crosstheages / public-data

11 stars 3 forks source link

Making Awakening Fair #349

Open DorkHelmet opened 1 year ago

DorkHelmet commented 1 year ago

A guild mate just recently expressed his dismay with awakening, after having spent 70,000 Trisel (that he had been collecting across two accounts and one month of playing.. A LOT of playing), to get ONLY C awakenings (this a top 10 player) and he was so dismayed by this result that he decided to take a short break to try to bring back his motivation to even launch the game again..

This is of course how statistics work when every single event is distinct, every time that you gamble on the awakening there is a 60% chance it will be a C. Of course, over a large enough set of awakenings this will even out, and statistically across ALL awakenings I'm sure the statistics actually do show it to be so, but for the individual, it will not have to be so because the distribution follows a standard Gaussian distribution in the shape of the famous bell curve:

image

Most people will fall in the middle, so 64.2% will get pretty close to the expected results e.g. 60% C, 27.5% B, 10.5% A, and 2% S. BUT, the other 35.8% of people will NOT be in that part, half of those will actually be wondering what I'm blabbing about because they are in the top part and are actually getting better results and are happy, but the other half are nodding vigorously now.

To give an example, there will be a time when some really unlucky VERY patient person will have awakened 1000 cards but NEVER got anything better than a few Bs, and on the other side, there will be people that have awakened 1000 cards but have only gotten like less than 100 Cs and the rest being B+. This is how a bell curve distribution work. Obviously, the person that is very unlucky is unlikely to stick around...

So, how can this be mitigated, and made to feel fairer?

I propose the following:

current numbers as mentioned before: C = 60%, B = 27.5%, A = 10.5%, S = 2%

1 counter value lowers the % of C by 2%, and distributes that 2% evenly by the current ratio across the other grades, the calculation would look like this: Where $P_B, P_A, P_S$ are the percentage to get B, A, S, and $i$ is the counter.

$$P_B= 27.5 + (i * \frac{0.275}{0.4})$$

$$P_A= 10.5 + (i * \frac{0.105}{0.4})$$

$$P_S= 2 + (i * \frac{0.02}{0.4})$$

example %

Grade i = 1 i = 5 i = 10 i = 20 i = 30
C 58% 50% 40% 20% 0%
B 28.875% 34.375% 41.25% 55% 68.75%
A 11.025% 13.125% 15.75% 21% 26.25%
S 2.1% 2.5% 3% 4% 5%

The reason to have a rarity counter per rarity is to prevent gaming of the system by awakening common/uncommon cards in the hope to later awaken a mythic at a much higher %, But also in the reverse e.g. it kind of sucks to "waste" your "awaken bonus" on a common card! vs a mythic, so this way it makes sense to have a counter per rarity.

the above formula can of course be tweaked a lot, this is just a suggestion, as you could add halt positions vs taking it all the way down to 0% chance for a C, e.g. maybe C can not go below 10% or 20% (though this still allows for some exceptionally unlucky people to just luck out), our maybe its 1% adjustment per counter, or maybe its 5%.

Another possible adjustment is to have the counter slide based on rarity, maybe Mythics have a multiplier of 5 instead of 2 and a "common" has a multiplier of 1. There are a lot of possibilities, BUT the fundamental thing is that people can feel that awakening a "C" kind of sucks BUT any future awakening has a higher chance to be something better, and every "failure" actually increases your chance for future awakenings.

Shin7-7 commented 1 year ago

You would be my Savior LordHelmet! I did not do the 70k for only C as our friend, but I get the feeling to much often...

mirumoto1 commented 1 year ago

Your proposal tends to increase the number of B, A, and S cards over time. But it's a good idea