LumiRank | Stuart98
@Stuart98SSB
Followers
2K
Following
5K
Media
161
Statuses
1K
Algorithm creator and other stats stuff at @LumiRank, moderator of @r_smashbros Previously: Creator of ΩRank, fact-checker and TTS maintainer for the PGR.
@stuart98 on discord
Joined July 2022
RT @tradewarhj: “Why Do People Care So Much About Rankings?” - A Measured Response.
nutradewar.medium.com
I redownloaded Medium on my phone for this.
0
6
0
RT @doramigi_1008: LumiRank 1位！！！！！！！！！.ずっと目標にしてたからめちゃくちゃ嬉しい！！！でもこれに満足せず後期も頑張りたい！！！俺すごいおめでとう😭😭
0
994
0
RT @SuperSmashCon: It's time for the Top 10 reveal of the @LumiRank 2025 Half Year Rankings!. Hop in ⬇️.
0
5
0
RT @SuperSmashCon: These are the best of the best. The Top 10 @LumiRank 2025 Half Year Rankings will be presented tonight at #Supernova202….
0
17
0
0
46
0
Apparently I need to start running my MTL'd tweets through more translators so that when I try to say "you fixed the flaw in your season that caused you to be low" I don't accidentally say "I fixed the flaw in the ranking that caused you to be low." Smh.
2
3
39
I went down the prototype all-time ranking and my new answer is Rotsuku the yoshi. Retrorank puts him top 100 for all of prequarantine (while Brood misses by a bit due to inconsistency) and has barely gone to anything postquarantine but has done pretty well when he has.
I mentioned Shu for Melee/Smash 4 and Prand Grix for Ultimate but the answer here is actually Brood. Not ranked in the PGR or OrionRank for 2019 and obviously had a lot of Smash 4 history. Thanks to PSIguy for this.
0
2
31
0
55
0
Makoto the Sheik is that you?.
Finally made it, 87th in the world I am no longer in a crisis core, Sephiroth🗡️ the number 1 soldier on the rise ✨. The battle doesn't end here let's go even further beyond!!!. Remember anyone can do it but it's going to be easy, a lot of strong warriors out there. @SchuStats
1
1
6
How much discourse will I cause if I post the algorithm's output if you freeze player values at what they were last season (equivalent to just not iterating at all).
9
3
116
0
62
0
0
95
0
as well as their lifetime attendance overall, with higher attendance in a season giving higher weight to it and higher overall attendance giving more weight to their best seasons.
1
0
9
Generating an all-time score for each player will work like a simplified version of how the algorithm scores players during a season: Weighted average of their score for every season, based on the base weight of each season, the player's overall attendance during that season,.
1
0
9
Establish a base weight for each season using the total value of all tournaments in that period (so eg 2023 will be more valuable than 2020.2-2021). Take each player's unscaled, attendance agnostic score from each season, with a 0 used in seasons they had no attendance.
1
0
12
Ranking periods:.December 2018-2019.1, 2019.2-2020.1.2020.2-2021.2022.2023.2024.1, 2024.2.2025.(?). 2023 onward use official LumiRank releases for each period, prior periods are fully retiered and reevaluated using the current LumiRank algorithm.
1
0
17
Been doing a lot of thinking on LumiRank all-time. Nothing set for when we make and release it yet but we do intend to at some point. Current plans (subject to change, a thread):.
3
4
93
RT @TheThiefOfLight: Hey, do y'all remember when I made a 70 minute documentary about the history of Kagaribi 3 days before the biggest Sma….
0
20
0