year + 2 months. i chose this timeframe because a) too lazy to do more and b) i'm curious to understand what its RECENT fandom trends have looked like and i thought a year would give enough context in showing new narratives/ships without restraining the sample size too much. then i aggregated all the instances (i had 1457 total tags) in a spreadsheet and did some very arbitrary cleaning... this is extremely not sophisticated work lol but basically >collapsed "(implied) (past) (one-sided)" etc. tag identifiers into one (in the end i still had some minor fic count discrepancies with the biggest ships which i'm attributing to server-side tag formatting shenanigans but the data seems mostly correct from the others i sanity checked), >deleted platonic relationships, wag ships, gm/hc ships, "reader" ships, and... that's it. also had to wrangle threesomes + hughesbro lastname formatting somewhat clumsily which makes me realize i didn't do this properly for the tkachuks but whatever just use context clues. it's fine. and i always struggle with team assignments sometimes because i think it's best to default to current team no matter what for objectivity + consistency's sake but it's also like... i'm very much aware that for some of these ships every single fic was written when all the players were under a certain team and i worry about eliminating important context in pretending otherwise. alas....... BYOCC (bring your own context clues).
tbh i don't know IF i really found anything out from doing this lol but it's just fun to make lists. i ended up with 88 ships that have had 7 fics written using their relationship tag since jan 01, 2022. you may be wondering "why did you choose that as the cut-off" and it's just that 5 seemed too low and then 10 felt too limiting... yeah it's that dire out in hockey land lol so i was like 7 it is. i don't know man. here's a slightly choosier overview of all the ships that have had at least 15 fics though:
not a surprising turnout but it's useful to have a lot of the cross-team ships contextualized in a larger table. that is, a majority of these ships follow the usual teammates/used-to-be teammates archetype (usually 2 young star/promising players, C and his A, or in the 1386 case both), but you obviously also have new-age cheesby (mattdrai) and the somehow still-breathing carcass of mceichel. then the next few "seemingly" cross-team pairings are all from the umich dynasty (opkj, brissbords) or its adjacent ntdp implications (jhtd). it does kill me that in the end no one is quite as cross-team bicycled as the Two Main Generational Talents but i feel like the circumstances surrounding how and why they were shipped with those people are so different... like crosby/mackinnon is such a different beast from mcdavid/strome lol but to see them stand out in a sea of teammate4teammate is kind of knocking me out. even though i do not care about him kind of sad that dstrome has been cosplaying as adjusted normie girldad ever since dildo episode because i feel like it would have been funny if he'd managed to revive rpf blorbo narratives for himself in washington lol. i expressly do not follow the caps though so i wouldn't even know if it happened
also kweekes voice Keep an eye out on Number 29 👁 the fact that they aren't that far off of bennguin lol......... the stars are in a new era fr it's just that when things peaked they Peaked and now every effort pales in nostalgic comparison (for collective fandom that is because i was bennguin hater as a teenager so i actually actively prayed on their downfall. but that's neither here nor there). maybe fandom can just be the same 5 people engaging in characterization sewer ball via convoluted tumblr tags
also looked at some team-specific stats to better contextualize team ficdom success/representation... florida feels like such a poser because it's really just tkachuk/guys who were traded but whatever... i just think it's interesting to see what the # of fic for the biggest ship on each team has been (as in any ship the team is represented in, even if it's cross-team!) because some of these teams are STRUGGLING; it's really fascinating to realize that no ship on the canucks has surpassed 16 in the past year+ when epbb enjoyed a relative amount of fandom success during their peak... 189 lifetime works but only 16 since the end of 2021 and 4 in the entire 2022-23 season lol.
^ you can also see at the bottom that only three teams (TBL, ARI, STL) don't have a single ship with at least 7 works since jan2022. which does kind of make you wonder when the cup-winning teams from 2009-2018 were PIT x3 (8771; C/A) CHI x2 (1988; C/A) BOS (pbbm; C/A, even if not at the time) LAK (jcmr) and WSH (819; C/A) and then since then we've had STL, TBL x2, and finally COL lol........... hm. the thing that also fascinates me is that like... ok i don't want to get too into it but i just find it interesting that 1988 are still as big as they are sometimes even though stl are like established villainous team and everyone hates binnington etc. and i'm not saying that People who still like the hawks and People who hate the blues are inclusive categories (actually they probably are but for divisional reasons not ethical consumption of fandom reasons) but it's like ....? clearly no amount of character death can deter the most passionate of fans so i think it really is old lore and habits dying hard.
also it seems like metro has the best parity/relative consistent spread across team fandoms while central is a true deadzone........ sadge
dallas breakdown
boston are also on top for "unique players represented" which is interesting because trevor zegras, jack hughes, etc. get bicycled a lot but boston truly have Established Pairings.
anaheim lol
oh maybe i should actually evaluate player bicycling one second. the sebastianvettelification of sidney crosby.............. guys with at least 24 works (from the ships with 7+ works):
this took me back to my "what's in a pick" analysis and how the median draft position for the top 20 lifetime ships was 3.5 (mean 18.08, mean for the better half of the pairing 5.65)...... was wondering before i color-coded the table whether propensities have been different as of late with recent trending ships, but it didn't feel like anything would have become more "equitable" just from my eye test and intuition LOL. and...
yeah that tracks.
i'll stop here 😔🙏 why does this latest tripleS ep kind of serve cunt severely btw...
edit: omg i need to fix giroux -> ott LMFAO i'll change it later maybe
no subject
Date: 2023-03-02 11:17 pm (UTC)i love that i'm like Here's my Vibes Read (made up) (highly personal) and then you go and find actual evidence to test a hypothesis.... slay research legend
it's also useful to see the lifetime of the archive kinda stuff bc it's difficult to remember that eg. richiecarts were so exponentially More Niche/limited than the big kazer+sidgeno popularpairs. how can there literally be under 300 richiecarts works ever... this is my mandela effect subreddit moment.
idk if you're happy to share but i'd be interested to know how you did the scraping of archive locked stuff! i don't know python but would be cool to see how you got around it :^) even seeing what proportion of fic is unlocked vs locked now compared to ten years ago would be interesting tbh
the shiplist for the last year is so interesting but also patches/mark stone is so random. is that basically the only contemporary old man ship that had to be invented *while they were already old men* and without extensive rookie lore? maybe since vegas was still so new as a team they got rookie-beginnings vibes just from that....
thank you for convincing me that stars have a chance and it's not an entirely futile endeavour! looking at what else is happening with other teams is better than reflecting on the past hey..... like At Least It's Not Minnesota. Even tho minny has such an incredibly rotateable top line lmfao. I kinda knew boston had the goalietandem ship going for them still but i hadnt truly respected that the other stuff was on the rise too (outside of marchy/bergy). boston stand out in my memory from 2010-14 era bc they had strong team-centred fests. idk if that was random that i recall that specifically, it's quite likely other teams were doing the same fest stuff, but they were that kind of edge-case of popularity where it seemed to be super productive/valuable at the time to keep things cohesive and churning along. who knows if there's still ghostly influence from that era
jr21rh24 barely squeaking over the 7fic mark :tearful::proud:
> no taylor hall appearance in recentfic
oh how the mighty have fallen
> no rasmus dahlin
king keep your head up your crown is slipping!!!!!!!!
anyway thank you for this, truly scratched my investigative itch incredibly satisfyingly, and very interesting to see your approach and what you wanted to compare and analyse. very cool!
no subject
Date: 2023-03-08 12:12 pm (UTC)also my methodology was so genuinely stupid aldsfkahfds i'm happy to share it's just extremely inefficient because i realized that sometimes trying to automate everything takes more time than doing parts of it manually... i basically had to download all the pages i wanted myself (not specific fics just the archive pages since you can extract from the previews) which took the brunt of my time but then i looped through everything with beautiful soup to extract and aggregate the relationship tags and send all the values to a csv. i have seen ao3 scrapers that let you do user session log-in stuff but i'd have to try it with a dummy account at some point because i'm paranoid LOL and also since you have to include a 5-second delay to not get blocked all scrapers are still pretty slow but i guess it'd be good to figure out an efficient version at some point just to have something i can reuse quickly that requires the least amount of active time/effort
but re: unlocked vs locked i actually had an old graph i made except for F1 which was a thought exercise i engaged in as well because i was like is sports rpf getting more public after seeing how insane people are on twitter sometimes/the perceived dissolution of the 4th wall... maybe i'll play around with a few more hockey stats later
also yeah the sample sizes are obviously significant here (or i should say actually insignificant) because sometimes a ship with 6 works is really just one person but like i still consider that meaningful in this context because it's beautiful for one person to singlehandedly make a ship happen you know... ANYWAY the pacioretty ship count is basically entirely from 2 people + a ton of Anonymous fic which i presume to be from a third singular person so there's that. respect for real
alsooo the stars are fun because despite bennguin being on top robo is the most shipped by virtue of being paired with jake AND roope. lfg 7 robohintz works!!! and oh man i totally forgot about the existence of fic fests... i just realized does this stuff also get organized on discord these days because i'm so anti-discord server i'm probably missing out on like massive amounts of behind-the-scenes Fandom Collection And Building that completely skews my understanding of digital growth. oh no... time comes for us all. we need to bring back public forums and accessible centralized content 😮💨
and truly taylor hall was not terrorizing the edmonton nightclubs tag teaming with ebs bartering gucci bags for anal and snorting so much coke connor mcdavid snapped after a year and had him vibe checked out to NEW JERSEY for him to fall off this hard... rough new world we're living in.
ras is fun and charming to me!!! i honestly think people just hate shipping swedes and i know we keep going back on forth on the state of euroshipping these days but it still baffles me a little bit... anyway sorry for late and incoherent reply but thank you as always for reading & enjoying! i appreciate it a lot T__T
no subject
Date: 2023-03-12 02:50 am (UTC)dling the actual archive pages and then scraping/sorting those is a nice middle point tbh and sounds v cool as a problem solution! i get what you mean about just Manually doing part of it being faster than spending days trying to automate it lmfao. it would be so useful if ao3 had actual back end api but like I Guess They Definitely Have Bigger Problems....
the f1 publicness graph im kinda like HUH.... idk what this means but huh!!!. thank you for the data vis though.
crying at robo of all people being team bike we'll manufacture the vibes if it kills us
every time i remember that so much fandom stuff is happening on discord i die a bit coz like... its not even just zoomers, i think a lot of the hags too are in hag servers and stuff.... re: our prev convos it really is maybe that cliques are now cliques in private on discord and only rare groups let it leak onto public platforms... TT_TT. even locked twitter being a whole other ecosystem that's basically inaccessible or fairly high-fence to get over is an issue.... which is good i guess since this is rpf we're talking about but lol
:^) :^)