Features

#PGRv2 Methods

Sat 14th Jan 2017 - 11:42pm

 

Methods

 

What is X-Factor?

The X-Factor is the literal difference between the objective rating competitors earned after having tournament set counts and placements calculated and the subjective rating competitors received by the vote of the panel.

The panel is comprised of community leaders, tournament organizers, and top players from around the world who were surveyed months before release. The panel was asked to rate competitors on a scale of 1 to 10 and the averages of all of these ratings were then taken and ordered from greatest to least.

The X-Factor rating is designed to dilenate objective standings versus community perception and does not determines a player's ranking on the #PGRv2.

 

Defining Parameters

We ranked players according to their placings and set counts against each other, firstly, in the 6 Tier 1 events in order to create a basic framework to analyze the 50 players.

A player’s #PGRv2 is a purely based on objective findings: tournament placings and set counts. We then assembeled a team of 30 panelists (names listed in Methods) who also ranked players according to personal opinion. The X-factor is the difference between the final spot on the #PGRv2 and panelist ranking. A +3 X-factor means the panelists ranked a player 3 spots higher than they ended on the #PGRv2 while a -2 means they were 2 spots lower on the panelist ranking. An X-factor of 0 means both the panelist ranking and the #PGRv2 were the same.

 

Our qualifications for a Tier 1 event (Majors/Premiers) include at least 2 of the following :

 

  • At least 512 entrants
  • 9 rounds in Winner's bracket or more
  • 20 PGR members or more

 

Afterwards, consideration was given to the Tier 2 (Nationals/Super Regionals) listed in figure 2 of the appendix, listed as #s 7-49, which include at least 2 of the following: 

 

  • 256 to 512 entrants
  • 8 rounds in winner's bracket or more
  • 10 PGR members or more

 

With these placings in hand, rankings for players were ordered with their highest placings before them with notable players having the highest amount of placings that were within the top 8 at majors.

 

Basis for Metrics & Standards

After placings were organized, players that were not in attendance for most of the majors were considered for their placings in Tier 2 events (Appendix Figure 2).

There was no such player that was ranked without at least some sort of placing at 1 of the 6 Tier 1 events.

After a preliminary list was created and the highest placing players were evident, namely 1-10, set counts were researched.

Each player that had 1 won set or more over the top 10 was taken into consideration and the top 10 players that had wins on each other was then totaled with the results of their set counts being recorded.

It was at this point that the top 10 was solidified and ranks 11 - 50 were ordered according to this standard of placings contextualized by set victories” meaning, ranking values were higher depending on who was beaten/lost to in order to get there.

For example, if a player had 3rd at an event but did not face anyone in the Top 10 to get there, that placing was undervalued compared to someone who received 5th at a slightly larger event in which 1 - 3 names on the Top 10 were beaten in order to get there.

Then, throughout #11 - 50, sets were evaluated with the biggest emphasis being placed on the set wins that were on people that were above their own ranking.

 

Meaning, 11 - 20 was largely influenced on their wins over the 1 - 10.

The 21 - 30 was largely influenced on their wins over the 1 - 20.

The 31 - 40 was largely influenced on their wins over the 1 - 30 and etc.

 

Due to this methodology, it was rare for someone to have repeated documentations  of losses to anyone that was below them and if they did this would, for all intents and purposes, be considered “an upset” in which a player that was expected to lose to another player of higher ranking actually wins.

For example, if at a major in 2016 Ryo (PGRv1 #45) takes a set from SlayerZ (PGRv1 #29), this would be considered an upset while the reverse would be considered predictable.

With all of these dynamics, metrics, and methods in place, the top 50 was created based solely on objectivity and no subjective values were implemented -- such as the survey that 30 respondents answered.

Closing

In conclusion, the list was an aggregate ordering of players from around the world that competed in the 49 events listed in Figure 2 and were ranked strictly according to placings and set counts against each other.

By these standards, each person above the other in the list should have either placed higher or more consistently at most events than the player after them, have more impressive wins over notable players, or have higher set counts against notable players.

If they were missing one of the components listed above, there was usually an overwhelming focus in another component.

If a player did not fulfill those criteria, then they could not be considered to be higher than they were listed or listed at all as speculation would have to be the only basis by which to judge them and we avoided speculation entirely.

This, in part, explains the inflated or harsh ratings people received on the survey since that was completely subjective.

 

Obstacles:

 

It is through establishing these standards that we faced the following limitations:

  1. Missing data - players that simply did not attend an event
  2. Limited data - players that did not attend enough events
  3. Limited bracket interactions - players did not all face each other this season due to different bracket runs
  4. Inconsistent data - players placing abnormally high/low due to upsets, DQs, or breakthroughs

 

These limitations were met with swift deliberation and the objective metrics that were listed above were the only metrics considered for this study.



Appendix

 

1. Panelists:

 

Bear
Tweek
Larry Lurr
Pugwest
Dabuz
NAKAT
CaptainZack
BBES | TMPR
Zinoto
EMG TorontoJoe
Tetra
SilentDoom
vyQ
Keitaro
EE freaking Visu
VGBC | Tantalus
VoiD
Samsora
Rorec
Austy
ANTi
ESAM
Vayseth
Gunblade
Haruki
Max Ketchum
suar
PG Dom
Zan
juddy

 

2. Tournaments that were considered:

 

  1. CEO 2016
  2. EVO 2016
  3. Super Smash Con 2016
  4. The Big House 6
  5. UGC Smash Open
  6. 2GGT ZeRo Saga
  7. Umebura 23
  8. Sumabato 10
  9. Battle Arena Melbourne 8
  10. LVL UP EXPO 2016
  11. Get On My Level 2016
  12. Momocon 2016
  13. Texas Gaming Championships 8
  14. Midwest Mayhem 3
  15. 2GGT Mexico Saga
  16. KTAR XVIII
  17. Smasn 'N' Splash 2
  18. Sumabato 11
  19. Apex 2016
  20. Low Tier City 4
  21. WTFox 2
  22. Neokan Party 2
  23. Midwest Mayhem 4
  24. Umebura 24
  25. Smash Factor 5
  26. 2GGT KTAR Saga
  27. Sumabato for The Big House
  28. Clutch City Clash
  29. Endgame
  30. Umebura SAT
  31. Sumabato 12
  32. Rio Egames
  33. Shine 2016
  34. Umebura 25
  35. Syndicate
  36. Glitch 2
  37. 2GGT Abadango Saga
  38. Sumabato 13
  39. Sumabato 14
  40. Eclipse 2
  41. Collision XIV
  42. Little Big House 2
  43. Midwest Mayhem 5
  44. Canada Cup
  45. Olympus 2016
  46. KTAR XIX
  47. Smashdown World
  48. Showdown Battle Royale
  49. 2GGT ZeRo Saga

**bolded signify Tier 1 Events



3.Survey used (now retired):

https://goo.gl/forms/KVGGqVbKnhFkb3kx1




 

 

Mashumaro

Mashumaro

David Wu

Your Comments

Please register or login to post comments