Feedback New relic drop table?

[Achaeus]

Kind of on topic.... Has the drop rate for relics dropped? I normally see much more than I have since the change.
I know my ToR is low level but I definitely noticed a reduced number of relics.
k.thx.

Oldtyme72 it varies from week to week, don't get discouraged! I am not sure what time frame the chance of getting a relic is based on, but I notice that some weeks I get very few and other weeks I get quite a few. Last week I had very few drops and this week it seems I can't move without getting one, lol. So hang in there, you will get relics again.

 

[podkap1970]

Jade Relic: Reno Kit

 

[Andi47]

Jade Relic: Reno Kit

Thanks, is added to my list

 

[DeletedUser5429]

ToR level 62, and very few relics this week.

 

[DeletedUser730]

Hi, I got one yesterday.
Same thing: Level 4, gold relic, 100 FPs.
The 100 FPs reward was not said to be removed, only its probability was lowered.

Got one today in lvl 2. Definitely not removed.

 

[Andi47]

Updated table for Jade Relic (+ Sacred Sky Watch and Storage Kit)

 

[piohh]

I get only 1 gold Relict for the first 3 Rounds this week. My ToR is Level 8

 

[podkap1970]

i got 2 gold relics in 4th level, both had medals inside .....

 

[Andi47]

updated the start posting.

 

[qaccy]

I agree that it'd be nice to hear some actual values for the odds of each relic prize, however with such a small sample size you aren't personally indicating anything other than that you're probably a bit 'unlucky' with getting farms. Now, if you were still at 0 farms after collecting 200 jade relics, you might be able to start claiming something but 17 is inconclusive to say the least.

Edit: NormaJeane - 2017-09-22 - removed reference to deleted post.​

 

[Manganite]

Based on some data one can do a so called maximum likelihood estimate of the drop rate within a confidence interval of 95%. For 157 relics in 9*64 attempts this gives an interval of 23.7% - 31.1%, which is just 0.4% below the supposed drop rate of 31.5%.

A confidence interval of 97% gives an upper limit of the expected 31.5%.

One can do this also the other way round. If you would have won the expected 181 relics, the drop rate would be with 97% in the interval 27.26% - 35.81%.

Therefor, the data are statistically absolutely compatible with the supposed drop rate of 31.5% and not a contradiction in any way.

Edit: NormaJeane - 2017-09-22 - removed reference(s) to deleted post.​

 

[Manganite]

There´s nothing wrong with the system, it´s the nature of randomness that the exact expectations values will only be reached for a very large numbers of trials. If you would wait a few months you would see that your observed expectation values will narrow down to the theoretical ones.

It is also due the nature of randomness that you observe some values far below, some above and some very close to the theoretical ones.

It is exactly the opposite, if all values for all players would be always very close to or even exactly on the expected values, then something would be badly going wrong and the whole system would be based on cheating...

 

[Manganite]

Of course I looked at your stats. But you have to draw the correct conclusion from them.

The whole mechanism of the Temple is based on two completely independent random decisions. In the first step, the decision is made if you get any relics at all with a certain probability. If the decision is positive, the next decision is made, it is which type of relic you get.

That means you have to compare the drop rate for the specific relics to the number of relics you actually got not to the number of total relics you expected to get! Therefore you have to set the 52 gold relics in relation to the 157 relics you got in total. That gives you a drop rate for gold relics of 33.6% which is around 4% higher than the expected drop rate 29.3%. For silver relics the difference is almost -7% and for the jade relics roughly +3%.

You see, the differences are always a few percent and sometimes plus, and sometimes minus. So randomness works fine in your case...

 

[DeletedUser]

The drop rate of relics seems less than it was on my end.
Hard to say though, considering odds and all, if this is a pattern I should give any mind.

 

[Manganite]

The system is based on randomness. It generates a random number between 0 and 1 and compares it with the drop rates of your temple. The whole thing will look like this:

if rand()<%relic

r = rand();
if 0<=r<%silver

you get a silver relic;​

elseif %silver<r<%silver+%gold

you get a gold relic;​

else

you get a jade relic;​

end​

end

That's it...

Edit: NormaJeane - 2017-09-22 - removed reference(s) to deleted post.​

 

[Manganite]

Just for fun I made Monte Carlo simulation based on the data posted above. I assumed 100000 players doing 64 trials for 9 weeks. The temple has relic drop rate 31.5% and 54.11% for silver, 29.3% for gold and 16.59% for jade relics.

As one can see this gives nice symemtric distributions around the expected values. But of course, they are quite broad. For the total number of relics, expectation value 181, the lowest value is 134 relics, the highest is 233. So people can get up to 50 relics more or less than they would expect.

The value of 157 total relics that was claimed above to be almost impossible to get, or only if the 'system' is broken, is reached by 322 out of the 100000 players. 1423 players win 157 or even less relics. That's 1.42% of all players. That seems to be quite a low number, but on the other hand, only 3547, or 3.5% of all players got exactly the expected number of relics. That's just a bit more than twice the number of 'looser'...

The results are similar, if one has a look at every single type of relics. Maybe most interesting for the jade relics. One would expect 30 relics per player, but minimum is 8, maximum is 58! That's a difference of 50...

 

[RedRed]

Really interesting thread. BTW, just to give some order of magnitude, I'm monitoring the drop rates of diamonds from wishing wells and I've had to collect 14.000 of them before reaching the 1% of average drop rate, therefore I'd don't expect to be in a small range near the expected values with a sample of some hundreds cases

Spoiler

 

[Manganite]

Really interesting thread. BTW, just to give some order of magnitude, I'm monitoring the drop rates of diamonds from wishing wells and I've had to collect 14.000 of them before reaching the 1% of average drop rate, therefore I'd don't expect to be in a small range near the expected values with a sample of some hundreds cases

That's exactly the point. People tend to draw conclusions from their own numbers, which are naturally to small and not representative. They vastly overestimate how fast games based on randomness converge to the expected values and then they think the 'system' is somehow wrong.

I did another simulation to show this. 10000 players, 64 trials per week for one year with a relic drop rate of 31.5% which means an average of about 20-21 relics per week.

As you can see, after one week players will get everything between 8 and 35 relics. The mean values after 10 weeks still show a variation of +/- 4 relics (which means a deviation from the expectation value of +/- 20%!). And even after one year the maximum variation is +/- 2 relics, which means a deviation of +/- 10%! Or in total numbers, the expectation value is about 1000 relics, but some people will only get 900, while someone else will get 1100...

The reason is simply the time behavior of such statistical processes, which approach their expectation value with an exponential decay. That means, in the beginning it converges quite fast but in the end it took for ages until everyone approaches the expectation value....

 

[Manganite]

First, if you would significantly raise the chance to win a relic at every level nothing really would change, still people with a much lower level temple would maybe get more relics than you or someone else. Second, there is a high risc to break the balance in the game. Let´s assume you rause the chance by 0.5 % at each level, you would end up with a 60-70% chance of winning a relic. So basically at two out of three encounters in the GE. I think that is far too much...

Regarding the 9 weeks, of course, for you it is long time of your game life, but it is nothing to draw any serious conclusions out of your data. Especially to claim, that the whole system is broken. And it doesn´t help that others are complaining about something else. They also don´t get the point when they draw general conclusions from their individual unlucky experiences.

 

[thephantom]

Doesn't the fact that you are getting a bigger amount on one world, but a lower amount on another world, already prove that it is random?

Probability is very complicated, but I think if one just looks at the case of a coin toss it's not too difficult to see how it works. Since a coin has 2 sides, we say that the probability of getting either side is 50%-50%. Does that mean that after 10, 100, 1000, 10000, or 100000 coin tosses you will have actual 50%-50% winning of both? Nope.

In the long run, after a huge amount of attempts, the number will lean toward 50%-50%. It is possible that you will not actually have 50%-50%, that is just the expectation we calculate and the average is around there, but you don't know if your winning numbers are at the average, below or above it. Just as manganite's data shows, it is possible for one person to be below the average, as well as people to be above the average. That by itself does not show any error, it's kind of what is expected.