Guides
Community created guides, helpful strategies, and more.
TOPIC | data on Gathering item frequencies
@kengee Thanks! c:
@LostCure Thanks for the data and continued bumps!
@Bwee Are you sure you found a rooster hunting at level 14? That's not supposed to happen... Same for an earthworm at level 15 insect-catching
@LostCure Thanks for the data and continued bumps!
@Bwee Are you sure you found a rooster hunting at level 14? That's not supposed to happen... Same for an earthworm at level 15 insect-catching
Woah, thats pretty awesome, good job @Nika! Looks like we might be needing more level variation, so hey people fill out these sheets please!
Woah, thats pretty awesome, good job @Nika! Looks like we might be needing more level variation, so hey people fill out these sheets please!
@nika Wow, this is really neat! I'm interested to see what kind of data we'll find--I'll start adding some data of my own from late-teens-level insect catching too :o
@nika Wow, this is really neat! I'm interested to see what kind of data we'll find--I'll start adding some data of my own from late-teens-level insect catching too :o
I'm on FR time! Pronouns: she/her
Accents | Tumblr | Commissions
Excerpts from a Longneck Journal - beastclans-POV illustrated fic, finished
Accents | Tumblr | Commissions
Excerpts from a Longneck Journal - beastclans-POV illustrated fic, finished
@Nika
Most of my gathering areas are in the really low levels but I keep forgetting to record them ( foraging lvl 1, insects 6, fishing 7, hunting 9, scavenging's at 14, digging at 16). I will try to be better about it so you can have delicious data. Great work, though! I'm loving all this info! :D
I wonder if it's possible / if there's enough data now to start doing some moderate stats? A t-test between two items of equivalent rarity shouldn't be significantly different from each other (p greater than 0.05). Or determining how many more responses are needed for the test results to be accurate...
Most of my gathering areas are in the really low levels but I keep forgetting to record them ( foraging lvl 1, insects 6, fishing 7, hunting 9, scavenging's at 14, digging at 16). I will try to be better about it so you can have delicious data. Great work, though! I'm loving all this info! :D
I wonder if it's possible / if there's enough data now to start doing some moderate stats? A t-test between two items of equivalent rarity shouldn't be significantly different from each other (p greater than 0.05). Or determining how many more responses are needed for the test results to be accurate...
@Nika
Most of my gathering areas are in the really low levels but I keep forgetting to record them ( foraging lvl 1, insects 6, fishing 7, hunting 9, scavenging's at 14, digging at 16). I will try to be better about it so you can have delicious data. Great work, though! I'm loving all this info! :D
I wonder if it's possible / if there's enough data now to start doing some moderate stats? A t-test between two items of equivalent rarity shouldn't be significantly different from each other (p greater than 0.05). Or determining how many more responses are needed for the test results to be accurate...
Most of my gathering areas are in the really low levels but I keep forgetting to record them ( foraging lvl 1, insects 6, fishing 7, hunting 9, scavenging's at 14, digging at 16). I will try to be better about it so you can have delicious data. Great work, though! I'm loving all this info! :D
I wonder if it's possible / if there's enough data now to start doing some moderate stats? A t-test between two items of equivalent rarity shouldn't be significantly different from each other (p greater than 0.05). Or determining how many more responses are needed for the test results to be accurate...
@Jacq More lower-level data would be great, please record it if you have the time! :)
Re: stats - hmm, do you think a t-test (or anova) is going to be all right for this kind of case? I was originally going to use a chi-squared test of goodness of fit (similar except without assuming normality), but then I realized I can't really do that, because the stack system means each occurrence of an item isn't independent.
Like, think about an extreme case: if you had a really low chance of a stack, but the stack sizes can be anything from 1 to 100, with equal probability. In that case, you could easily get 2 items once and 80 items another time, and if you compared these two numbers using a t-test or anything similar, they would show up as significantly different, but actually this isn't an abnormal result!
So I think to do proper statistics that way, we'd have to count how many times a STACK of a given item shows up, because that's the primary number here - and then we could independently compare the number of stacks of different items seen, and their stack sizes. But counting each item separately even if there were multiple per stack basically increases our numbers by (probably) about 2.5-fold, which would make the differences look artificially significant. Am I making sense here? ;)
I may be wrong about this, of course. I do statistics at work, but it's not my primary field. And there may be other methods I'm missing...
Re: stats - hmm, do you think a t-test (or anova) is going to be all right for this kind of case? I was originally going to use a chi-squared test of goodness of fit (similar except without assuming normality), but then I realized I can't really do that, because the stack system means each occurrence of an item isn't independent.
Like, think about an extreme case: if you had a really low chance of a stack, but the stack sizes can be anything from 1 to 100, with equal probability. In that case, you could easily get 2 items once and 80 items another time, and if you compared these two numbers using a t-test or anything similar, they would show up as significantly different, but actually this isn't an abnormal result!
So I think to do proper statistics that way, we'd have to count how many times a STACK of a given item shows up, because that's the primary number here - and then we could independently compare the number of stacks of different items seen, and their stack sizes. But counting each item separately even if there were multiple per stack basically increases our numbers by (probably) about 2.5-fold, which would make the differences look artificially significant. Am I making sense here? ;)
I may be wrong about this, of course. I do statistics at work, but it's not my primary field. And there may be other methods I'm missing...
@Jacq More lower-level data would be great, please record it if you have the time! :)
Re: stats - hmm, do you think a t-test (or anova) is going to be all right for this kind of case? I was originally going to use a chi-squared test of goodness of fit (similar except without assuming normality), but then I realized I can't really do that, because the stack system means each occurrence of an item isn't independent.
Like, think about an extreme case: if you had a really low chance of a stack, but the stack sizes can be anything from 1 to 100, with equal probability. In that case, you could easily get 2 items once and 80 items another time, and if you compared these two numbers using a t-test or anything similar, they would show up as significantly different, but actually this isn't an abnormal result!
So I think to do proper statistics that way, we'd have to count how many times a STACK of a given item shows up, because that's the primary number here - and then we could independently compare the number of stacks of different items seen, and their stack sizes. But counting each item separately even if there were multiple per stack basically increases our numbers by (probably) about 2.5-fold, which would make the differences look artificially significant. Am I making sense here? ;)
I may be wrong about this, of course. I do statistics at work, but it's not my primary field. And there may be other methods I'm missing...
Re: stats - hmm, do you think a t-test (or anova) is going to be all right for this kind of case? I was originally going to use a chi-squared test of goodness of fit (similar except without assuming normality), but then I realized I can't really do that, because the stack system means each occurrence of an item isn't independent.
Like, think about an extreme case: if you had a really low chance of a stack, but the stack sizes can be anything from 1 to 100, with equal probability. In that case, you could easily get 2 items once and 80 items another time, and if you compared these two numbers using a t-test or anything similar, they would show up as significantly different, but actually this isn't an abnormal result!
So I think to do proper statistics that way, we'd have to count how many times a STACK of a given item shows up, because that's the primary number here - and then we could independently compare the number of stacks of different items seen, and their stack sizes. But counting each item separately even if there were multiple per stack basically increases our numbers by (probably) about 2.5-fold, which would make the differences look artificially significant. Am I making sense here? ;)
I may be wrong about this, of course. I do statistics at work, but it's not my primary field. And there may be other methods I'm missing...
@nika
I've been worrying about hunting because I want those familiar components, but I could perhaps switch to foraging, fishing, and/or insect catching, as I am lower level in those areas. And this is really interesting. Thanks for doing it. :)
I've been worrying about hunting because I want those familiar components, but I could perhaps switch to foraging, fishing, and/or insect catching, as I am lower level in those areas. And this is really interesting. Thanks for doing it. :)
@nika
I've been worrying about hunting because I want those familiar components, but I could perhaps switch to foraging, fishing, and/or insect catching, as I am lower level in those areas. And this is really interesting. Thanks for doing it. :)
I've been worrying about hunting because I want those familiar components, but I could perhaps switch to foraging, fishing, and/or insect catching, as I am lower level in those areas. And this is really interesting. Thanks for doing it. :)
@nika
Ahh, you're right. I forgot about the difference between # of items and stack size. You could potentially fudge it by averaging the stack sizes to 2 (because the range is 1-5, right, no chance of 0 stacks?) With enough data the sizes of the stack should basically normalize themselves anyways. It would acheive the same result without inflating the numbers too badly (or at all), but could still skew the significance results for the rarer items. Testing for alpha would work, but I'm not sure if that's an easy test to do in Excel.
Hmm, I'd have to think/sleep on this. I think Chi-squares are better than an anova in this case, but it's still a tricky dicky issue. (Stats is also not my primary field, though, but I have to do them or no one takes my whimsical ideas seriously ;) )
May I steal these sheets and try to look at them in R (someday, no promises of that day being soon)?
(data nerd fist bump)
Ahh, you're right. I forgot about the difference between # of items and stack size. You could potentially fudge it by averaging the stack sizes to 2 (because the range is 1-5, right, no chance of 0 stacks?) With enough data the sizes of the stack should basically normalize themselves anyways. It would acheive the same result without inflating the numbers too badly (or at all), but could still skew the significance results for the rarer items. Testing for alpha would work, but I'm not sure if that's an easy test to do in Excel.
Hmm, I'd have to think/sleep on this. I think Chi-squares are better than an anova in this case, but it's still a tricky dicky issue. (Stats is also not my primary field, though, but I have to do them or no one takes my whimsical ideas seriously ;) )
May I steal these sheets and try to look at them in R (someday, no promises of that day being soon)?
(data nerd fist bump)
@nika
Ahh, you're right. I forgot about the difference between # of items and stack size. You could potentially fudge it by averaging the stack sizes to 2 (because the range is 1-5, right, no chance of 0 stacks?) With enough data the sizes of the stack should basically normalize themselves anyways. It would acheive the same result without inflating the numbers too badly (or at all), but could still skew the significance results for the rarer items. Testing for alpha would work, but I'm not sure if that's an easy test to do in Excel.
Hmm, I'd have to think/sleep on this. I think Chi-squares are better than an anova in this case, but it's still a tricky dicky issue. (Stats is also not my primary field, though, but I have to do them or no one takes my whimsical ideas seriously ;) )
May I steal these sheets and try to look at them in R (someday, no promises of that day being soon)?
(data nerd fist bump)
Ahh, you're right. I forgot about the difference between # of items and stack size. You could potentially fudge it by averaging the stack sizes to 2 (because the range is 1-5, right, no chance of 0 stacks?) With enough data the sizes of the stack should basically normalize themselves anyways. It would acheive the same result without inflating the numbers too badly (or at all), but could still skew the significance results for the rarer items. Testing for alpha would work, but I'm not sure if that's an easy test to do in Excel.
Hmm, I'd have to think/sleep on this. I think Chi-squares are better than an anova in this case, but it's still a tricky dicky issue. (Stats is also not my primary field, though, but I have to do them or no one takes my whimsical ideas seriously ;) )
May I steal these sheets and try to look at them in R (someday, no promises of that day being soon)?
(data nerd fist bump)
@nika Hey there. I just checked my text file for my gathering. For earthworm I wrote down in my file that in one turn I had found 2 mealworms and 3 pillbugs in the same turn. And for rooster I wrote down that in one turn I only got 1 clucker. Unless clucker wasn't supposed to be written as rooster? It's been some time since I entered those in but I'm fairly careful with the information I input. Haven't been able to record anything recently because of vacation but starting tomorrow I'll be back in the game!
@nika Hey there. I just checked my text file for my gathering. For earthworm I wrote down in my file that in one turn I had found 2 mealworms and 3 pillbugs in the same turn. And for rooster I wrote down that in one turn I only got 1 clucker. Unless clucker wasn't supposed to be written as rooster? It's been some time since I entered those in but I'm fairly careful with the information I input. Haven't been able to record anything recently because of vacation but starting tomorrow I'll be back in the game!
@Meredith44 Fishing/etc data would be good, but I wouldn't want to deprive you of familiars! :) This is a long-term project anyway.
@Jacq Hmm, you're right, maybe dividing all the numbers by 3 would be okay, assuming the stack size distribution isn't skewed and doesn't change between levels/rarities/anything... I may do some private collecting of data including stack sizes to see what it looks like.
And yeah, feel free to play around with the data! Just tell me what you get. :D I do more complicated stuff in python rather than in googledocs - I'm only now learning the basics of doing googledocs programming-type stuff.
/fist bump! :D
@Bwee Ohh! Sorry, mealworms are listed as "mealworm" (2-star rarity) - earthworms are another item (earthworm/blackworm/redworm/greenworm). And cluckers were listed as "hen" (I just updated that to "hen/clucker/orpington" for clarity) - roosters are a different item. I moved your data to the right rows.
Sorry about the confusion, and thank you for keeping such detailed records that you could look that up!
@Jacq Hmm, you're right, maybe dividing all the numbers by 3 would be okay, assuming the stack size distribution isn't skewed and doesn't change between levels/rarities/anything... I may do some private collecting of data including stack sizes to see what it looks like.
And yeah, feel free to play around with the data! Just tell me what you get. :D I do more complicated stuff in python rather than in googledocs - I'm only now learning the basics of doing googledocs programming-type stuff.
/fist bump! :D
@Bwee Ohh! Sorry, mealworms are listed as "mealworm" (2-star rarity) - earthworms are another item (earthworm/blackworm/redworm/greenworm). And cluckers were listed as "hen" (I just updated that to "hen/clucker/orpington" for clarity) - roosters are a different item. I moved your data to the right rows.
Sorry about the confusion, and thank you for keeping such detailed records that you could look that up!
@Meredith44 Fishing/etc data would be good, but I wouldn't want to deprive you of familiars! :) This is a long-term project anyway.
@Jacq Hmm, you're right, maybe dividing all the numbers by 3 would be okay, assuming the stack size distribution isn't skewed and doesn't change between levels/rarities/anything... I may do some private collecting of data including stack sizes to see what it looks like.
And yeah, feel free to play around with the data! Just tell me what you get. :D I do more complicated stuff in python rather than in googledocs - I'm only now learning the basics of doing googledocs programming-type stuff.
/fist bump! :D
@Bwee Ohh! Sorry, mealworms are listed as "mealworm" (2-star rarity) - earthworms are another item (earthworm/blackworm/redworm/greenworm). And cluckers were listed as "hen" (I just updated that to "hen/clucker/orpington" for clarity) - roosters are a different item. I moved your data to the right rows.
Sorry about the confusion, and thank you for keeping such detailed records that you could look that up!
@Jacq Hmm, you're right, maybe dividing all the numbers by 3 would be okay, assuming the stack size distribution isn't skewed and doesn't change between levels/rarities/anything... I may do some private collecting of data including stack sizes to see what it looks like.
And yeah, feel free to play around with the data! Just tell me what you get. :D I do more complicated stuff in python rather than in googledocs - I'm only now learning the basics of doing googledocs programming-type stuff.
/fist bump! :D
@Bwee Ohh! Sorry, mealworms are listed as "mealworm" (2-star rarity) - earthworms are another item (earthworm/blackworm/redworm/greenworm). And cluckers were listed as "hen" (I just updated that to "hen/clucker/orpington" for clarity) - roosters are a different item. I moved your data to the right rows.
Sorry about the confusion, and thank you for keeping such detailed records that you could look that up!
@nika my fishing and foraging are both level 12 and insect catching is lvl 8 so I'll definitely be adding data from those areas from now on. I also added yesterday's data cause I forgot to add it in. I got to lvl 23 Hunting(didn't gather more from there) but I'm satisfied with it personally for now so I'm moving on to fishing and the other lower level areas.
@nika my fishing and foraging are both level 12 and insect catching is lvl 8 so I'll definitely be adding data from those areas from now on. I also added yesterday's data cause I forgot to add it in. I got to lvl 23 Hunting(didn't gather more from there) but I'm satisfied with it personally for now so I'm moving on to fishing and the other lower level areas.
|
FR time +9
They/Them |
|