Competitiveness stats after first full GA
After complaining about the matchmaking all last week, I decided to collect and analyze some data on my competition in GA, to try to determine the statistical correlates of GA victory. Immediately after matchmaking, I collected all the data shown in my and my opponents’ public profiles in-game, and then manually counted our numbers of G12s and zetas.
What I found was that, in my group of eight, neither number of G12s (r^2 = .08) nor number of zetas (r^2 = .04) correlated very strongly with number of GA wins.
Instead, the highest correlates of victory were statistics that indicate how long the player has been playing: number of raids (r^2 = .67), number of guild tokens (r^2 = .64), normal battles (r^2 = .47), and donations (r^2 = .44).
The highest correlation that does not appear to be closely associated with player longevity is that between arena rank and GA wins (r^2 = .33).
Based on this, I would suggest that if the devs want to create more competitive groupings, they should consider each player’s longevity and its correlates. Just anecdotally, the player who won our group had a below average number of G12s, above average number of zetas (not the most in the group), but had completed 200 more raids and 1200 more GW battles than the next person. Based on my data, that looks like it may be a hidden advantage in GA, that the matchmaking algorithm could account for.
Obviously this is all preliminary, and based on a small sample size. If people find this helpful, I’ll continue to update this thread as I collect more data from my future GA matchups.
What I found was that, in my group of eight, neither number of G12s (r^2 = .08) nor number of zetas (r^2 = .04) correlated very strongly with number of GA wins.
Instead, the highest correlates of victory were statistics that indicate how long the player has been playing: number of raids (r^2 = .67), number of guild tokens (r^2 = .64), normal battles (r^2 = .47), and donations (r^2 = .44).
The highest correlation that does not appear to be closely associated with player longevity is that between arena rank and GA wins (r^2 = .33).
Based on this, I would suggest that if the devs want to create more competitive groupings, they should consider each player’s longevity and its correlates. Just anecdotally, the player who won our group had a below average number of G12s, above average number of zetas (not the most in the group), but had completed 200 more raids and 1200 more GW battles than the next person. Based on my data, that looks like it may be a hidden advantage in GA, that the matchmaking algorithm could account for.
Obviously this is all preliminary, and based on a small sample size. If people find this helpful, I’ll continue to update this thread as I collect more data from my future GA matchups.
