Deleted User - 341074
Guest
2 accounts zero hits on this as well for me. Better than average on the 82s more than made up for it though.
-
Good day, Stranger! — Are you new to our forums?
Have I seen you here before? To participate in or to create forum discussions, you will need your own forum account. Register your account here!
Deleted User - 1759805
Guest
Out of curiosity, why so many Cults? Is it for disenchanting, or are you actually planning to use them?
DeletedUser5093
Guest
That's a pretty bold claim.
I still maintain that my method is mathematically valid. At this point I think we'll just have to agree to disagree. Deleted User - 1759805
Guest
No, it's not valid. The good thing about science is that you can prove (or disprove) claims. You can try to make a simulation yourself, and see if it can confirm your approach. You won't see the numbers that you predict, starting with expected values. We just can't use SE per daily prize in this way. Deep down, it boils down to the fact that inverse of the sum is not the same as sum of inverses.That's a pretty bold claim.
But here is a simple illustration why the SE per daily prize approach doesn't work. Lets assume we only have 1 chest per selection (instead of 3), and only 2 different chests that pop up equally likely. Both chests cost the same to open, let's say 100, but chest A has probability of success at 100%, but chest B has probability of success of 1%. So obviously chest A has SE per daily prize of 100, but chest B has SE per daily prize of 10,000. You can probably see where I am going with this example
So your spreadsheet would only have 2 scenarios, and they will look like this:
1: chest A - cost per daily prize = 100
2: chest B - cost per daily prize = 10,000
So far, so good, but now you average those numbers across scenarios, and assume that this would be average cost per daily prize for the whole simulation. It obviously isn't. (100+10,000)/2 = 5050. Do you really think that in this setup it takes 5050 SE on average to hit success? The real number is slightly less than 200, and it's easy to see why. Every 100 SE that we spend opening, we either hit chest A and get the prize, or we hit chest B and get nothing, most likely. So on average, we get the prize every 200 SE. That's quite a difference from 5050. And I can make this number arbitrarily large by simply making probability of chest B arbitrarily small. The real average cost per daily will only move closer to 200.
To step back from math, the problem with averaging costs per daily prize is that we effectively assume that once we hit a particular chest, we keep hammering at it until we hit success. Which is a problem with bad chests as it can take a long time (see above). In reality, you only spend the cost of the chest, and then move into a different chest which likely is going to be a better option.
m4rt1n
Adept
I notice some unfairness with the quests, especially the latest ones requiring lots of T1 and the scout 2 provinces.
Elves T1 Steel and Humans T1 Steel and Planks boosted have to use 100% more space to fit in the extra buildings that are absolutely necessary for anyone now with sentient goods buildings as they don't produce 1 and 2 day productions.
Scout 2 provinces for me in ch14 is going to take about 6 days to complete and I see no alternative to the quest given either.
Elves T1 Steel and Humans T1 Steel and Planks boosted have to use 100% more space to fit in the extra buildings that are absolutely necessary for anyone now with sentient goods buildings as they don't produce 1 and 2 day productions.
Scout 2 provinces for me in ch14 is going to take about 6 days to complete and I see no alternative to the quest given either.
LazyTony
Sorcerer
Elves steel lev 1 is 2x2. Its planks and marble that are 1x2 for elves.
There are a few imbalances between the races though. I would love to start again with a human city so I could have human mages, easily the best unit in the game (tourneys are very much my thing). Also, humans have a smaller builders hut than elves.
My 2 scouts will take 7.5 days
There are a few imbalances between the races though. I would love to start again with a human city so I could have human mages, easily the best unit in the game (tourneys are very much my thing). Also, humans have a smaller builders hut than elves.
My 2 scouts will take 7.5 days
DeletedUser7309
Guest
To focus on this event I am neglecting other aspects like research; I have no more ancient wonders to build (even ones I don’t want) but am still producing rune shards.
When I unlock a new AW will my collected shards magically appear if have they gone into the abyss never to be seen again ???
I have accumulated 2k SE ready to bid for the wishing well daily prize tomorrow but after I might quit the event
When I unlock a new AW will my collected shards magically appear if have they gone into the abyss never to be seen again ???
I have accumulated 2k SE ready to bid for the wishing well daily prize tomorrow but after I might quit the event
Last edited by a moderator:
Sir Derf
Adept
I'm already, still, sitting on 13 Elven and 4 Orcish Floats gotten for disenchantment purposes.
I am strongly considering using at least some of them, but yes, they are also disenchantent fodder.
Killiak
Artisan
Rune Shards are specific to each AW, and will be stocked up indefinitely, even if you do not have the AW unlocked quite yet.
Broken shards, however, have a maximum of 10 (ten broken shards can be used to make one rune shard, which is then auto-inserted).
Any above said limit are gone with the wind.
DeletedUser7309
Guest
Not what I wanted to hear but thanks
Timneh
Artisan
Not 100% correct as the 10 broken shards are not auto-inserted, You have to click the forge shard button and then it gets insertedThis might be being a bit picky but i just want to avoid any danger of some people getting confused if they get 10 broken shards and a rune shard does not get inserted automatically.
Timneh
Artisan
The wishing well comes again near the end of the event, day 26 i think.
Deleted User - 341074
Guest
Yup, I would have had no chance on Beta if I wasn't an elven planker.
Make it fair:
"Produce 5 Giant Globes OR 10 Admirable Altars"
Remove the boost condition, let players build whatever flavor of turds they want.
Deleted User - 1759805
Guest
Still won't be quite "fair". Don't know about others, but I use my real manufactories in quest productions. I only run 3 T1, but some people run 5-6 or more. That's a big help. Unless you force building non-boosted for quests - in which case Human marble will be screwed.
Deleted User - 341074
Guest
true, but closer than now.
Paladestar
Enchanter
Always interesting comparing the RNG when doing this event with multiple cities. About 6350 essence spent on both cities (all sequential quests done, all daily quests done, no diamond-bought essence and no Stash Outpost):
City 1: 21 wishing wells, 7 artifacts. 42 feathers to go until phoenix is lvl 10 (not looking so good). No essence left.
City 2: 30 wishing wells, 8 artifacts. 11 feathers to go until phoenix is lvl 10 (easy mode). 56 essence left.
City 1: 21 wishing wells, 7 artifacts. 42 feathers to go until phoenix is lvl 10 (not looking so good
). No essence left.City 2: 30 wishing wells, 8 artifacts. 11 feathers to go until phoenix is lvl 10 (easy mode
). 56 essence left.
m4rt1n
Adept
I have been here a while and still do not understand everyone's obsession with the RNG system Inno uses.
Its easy, it works in the favour of Inno, as it is their RNG, not one you use in the calculations.
From years of experience Inno's RNG gets a little "sticky" also giving the same prize multiple times in a row, well against random odds.
That can mean loads of prizes in a row at 5% each, or loads of nothing at 30% odds. Inno's RNG in my opinion has more factors included than the number crunchers here factor in.
Its easy, it works in the favour of Inno, as it is their RNG, not one you use in the calculations.
From years of experience Inno's RNG gets a little "sticky" also giving the same prize multiple times in a row, well against random odds.
That can mean loads of prizes in a row at 5% each, or loads of nothing at 30% odds. Inno's RNG in my opinion has more factors included than the number crunchers here factor in.
Deleted User - 341074
Guest
But why? There's no gain for inno to do that. The posted odds work well enough. It's like at the casino I used to hear players insist that machines or even dealers were cheating when they obviously weren't because there's no point. The posted odds are enough for the casino to make money, just like the posted odds in Elvenar are enough to give the desired number of prizes.
DeletedUser8573
Guest
Used about 5.5k essences today.
- I got 1000+ extra essence from chests (at least 5x200)
- 32 Wishing Wells (30 from daily rewards)
- 7 Phoenix upgrade things
DeletedUser5093
Guest
Sometimes if a program requires a random number it's easier to use pseudo-random numbers instead, as this can save on computing power and other similar resources. If anything the benefit would lie there, it wouldn't be about improving odds in favour of Inno. If the pseudo-random number generator is well designed then there's no real difference for most applications, and both methods work equally well. If it's poorly designed then you can occasionally see quirks in the randomness, such as some of the issues people are reporting here. I would caution against leaping to this conclusion though. The plural of anecdote is not data. We simply don't have enough information to determine if these issues are a result of bad code, or simply a fluke. It's not just about whether or not these implausible scenarios occur, but also the frequency with which they occur, and there's no way to know that without also knowing about all the times they don't occur.
If people were really keen to test the randomness of the chest and prize generation I could set up a thread to collect data, with some basic instructions of what information people need to record (and how to do it so as to ensure a true representation of events, not just the occasions when things are really great or really bad). If enough people post their observations from their own cities then the various number crunchers on these forums can use this data set to run their own analyses and see just how random things really are.
The data collection form might look something like this:
Initial SE: [the amount of SE you have stockpiled before you buy the first chest in this test]
Round 1:
Chest options: [chest1], [chest2], [chest3].
Chest chosen: [one of the above chests].
Prize won: [item].
Round 2:
[etc]
Final SE: [the amount of SE left after all the above rounds - mostly just a way to check the above data has been recorded correctly]
Collecting this data would tell us how frequently each chest appears, and how frequently the different prizes are given from each chest.
Would there be any other information that would be useful to record? If not necessarily for this test, then for other ones?
Note that no one should feel the need to record mass quantities of data to do this right. Although that will undoubtedly be useful, lots of people taking small samples will be just as useful, as it all adds up.
So would people actually be keen to give this a go?