Coin Distribution

[Wildfire2099]

I was wondering what the ratio was for how many coins you get per drop... so I decided to do a bit of a test.....


1 Coins: 25 times (12.5%)
2 Coins: 31 times (15.5%)
3 Coins: 30 times (15%)
4 Coins: 43 Times (21.5%)
5 Coins: 40 Times (20%)
6 Coins: 31 times (15.5%)

So, after 200 drops.. it looks like it COULD be an equal chance (that's a smallish sample), or it could be a bit of a 
normal distribution.. probably even though, as in a normal distribution, 3 and 4 coins would be the higher totals

Incidently, I got 9 where clearly the triple hit (4.5%).. my skill is at 5.05%.  That seems pretty on target, since 1 (or more) of the 3 and 6 drops could be 1 or 2 tripled... In fact, it's likely at least one was, so that's make it just right.

Only 1 treasure (out of the 200... coins from that weren't counted).. that quite a bit down from the expected (should have been 3), but with small numbers and a small sample, that doesn't mean too much.

Anyone else have any thoughts or knowledge?

[Analog]

If it's an even chance, 1 coin drops should be the least common, followed by 2 coins, due to 3x muddying the results with realistic numbers. 3, 4, 5, and 6 coin drops could be retrocalced upon seeing 9, 12, 15, and 18 coin drops. Consequently, 3 and 6 coin drops would appear highest since there is no way to determine whether a 3 or 6 was tripled or natural.

I would assume this is the distribution you'd see if you were to record at least 1000 kills, but probably need more than that.

An easier way to ballpark is to shut off treasure and measure average kill rate and average gold rate. After recording a long time frame and comparing your rates, it should be:

Code:

gold gained / kills = gold per coin * 3.5 * triple bonus

...given a high enough sample size if it's a random +0 to +5 drop.


Rock on.

[WhiteK]

my personal idea, i think we should have a new upgrade after reach lv max on coin amount, a coin drop amount upgrade, each upgrade make the amount drop 1-3 lower and increasing the amount 4-5 :v everyone happy

[Wildfire2099]

Analog: That would give you the average drop, but not the distribution... helpful, but not what I was going for.

WhiteK: That's an excellent idea!

[Analog]

(06-24-2015, 02:02 AM)Wildfire2099 Wrote: Analog:  That would give you the average drop, but not the distribution... helpful, but not what I was going for.

WhiteK:  That's an excellent idea!

You can calculate some parts of the distribution from the average drop, but I agree it's not going to get you what you want. The only way to do it is to get a massive sample size (ideally with help from other reliable reporters) to get what you really want. The 3s and 6s should be ~5% higher then 3s, 4s, 5s, and ~10% higher than 1s and 2s, but at least by calcing average drop you can remove the triple coins modifier (right off the total), so that could be a secondary check.


Rock on.

[WhiteK]

Well it more like impossible to calculate when ur x3 is 9% :< try my self sometimes it drop 12 coins 6 times in a row and 3-6 coins keep drop so a x3 notice will be help full if Ray add that to the next patch :v

[Wildfire2099]

Or he could just tell us so we don't have to try to figure it out

[Analog]

(06-24-2015, 01:20 PM)Wildfire2099 Wrote: Or he could just tell us so we don't have to try to figure it out ;)

I thought he did tell us. It is random from +0 to +max coin count level.


Rock on.

[Wildfire2099]

Ahhh... well why didn't you say so

That's kinda what I thought, but good to confirm.

[Analog]

(06-25-2015, 04:49 AM)Wildfire2099 Wrote: Ahhh... well why didn't you say so ;)

That's kinda what I thought, but good to confirm.

Sorry, I thought you knew and were trying to confirm.

But I'm not 100% sure Ray said it, I could be imagining things.

And even if he did say it, doesn't hurt to double check his work!


Rock on.