Gambler's Fallacy... new change nerfs GM n Cav crystals
HeroUniverse45
Member Posts: 16 ★
The odds don't stack... while you've introduced a small chance for a Nexus crystal, you've decreased the chance for a normal 5* pull. The 2 do NOT stack, they are independent of eachother. O.8 and 0.2 do not equal 1... your highest chance has been reduced to 0.8.
29
Comments
Wouldn’t the 5* Nexus give you a choice of 3 different 5*, so the odds between the non-nexus 5* and the Nexus 5* added together still adds up to what it was before (not less).
A better crystal would have been a 6* that stays at 1% WHEREIN there's a small chance for a Nexus smh
If you can’t understand how the chances of pulling a 6* from a Cav haven’t changed, you should probably get back to school.
I should report you for harassment in re: to the personal attack.
When these changes come into effect, you will still have a 1% chance to pull a 6*. Only now 20% of your 1% is a chance at a 6* Nexus.
Honestly, it really isn’t that difficult to grasp.
Whatever makes you happy.
Those odds are being described as 1% chance for 6*, with 0.2% chance of Nexus.
But the chance to get a 6* at all is unaltered.
I actually asked for clarification on how the implementation would work, rather than guess. Kabam confirmed that how I'm describing it now is how it is intended to work in-game. In fact, I was asked to bold the above for emphasis.
Since 1.0 - 0.2 = 0.8, this is exactly the same odds listed in the announcement.
In GMC the game first decides between 3*/4*/5*
If 5* was decided, now there's a small chance that you don't get a 5* champ but 5* Nexus crystal instead.
There are lots of different ways to implement reward tables like this, but a common one is through weights and pigeonhole partitions. Algorithmically, it looks like this. You construct a reward table with the reward entries, and a weight for each. 3* would have a weight of 50, 4* would have a weight of 38, 5* would have a weight of 11, and 6* would have a weight of 1. Then the computer would roll ONE number, and scale it to a pigeonhole. Which is to say, we add up the weights (which in this case is 100) and we then roll a random floating point number between zero and one. Then we compare the result to 50/100 (the first weight divided by the total weight). If it is less than or equal then that's what you get. If not we compare to (50+38)/100. If it is less than that, then that's your reward (it wasn't less than 50/100, so this means the random number is between 50/100 and 88/100). If not, we keep going, comparing to 99/100, and then finally 100/100 (which of course has to match if everything else fails).
In this case, the 6* weight of 1 is changed to 0.8, and a 6* Nexus weight was added with a weight of 0.2 The same ONE random float is rolled, and the same pigeonhole algorithm is run, and now you still have a one out of one hundred chance to get some kind of 6*, but now there's a chance for 6* basic and a chance for 6* Nexus.
Whatever you're thinking about "variables" you are just confusing yourself needlessly. Either you have a mental image of how the random reward table works that is wrong, or your notions of how probability work is wrong, or both.
There is still a 100% chance you get a champion from that crystal. It’s the same thing as the featured crystal odds; first the game determines the rarity it rewards, then it determines if it will award the “weighted” champ or pick randomly from the pool.
To put it in simpler terms, there’s a 99% chance you won’t get a 6*, but a 1% chance you will.
There is also a 20% chance that 6* you pull will be a nexus, and an 80% it won’t be.
Featured
3* 0%
4* 60%
5* 35%
6* 3%
6* nexus 2%
Standard Cav
3* 45%
4* 32%
5* 20%
6* 2%
6* nexus 1%
Kabams changes:
3* 50%
4* 38%
5* 11%
6* .8%
6* nexus .2%
Real life reality drop rates on standard & featured Cavs
3* 70%
4* 22%
5* 7.5%
6* .5%
6* nexus .000000000001%