• 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!

Crafting Recipes

Deleted User - 1759805

Guest
Dony doesn't guess about such things. It's either viewable from the game files or he got it from an employee.
I'd sooner bank on his information than a support ticket or a post from a volunteer forum moderator.
Not once in 3 years has he shared bad info.
I don't think the actual current WW percentages are in question - I believe many people verified these as long-running averages (at least for diamonds). The point is, unlike event chests or mystery object these are not officially stated. This means that these can be adjusted at any time without notifying anyone, as no one ever promised that you're going to get, say, 10% chance of diamonds from the WW.

Now they probably can't do the same with the mystery object without making changes public (at least by changing the description) - and hopefully some warning ahead of time.

It's like relying on undocumented/unsupported features or API. Sure, it works - until it doesn't, and this can happen suddenly and without warning. Many a tear has been shed this way ;)
 

m4rt1n

Adept
The wishing well comparison also shows another example of differing odds across the field. Just like event chests offer differing odds, so in my opinion there is no reason why the Crafting Recipes would be all the same odds or different odds, either way for appearing.
@Muf-Muf @Dizzy Lizzie @Rilian an official answer would be helpful please. :)
 

Deleted User - 341074

Guest
This means that these can be adjusted at any time without notifying anyone, as no one ever promised that you're going to get, say, 10% chance of diamonds from the WW.
There was even an announcement when it got changed from 10 to 25 diamonds.
But I'm pretty sure(without bothering to check) that the EULA has wording to the effect that "inno can change stuff whenever."
spire2.0.png
 

Deleted User - 1634960

Guest
@m4rt1n @SoggyShorts and all others.
Soggy has just posted a graphic quoting Marindor on this issue. As a disagreement between Marindor and Muf-Muf is a truly ugly sight, I think we can say that this is probably true, although I would have trusted Dony on it as well (Glad I saw that before I posted what I had been working on).;)

For those still doubting, with 100 Recipes, 5 per swap and 21 swaps a week, It would only take about 4 months of effort, meticulously recording every recipe that showed, to reach a statisticly valid sample. Any of you are welcome to go for it. :D
 

Deleted User - 1759805

Guest
There was even an announcement when it got changed from 10 to 25 diamonds.
Number of diamonds received is directly observable, and those changes can be immediately noticed. Probabilities... not so much ;)

But I'm pretty sure(without bothering to check) that the EULA has wording to the effect that "inno can change stuff whenever."
Most likely. But in my book there is a big difference between changes that are pre-announced, or at least immediately observable (e.g. mystery object percentages, WW number of diamonds received) vs the ones that are not directly observable, but still apply to me (i.e. all unlisted probabilities).

And thanks for the screenshot. It is good to know that at least at that point in time all recipes were equally likely. The point about future changes still stands, but if Inno didn't implement tuning mechanism for those probabilities from the beginning there is less chance that it will happen suddenly.
 

Deleted User - 1759805

Guest
For those still doubting, with 100 Recipes, 5 per swap and 21 swaps a week, It would only take about 4 months of effort, meticulously recording every recipe that showed, to reach a statisticly valid sample. Any of you are welcome to go for it. :D

Depends what you want to check - you might be greatly overestimating the effort required ;) I only care about probability of a single recipe - Unleashed Unit Upgrade. Those are very easy to track. You only use UUU at most once a week (and I certainly keep records of that), so with that you only need to know your starting UUU inventory and final UUU inventory. Still takes time of course, but that's it, everyone can do it without too much effort.

You obviously need to make sure that you don't miss any MA rolls, and craft every single UUU that pops up.
 

DeletedUser5532

Guest
I sometimes think that the % chance if winning something is mixed up with the probability of winning something. If there were 100 different recipes, and you managed to check all 3 that you get each day, then the probabilty of seeing a recipe in any given week is approximately 20%. So at the start there is a probability that every recipe will be seen once in every 5 week period. But as the probability will change on each new set, the % chance never changes.
The % chance of seeing the exact same set of five a second time is statistically very low, but no matter how many times it were to happen, the % chance would never change, being the exact same each time. The probability of this happening would decrease massively each time, but the game does not adhere to probability, only on the % chance.
 

Deleted User - 1759805

Guest
I sometimes think that the % chance if winning something is mixed up with the probability of winning something. If there were 100 different recipes, and you managed to check all 3 that you get each day, then the probabilty of seeing a recipe in any given week is approximately 20%. So at the start there is a probability that every recipe will be seen once in every 5 week period. But as the probability will change on each new set, the % chance never changes.
The % chance of seeing the exact same set of five a second time is statistically very low, but no matter how many times it were to happen, the % chance would never change, being the exact same each time. The probability of this happening would decrease massively each time, but the game does not adhere to probability, only on the % chance.
I am not sure how did you derive your numbers. There are 5 recipes shown every 8 hours - so we see changes 3 times a day. This means you can see 5*3*7=105 recipes per week. Let's round down and say 100 recipes per week, so it's easier to calculate.

If we assume that there are exactly 100 possible recipes out there, and they're picked randomly with equal probability, then it is quite obvious that you would expect to see each recipe about once per week - on average. Not every 5 week period.

Now, probability of seeing any particular recipe in a weeks time is different. It is complementary to the event of not seeing a particular recipe for a week. Which incidentally means that in 100 draws, we've got one of the other 99 recipes every singe time. Probability of that is 0.99^100 = about 37%. This means that probability of seeing any particular recipe in a week is about 63% (not 20%).
 

DeletedUser5532

Guest
If we assume that there are exactly 100 possible recipes out there, and they're picked randomly with equal probability, then it is quite obvious that you would expect to see each recipe about once per week - on average. Not every 5 week period.

I clearly had brain fog when i was writing this, yes, the probability is that they could all be seen in a week, not 5 as i said.


Now, probability of seeing any particular recipe in a weeks time is different. It is complementary to the event of not seeing a particular recipe for a week. Which incidentally means that in 100 draws, we've got one of the other 99 recipes every singe time. Probability of that is 0.99^100 = about 37%. This means that probability of seeing any particular recipe in a week is about 63% (not 20%).

The probability of seeing new recipes grows from the 1st 5 that we see, which means the probability changes on every single change of recipes, it is never a fixed number, which was the reason I entered my thoughts onto the thread, the % chance of seeing different recipes on each new change is 5%, and the % chance of seeing the exact same 5 is also 5%.
 

Deleted User - 1759805

Guest
The probability of seeing new recipes grows from the 1st 5 that we see, which means the probability changes on every single change of recipes, it is never a fixed number, which was the reason I entered my thoughts onto the thread, the % chance of seeing different recipes on each new change is 5%, and the % chance of seeing the exact same 5 is also 5%.
I am not sure what you mean. If what Inno is saying about equal probabilities is true, the chance of rolling any single recipe is independent of what was rolled previously. Even if you roll 5x UUU in a row, it has no bearing on what will come next. The probabilities stay the same. The outcomes will not.

Still not sure where 5% is coming from. Do you really think that probability of rolling the same 5 recipes in the next 8 hours is 5%? (HInt: it is significantly less. Like, a lot less).
 

Deleted User - 1634960

Guest
There are no small numbers in this.
The chances of any particular 5 Recipes appearing can be calculated.
I am using the following assumptions:
  1. There are 100 Recipes
  2. Recipes will not be exactly duplicated in the set.
  3. The chances of each Recipe are equal.
The odds of one of your 5 Recipes coming out on the first draw are 5:100, or 1 in 20
The odds of the second draw being one of your 5 ar 4:99 or 1 in 25
The odds for the 3rd draw are 3:98 or 1 in 33
The odds for the 4th are 2:97 or 1 in 49
The odds for the 5th are 1 in 96

These are multiplicative, so the final odds for your particular 5 are 20x25x33x49x96 or 1:77,616,000
It is numbers like these that make lotteries so profitable for the organisations running them ;)
 

DeletedUser5532

Guest
I am not sure what you mean. If what Inno is saying about equal probabilities is true, the chance of rolling any single recipe is independent of what was rolled previously. Even if you roll 5x UUU in a row, it has no bearing on what will come next. The probabilities stay the same. The outcomes will not.

Still not sure where 5% is coming from. Do you really think that probability of rolling the same 5 recipes in the next 8 hours is 5%? (HInt: it is significantly less. Like, a lot less).

This is where the confusion comes from, as the probability is very different from the % chance.

To answer your question, the % chance of getting the same 5 is 5%, the probability is significantly lower.

To give this another context:

If you were to have 100 different balls in a bag, and took one out, the % chance is 1% that you would get ball number 1. The probability at this point is the same, 1%.

Now imagine that you do this 10 times, and each time you get the same ball. When you go to try for the 11th time, the % chance would still be 1%, this cannot change. However, the probability of this happening is extremely small, and every time that number 1 ball comes out, the probability of it happening again gets even smaller.

Rilians numbers are correct, but you need to change the word odds to probability.
 

Deleted User - 1759805

Guest
Now imagine that you do this 10 times, and each time you get the same ball. When you go to try for the 11th time, the % chance would still be 1%, this cannot change. However, the probability of this happening is extremely small, and every time that number 1 ball comes out, the probability of it happening again gets even smaller.

No it doesn't - you said so yourself, correctly. Probability has a well-defined mathematical meaning, while "% chance" is a layman's term for probability, it does not define something new.

You're mixing probabilities of different events. Probability of pulling ball #1 out of the 100 in a single draw is indeed 0.01, and because draws are independent from each other (a priori assumption) it does not matter if we pulled ball #1 10 times before - or even 100 times. These are realized outcomes and have no bearing on future draws.

Now probability of a composite event where we pull ball #1 2x times in a row is significantly less - 0.01^2, 10 times in a row is 0.01^10 etc - so very small. But if one of these event would to happen, probability of the next ball being #1 again (conditional probability) is still 0.01 - because observations are independent.
 

DeletedUser5532

Guest
I guess we have to agree to differ on this one. It is quite a few years since I got my BEng degree in Mechanical Engineering, and they may well have changed the terminology, but I would rather stick with what I was taught back in 1995. Gee, that makes me feel soo old.
 

Deleted User - 341074

Guest
As a former casino manager I can confirm that if the ball lands on Red 23 twice in a row that the probability of it landing on 23 next spin is still 1:38
The probability of it landing on red 23 three times in a row is much smaller, but we don't need that--it already landed twice.

Think of flipping a coin
If you got heads 100 times in a row, what is the probability of it coming up heads next toss? Still 50:50 because we don't need 101 heads, just 1.

It's all about where you are in time.
If you are standing a the table AFTER 2 red 23 spins, the chances of red 23 next spin is 1:38.
If you just walked up and there have been no spins yet and you want to know the chances of 3 red 23s in the next 3 spins the chances are less than 1 in 50K
 

davemc

Bard
This is probably i bit of a simpler version as maths is no longer my strong point. I used to manage a bingo hall and we forever got complaints saying the numbers were fixed. But during a game a number comes out at the start there is a 1:90 chance of that number. But the number does not go back into the pot so the next number the chance reduces to 1:89. Now do the crafting reciepies go back into the pot? if so the chance of that will never reduce. It then comes down to probability that you get the one you one but the odds will remain the same
 

DeletedUser3097

Guest
A very interesting thread- one factor which hasn't been taken into account is this. Some of the recipes are one-off unique recipes so once used will not reoccur (unless the player destroys the original piece). An example of this is the Chessboard. This will skew the figures given above.
 

LazyTony

Sorcerer
@Dizzy Lizzie Are there any other recipes that aren't offered when the item is in our city or inventory? AFAIK the chessboard pieces are the only ones, would be great if there were some others I could remove from the pool to increase my chances of stuff I want.
 

Pauly7

Magus
I've never been much interested in that chess set, but I suppose there could be an argument for building it anyway temporarily, when found with a bit of free space, just to increase the odds of other things appearing.

Incidentally I'm still not at all convinced about the percentage chance being even between them all. All we have to go on is one comment from Marindor once upon a time, but if the odds were the same I think the devs would be a bit more forthcoming with that information.

I know all too well that unusual looking runs of bad luck or good luck can appear in any such situation, but I have never seen a run of good luck when it comes to unit upgrade buildings. I don't think I've seen any one of the three of them for over two weeks now and there have been pet foods all over the place. If that's bad luck then OK, but if so I wouldn't expect the bad luck to just continue ad infinitum. Perhaps I should keep a record for a long period as @Rilian suggests, but I'm not sure I have that inclination. Also, I don't think there would be anything wrong with some things being more likely than others, but I also don't know why they wouldn't just say so like with event chests.
 

m4rt1n

Adept
I've never been much interested in that chess set, but I suppose there could be an argument for building it anyway temporarily, when found with a bit of free space, just to increase the odds of other things appearing.

Incidentally I'm still not at all convinced about the percentage chance being even between them all. All we have to go on is one comment from Marindor once upon a time, but if the odds were the same I think the devs would be a bit more forthcoming with that information.

I know all too well that unusual looking runs of bad luck or good luck can appear in any such situation, but I have never seen a run of good luck when it comes to unit upgrade buildings. I don't think I've seen any one of the three of them for over two weeks now and there have been pet foods all over the place. If that's bad luck then OK, but if so I wouldn't expect the bad luck to just continue ad infinitum. Perhaps I should keep a record for a long period as @Rilian suggests, but I'm not sure I have that inclination. Also, I don't think there would be anything wrong with some things being more likely than others, but I also don't know why they wouldn't just say so like with event chests.

You don't need to build it to take it off the menu, just store it in the inventory.
 
Top