These rankings are panel-based, and our panelists were asked to weigh both 2021 performance and 2022 potential in these rankings (to see the exact prompt, read to the FAQ below). Because this was not an official ranking period, and players clearly treat tournaments in official ranking periods differently, from their attendance choices to tournament mindset to other factors, this ranking will not be presented as a ranked list like the PGRU or the MPGR, nor should this be treated as the equivalent to a PGRU or an MPGR.
Instead, the players will be presented in a tier list form, with the only "ordered" player being the top ranked in the S+ tier for each region. The full list includes 30-32 players per region, with S+, S, A, B and C tiers for each region and game. The decision to move from a fully ranked list to unordered tiers reflects our attempt to account for the uncertainty our competitive scenes were experiencing while still honoring the achievements of the players who were able to show up and compete at the majors that were held in 2021.
The players you see on these lists, to be released over the next four weeks, are not a complete list of players who will be in contention for the new PGRU and MPGR seasons come 2022, but based on 2021 performance and potential, our panelists believe these are the players who have shown they will for sure be part of next season's conversation.
For the full schedule of releases and the methodology behind these rankings, read below for a summary from PGstats's Head Researcher, Andrew Nestico:
Have questions about Path to the PGR: Contenders Tier List? Hopefully the answers you want are here!
#PGRContenders will be releasing as follows:
B/C tiers (article only): Feb. 15
S+/S/A tiers (article and video): Feb. 17
We reached out to Japanese TOs, and due to the amount of uncertainty surrounding offline play in 2021, the Japanese Melee and Ultimate scenes both requested not to be included in these rankings.
#PGRContenders consisted of separate regional panels for the North America and Europe regions. Ultimate NA, Melee NA, and Melee EU panelists underwent the same balloting process that has occurred for past #MPGR releases, while the Ultimate EU panelists were granted their request to set their list by consensus in a group chat. The final result is the same for both, and tiers are unordered besides calling out the overall #1 player for each region and game.
Panelists were prompted with the following question:
“Based on results in 2021, rate each player from 10 to 4 on the following scale: S+ (10-9.6), S (9.4-8.6), A (8.4-7.6), B (7.4-6.6), C (6.4-5.6), untiered (5.4-4.0); you may use x.5 to denote that you feel the player is between two tiers. Scores will be averaged, so if you feel a player is firmly in a given tier, you should score them x.0, with other values being available if you think a player performed slightly better or worse than the others in their tier. Your rating of a player should take into account both your opinion of their performance in 2021 and your expectations of them going forward into 2022.”
Approximately 50 players qualified for inclusion on each individual regional ballot, and each panelist gave them a rating between 10.0 and 4.0. Panelists were also allowed to abstain from scoring a particular candidate if they felt they did not know enough about them; this had no effect on the player’s final score, but if enough panelists did not rate a player, they would be disqualified. (This time around, this occurred for one player who would have otherwise made the Melee NA list, Kurtis “moky” Pratt.)
Panelists scores were then normalized for consistency, and each player’s final score was calculated using what is essentially a weighted average, with higher weights given to an individual rating the closer it is to the interquartile mean of the given player’s scores.
To create the tiers, the resulting score lists were analyzed using K-means clustering. For all three regions, tier lists with varying numbers of tiers were generated and evaluated for best fit (all split points listed in the prompt are near a generated tier split, and there is a generated tier split after the 30th or qualified 31st player). Finally, generated tiers were combined into a single #PGRContenders letter tier if applicable. For example, one tier containing players with scores of 7.5 to 7.0 was combined with another containing players with scores of 7.0 to 6.5 into a final B Tier.
As noted, the Ultimate EU panelists set their tier labels directly, within the following constraints:
All players on the list satisfying eligibility requirements
All players within a given tier assumed to be tied
Additional flexibility of +-1 player outside a guideline with my approval except the minimum of 1 for S+ and 30 for the whole list”
Ultimately, the panelists requested that the length of the list be expanded to 32 players so they did not have to break a tie at the end of C Tier; this request was granted, and theirs is the only list that contains more than 30 players.
