Ok, so I just did some projections to simulate what would happen in the new rivalry system.
For my experiment, I created 198 guilds (enough for 33 rivalries each week).
I gave each a "starting quest score" (SQS) that represents a
typical # quests that they complete.
I did this in the following distribution:
- 120 guilds got a random SQS between 24-36
- 60 guilds got a random SQS between 36-48
- 18 guilds got a random SQS between 48-60
Don't get too worried about my distribution -- it doesn't affect the results too much.
To begin the simulation, I sorted the guilds by those starting quest scores and paired them into groups of 6 (just like our rivalries), and let that be the first week of results.
For each successive week, I did the following:
(1) I re-sorted all the guilds by the quest scores they just got, and re-grouped them into new rivalries of 6 guilds
(2) I gave each guild a "weekly modifier" (random between -4 and +4) to represent the fact that every guild's performance varies week-to-week
(3) I calculated their score for the new week = SQS + weekly modifier
(4) I then ranked them within their rivalries based on this week's scores
I repeated this exercise for 12 weeks.....
After 6-weeks of the simulation:
- 94% of all guilds had an average rank between 2.0 - 5.0 (i.e. within 1.5 of the 3.5 average)
- 76% of all guilds had an average rank between 2.5 - 4.5 (i.e. within 1.0 of the 3.5 average)
- 48% of all guilds had an average rank between 3.0 - 4.0 (i.e. within 0.5 of the 3.5 average)
After 12-weeks of the simulation:
- 99% of all guilds had an average rank between 2.0 - 5.0 (i.e. within 1.5 of the 3.5 average)
- 91% of all guilds had an average rank between 2.5 - 4.5 (i.e. within 1.0 of the 3.5 average)
- 60% of all guilds had an average rank between 3.0 - 4.0 (i.e. within 0.5 of the 3.5 average)
(GRAPHS BELOW)
Conclusions:
- As the weeks roll by, every guild that routinely performs within reasonable range of their SQS will have their average rivalry rank converge (as I predicted) towards 3.5th place
- That means, the
only way to have an average rank
better than 3.5 is to have
continuously increasing quest scores! Because once you plateau off, you'll immediately trend toward 3.5 again. So maybe you can "beat the system" for a week or two or maybe even three, but eventually your guild will hit their effort-wall, and you'll start getting 3.5th place just like the rest of us.
- Whatever short term gain you get from pushing a few extra quests to win a rivalry, will eventually be counterbalanced in the long term by extra rivalry losses.
-
EFFORT (towards rivalries)
HAS LOST ALL MEANING. Honestly, we might as well keep the guild quests, but scrap the rivalries altogether, and just give everyone a flat 13,500 tokens each week (to represent their average winnings for being 3.5th place every week).......
The old system had the flaw where people with strong ships/gear were tanking their infamy to generate easier rivalry match-ups.
However, by implementing this new system, you just threw the baby out with the bathwater....... :-(
View attachment 10131
A notes about my experiment:
The data actually would have been
even slightly tighter if I were to have randomly selected tie-breakers (rather than realize too late that my data fell victim to Excels tendency to preserve sort-order in the cases where the same two guilds tied two weeks in a row).
Click to expand...