affiliate_tree.txt

affiliate
  + referrals
  |  + referral
  |  |  + player
  |  |  |  + rake
  |  |  |  + rake
  |  |  + player
  |  |     + rake
  |  + referral
  |  |  + player
  |  |  |  + rake
  |  + referral
  |     + player
  |        + rake
  |        + rake
  |        + rake
  |  
  + payscale
     + tier
     |  +min   # 0
     |  +max   # 1000
     |  +rate  # 3%
     + tier
     |  +min   # 1000
     |  +max   # 2500
     |  +rate  # 4%
     + tier
        +min   # 2500
        +max   # null
        +rate  # 5%


# current way of doing things
1. When stats are uploaded, we check for affiliate exception rates
   If no exception rate is found, we use 0%

user    player    referred_by     rake  referral_rate   commission
A       X         foo             $50   1%              $0.50           # exception found for X-type player
A       Y         foo             $100  0%              $0.00
A       Z         foo             $0    0%              $0.00


2. After exceptions are accounted for, we want to use the payscale for the remaining 0% rates
   Sum rakes that have a 0% and check the payscale to see what rate they should get
   
  $50 @ 1% gets rolled into sum for tier determination, but stays locked at 1%
  $100 @ 0% gets rolled into sum
  $0 @ 0% gets rolled into sum
---------------------------------
  $150 = total rake
  

If the payscale is:
  0-100  => 3%
100-250  => 4%
    250+ => 5% 
   
Then a sum of $150 (50 + 100 + 0) referred rake would get 4%.

3. Update the 0% rates to 4% and calculate the commission column. Final table looks like

user    player    referred_by     rake  referral_rate   commission
A       X         foo             $50   1%              $0.50
A       Y         foo             $100  4%              $4.00
A       Z         foo             $0    4%              $0.00