Guides

[i][/i][b]I haven't been paying much attention to new changes, but I've been told the drop rates have been changed and normalized across venues, so this is no longer current unless someone decides to revive the research. (Feel free to copy the spreadsheet setup.)[/b] [size=7][b][url=https://docs.google.com/spreadsheets/d/16Cct1HgAZX1cPfkeqNIkr0nYTuGO8Fq5gNKYJzvA92U/edit] --> MAIN SPREADSHEET LINK[/url][/b][/size] [b]The question that prompted this thread:[/b] we have guides telling us how many or what percent of each element and Neutral enemies each Coliseum level has, but how should they be counted compared to each other when farming for festival currency? Do two Wind enemies give you a higher or lower chance of getting festival currency than three Neutral enemies? [b][u]Current approximate best answer:[/u] The drop rates depend on the Coliseum venue level. Enemies of the festival's element drop currency around 25-45% of the time, and Neutral enemies 15-35% of the time[/b] - lower percentages for lower-level venues, higher for higher-level, although the pattern still isn't entirely clear and we're missing data for some venues. Below is a graph showing the % currency drop rates from elemental and Neutral enemies by Coliseum venue (across all festivals, for venues where we have data from at least 100 enemies). [img]https://images.plurk.com/41HQgDUnSgoggbNUUHGJYN.jpg[/img] For the [b]most up-to-date plot, detailed data and more plots[/b] with error bars, or to [b]contribute your own data[/b] to the project, see [b][url=http://docs.google.com/spreadsheets/d/16Cct1HgAZX1cPfkeqNIkr0nYTuGO8Fq5gNKYJzvA92U/edit]THIS SPREADSHEET[/url][/b] - "raw_data" tab is where you can add your results, "summary" tab has the data totals, drop rates and plots (auto-generated from the raw_data tab), "currency/hour etc" tab is where people can add more direct measures like overall currency per hour or per battle in each venue. Note that sometimes there are [b]errors[/b] - we have at least two instances where the drop rate was almost certainly set incorrectly*, and there are documented cases of some enemies not dropping currency at all or the wrong element's enemies dropping it (which the admins usually fix after people point it out - but if only the drop rate is wrong, it's much harder to notice/prove the problem). [i]* those cases are marked as "BUGGED" in the raw data section of the spreadsheet and excluded from the summary/plots. [/i] [b]Info from admins:[/b] The admins confirmed that the drop rates from both elemental and neutral enemies are higher in higher Coliseum venues, but gave no details. ------------------------ If anyone wants to [b]help us gather more data[/b] during future festivals, that would be great! We can definitely use more data, the results are still very messy. You can:[LIST] [*]post your data here and ping me to add it for you [*]add data to the spreadsheet yourself! But you'll have to PM me your email address so I can invite you to edit (I used to just let anyone edit, but FR public spreadsheets have vandalism issues :/ )[/LIST] Ping me if anything is unclear. [b]What we're counting[/b] is neutral and festival-element enemies (not groups) and currency drops from each. IGNORE any groups containing mixed elemental+Neutral enemies - for why, see [url=http://www1.flightrising.com/forums/gde/1055074/19#10594486]Drezdin's post here[/url] for a summary and the conversation before that for more details - or ask me. You can count just elemental or just Neutral enemies instead of both, if you want, as long as you decide in advance which one you're doing for each half-hour or so. If you want to keep track of the data by enemy or pack, feel free to post that here or add another tab to the spreadsheet - it could be useful for more detailed analysis, but it's hard to incorporate into a single all-festivals spreadsheet sensibly. I won't say no to chest drop info either. And [i]please be careful about bugs[/i]! Multiple festivals have had cases of some enemies not dropping currency, or wrong-element enemies dropping it. It might be best to separate the data by enemy and by day, so that the bugged sections can be thrown away if needed. [b]Ping list[/b] is on the spreadsheet - you can add/remove yourself, or ask me if you don't have editing rights.

THIS POST CONTAINS SOME OLDER NUMBERS AND STATISTICAL ANALYSIS (later we switched to just using the auto-updating spreadsheet instead of running a once-a-month analysis by hand)

Here's the detailed data+analysis from Wind+Water 2014, with statistics:
(from multiple people, comparing two festivals and different Coliseum areas)

Notes on how the data works:
- what I'm actually trying to calculate is the probability of any one elemental or Neutral enemy dropping currency. I'm assuming that the drop rate is per enemy, not different between groups.
- the way the counts are done is the number of enemies that did and didn't drop currency, in different group types. So a group of 2 Water enemies that dropped 1 GSD would be "1 yes, 1 no", etc.
- a 95% confidence interval is the range inside which we're pretty sure the real number is. The more data we have, the tighter the range. You can calculate it here.

We have a lot of data from the Water festival, mostly from two areas (people who contributed data are listed in parentheses):

1) Forgotten Cave (nika, tiffdawriter, Sylvandyr)
2) Arena (nika, tiffdawriter, Sylvandyr)
Do we have good reason to think that the Water vs Neutral drop rates are different? YES - the probability that we'd get that much of a difference by chance (using the chi-squared test of independence) is below 1% for both Cave and Arena.

Do we have good reason to think that the drop rates differ between areas? NO - the values are very similar, and probability of getting such small differences by chance is >90% for both Water and Neutral. NOTE that this was either changed later, or was a weird coincidence - later data (see first post/spreadsheet) shows that there definitely ARE differences between areas.

So since we think the drop rates are the same across areas (well, at least between those two areas, but I'm going to extrapolate to the rest of them), I'm going to add up the data from all the areas - the two above, and other areas that we didn't have enough data from to do much separately:

3) Harpy's Roost (nika) 4) Waterway (nika, tiffdawriter) 5) Bamboo Falls (tiffdawriter) 6) Scorched Forest (Coldblooded)
Overall counts from all areas for this festival: And the mixed-group ones, which aren't very useful except to look at and make sure they're somewhere in between:
For comparison, here is my data from the Wind festival, from multiple areas (I wasn't counting different areas separately yet then):
Is there a reason to think the drop rates are different between the two festivals? MAYBE:
The Water vs Wind drop rates do look a bit different, but the 95% confidence intervals overlap quite a lot. There's a 16% chance of getting that much difference randomly, so it's sort of borderline - maybe the drop rate did change, maybe we just happened to get more lucky last festival than this one. If this was an academic paper, we wouldn't count the difference as significant. XD
The Neutral vs Neutral difference is minor, and there's a 62% chance of seeing that randomly, so we can assume that hasn't changed.

If we add the data from the two festivals together as our best estimate at the overall rate (which makes sense IF we assume that there wasn't a real change in drop rate between festivals): So I'd say those are our best current estimates.

@Ammoth Good to know someone other than me thinks it's useful! XD

And I'll definitely have to try it by area the next festival, since some people have been hypothesizing that lower areas may have lower currency drop rates per enemy...

@Drezdin ♥!

@Sylvandyr Cool, thanks!

@SaF Yeah, frequencies of mob occurrence are much harder, since there's so many possibilities rather than just a yes/no! There's another thread in Guides by Ammoth where we're getting together some people to work on that, so help would be awesome (and at least that doesn't have to be done during holidays!).
And oh no, fieldwork over an FR holiday! Why can't we all get time off for holiday weeks? XD (Maybe you can at least buy some currency and/or the actual holiday stuff, if the prices get low enough? I do that sometimes if grinding isn't working out that week.)