Talk:Extractinator

The items this can produce should be put in a list. - Cyb3rhusky

Common in ice chests?
Hey, just a heads up to anyone. I'm not sure what data is being used to declare it "common in ice chests" (on the wiki as of the time of this comment) but I have 28 ice chests, a very nearly fully explored ice biome (didn't bother with ores, just used spelunk potions looking for chests) and 3 wasted hours of my life that seems to indicate otherwise. I know random is random but if the author just happened to find a couple in ice chests you might consider revising this.


 * Each ice chest has a 5% chance of containing one, so with 28 ice chests there's a 23.8% chance of none of them having one. -- KhymChanur (talk) 21:20, 6 October 2013 (UTC)

Data
I'm really not sure how best to analyse this. At first, I did straight percentages, and worked out the number of coin "events" by subtracting the sum of everything else from the number of "events". Where an event is "put some silt in and get an item or some coins out".

And then I noticed that sometimes you get more than one piece of ore out. So I started again, and will present just the raw data for anyone else to analyse, if they know a better way of doing so.

My method for this, just for reference, is to have exclusively even stacks, so I never feed in a single block (just for consistency). I started with 30,000 silt blocks (15,000 pairs to input), so expect more data to arrive later. This is just for 6,024 blocks/3,012 pairs so far: It seems reasonably safe to say at this stage that ores are all about the same likelihood as one another. Gems are a little harder to make a broad statement on, other than that they are much less common than ores. Amber is a serious oddity, though. And I honestly have no idea how many times I got coins out, so I couldn't begin to work out the average for that, though a small sample of testing implies it's going to be ~20-40 at a time -- Sorceror Nobody (talk) 19:48, 7 October 2013 (UTC)

EDIT: In fact, just after resuming data collection, I've definitely gotten silver coins out. So that makes it damn near impossible to work out how often you get coins out, regardless of how many you get at a time. Well, unless you feed the silt in slowly. Have fun building up a statistically significant sample size with that, though : / -- Sorceror Nobody (talk) 19:51, 7 October 2013 (UTC)
 * The data I listed comes strait from the source code and is accurate. (In case you're curious you can even get platinum coins (0.02%) and gold coins (about 0.25%).) --0icke0 (talk) 20:02, 7 October 2013 (UTC)
 * Hm, I guess that's settled then.
 * That said, the table on the article really isn't very clear in what it means. I only now follow all of it properly, with the benefit of your statement and my own testing experience. So I imagine other people might be very confused by it. Would it not be more prudent to properly list all of the possible outcomes (different coins and also multiple items) with their respective percentages for the main table, and then below that, something like the existing table that "groups" them? -- Sorceror Nobody (talk) 20:28, 7 October 2013 (UTC)
 * This would be quite tricky, considering the way those chances are spread. For example if the outcome is a Copper Ore the amount is calculated as following: It starts with 1 copper ore, now there is a 1 in 20 chance that this amount is increased with 0-1 (equal chance for all those possibilities), than a 1 in 30 chance that the amount is increased with 0-2 (again equal chances), a 1 in 40 chance that it is increased with 0-3, a 1 in 50 chance that it is increased with 0-4 and a 1 in 60 chance that it is increased with 0-5. The average is therefore 1+1/20*((0+1)/2)+1/30*((0+2)/2)+1/40*((0+3)/2)+1/50*((0+4)/2)+1/60*((0+5)/2)=1.1775. However if you want to list the chance of every outcome for this ore you already need 16 rows. Doing so for all results (it gets even worse for coins) would result in a very large table. Therefore I listed it like this. --0icke0 (talk) 21:20, 7 October 2013 (UTC)