Recommendations
Play around with the weights and settings to check out different resultsMap
Selected:
Weights
weighted avg rating from users who like the diff
per descriptor shared with the diff
similar to descriptorMatch but divided by the total descriptor count (aka map with 2/3 descriptors rated higher than map with 2/20 descriptors)
when ranked within settings.proximityMonths months of the seed
per nominator shared with the set
per mapper shared with the diff
how much higher the fans rate the diff vs everyone else
share of the fans vs everyone so big fanbases of the diff get no bump in this
pearson correlation of how similar users rated both diffs
how close the diffs are in star rating
Settings
ratings at/above this count as "positive"
abs(diff rank date - TARGET) <= window
site-wide mean for bayesian avg which is 3 currently, change at ur own risk
n for bayesian avg
similar diffs need at least this bayesian avg rating
min fraction of fans
max fraction of fans
avoid overfiltering cuz of the share settings
n = u need 50% of the cohortLift value
diminish the effect of cohort if the fanbase is n large
exponent setting so 90% fans is more than twice vs 45% fans
similar to bayes avg, correlations are shrunk by n/(n+this)
diffs need at least this many users who rated BOTH maps
ignore candidates correlated below this (0 = anything negatively)
SR diff via fraction, so 0.5 = 50% of the diff's SR as the limit
Results
Pick a map and hit Get Recommendations