2. Regardless of how it happened, Leffen had one notable win at Genesis. Top 50 "didn't exist yet" but that's why the algorithm works retroactively and calculates the season all at once; a win on WaDi in February is worth the same as one in June at the same size event.
-
-
Antwort an @PracticalTAS @HugS86
Just calculating wins that are ranked on the PGR becomes incredibly silly when you do this. How am I supposed to control that other people get upsetted? That ZD dropped off later but when I won vs him he was seeded high? Placements + bad losses should be >>>>> wins early IMO
1 Antwort 3 Retweets 157 Gefällt mir -
If meru goes and gets 2nd at every EU tournament ever, but only Gluttonny goes to the US and gets PGRd, he retroactively has 0 wins of note? It's such a bad fucking system.
1 Antwort 2 Retweets 161 Gefällt mir -
you can only beat the people in front of you, and if anyone gets upsetted, the other player who then beats the one who upset a higher seed, shouldnt be punished harshly for it. Yeah if you live in america you can roll the dice more times and eventually get a high PGR bracket.
3 Antworten 5 Retweets 133 Gefällt mir -
Antwort an @TSM_Leffen @HugS86
1. I specifically said top-50 wins for a reason. Our list of "notable" players actually includes *190* people; among those are your wins on Kofi, LingLing, and, yes, ZD. Beating someone outside the top 50 is still valuable and it would be a huge oversight if that wasn't.
3 Antworten 1 Retweet 41 Gefällt mir -
2. You're also rewarded for outplacements; 17th at Genesis gets you partial points against 59 players you tied or outplaced, including Glutonny, Nairo, and Marss 3. No, Glutonny has multiple Meru wins, and as Meru was included in our 190 people, those are valuable.
1 Antwort 0 Retweets 49 Gefällt mir -
4. At the same time, you have plenty more chances to buster out and lose points (see: Wizzrobe). Going to a tournament is not points-positive for everyone, especially if you have already been to a lot or perform very poorly.
2 Antworten 0 Retweets 35 Gefällt mir -
Antwort an @PracticalTAS @HugS86
Reread the meru tweet... And obviously I didn't think the other wins/placements have 0 impact, the point is that placements should be by far the #1 most important stat, not wins.
2 Antworten 0 Retweets 39 Gefällt mir -
Antwort an @TSM_Leffen @HugS86
Rereading the meru tweet, it's unclear who "he" is referring to. And we try to keep the system balanced between placements vs. wins; if you think placements should be worth much more than wins, the top half of the list would not look like what you'd expect it. I have to disagree
1 Antwort 0 Retweets 27 Gefällt mir -
Antwort an @PracticalTAS @HugS86
The current list looks nothing like it should anyway, so that kinda argument is completely shit. Its just an inherent problem with small sample sizes + algorithms. Fairness over what gives you the top 10 you would expect.
1 Antwort 0 Retweets 18 Gefällt mir
Otherwise you're essentially just making a one man panelist ranking rather than actually making a fair system.
-
-
Antwort an @TSM_Leffen @HugS86
Sample size isn't that small, there are 10k relevant data points in the database. How about "fairness AND what gives you the top 10 you would expect"? You can tune the algorithm based on Smash 4 and still have it spit out a good result for Ultimate...oh, that's what I did.
2 Antworten 0 Retweets 27 Gefällt mir -
The difference between 1 and 31 is smaller than the difference between 31 and 111. The bottom end of the list is already bordering on statistically insignificant, that's why we tie everyone in Area 51 at 51st. Individual tournament runs move players 10+ spots down here.
3 Antworten 0 Retweets 25 Gefällt mir - 3 weitere Antworten
Neue Unterhaltung -
Das Laden scheint etwas zu dauern.
Twitter ist möglicherweise überlastet oder hat einen vorübergehenden Schluckauf. Probiere es erneut oder besuche Twitter Status für weitere Informationen.