Mining with profit : NS for ores

[Necy]

Indeed Mark
Any analyst will confirm that with 100 examples you have an idea of the result.
I see that enough when I'm testing the finders.
With 100 drops, any finder, all different times... around the same results.

You have an idea with 100 drops, but in my experience there is too much variance in such few loot events in this game to make any real conclusion.
I wouldn't trust resource distributions until at least a sample size of 1000, and it's actually one of the biggest gripes I have with these threads.

That said, Mark's sample size here should be more than enough.

 

[Leeloo Miner]

You have an idea with 100 drops, but in my experience there is too much variance in such few loot events in this game to make any real conclusion.
I wouldn't trust resource distributions until at least a sample size of 1000, and it's actually one of the biggest gripes I have with these threads.

That said, Mark's sample size here should be more than enough.

Well, must be a coincidence then that my unamped Igni return is also around 7% after the the last (unpublished) runs, where every run is 100 drops, even when we both where mining on dif planets.
Mark did the test cause ingame we talk and because we had the same results, he tried the amp 2 and unamped..

Did you read this test ?

Call for miners: join the mining train hit rate testing

A bunch of us have been discussing whether other people mining in the same area before you affects your hit rate, TT, etc., but there really isn't any good data on this. There are a few ways to test this, but the first test to science the answer out if pretty straightforward if we have a few...

www.planetcalypsoforum.com

 

[kingofaces]

You have an idea with 100 drops, but in my experience there is too much variance in such few loot events in this game to make any real conclusion.
I wouldn't trust resource distributions until at least a sample size of 1000, and it's actually one of the biggest gripes I have with these threads.

That said, Mark's sample size here should be more than enough.

100-200 can be more than plenty if you are doing actual statistical tests and have a a valid experimental design setup for those tests. Even a sample of 30 is good enough in many situations. There are examples of that in my signature. The short of it is that those tests are designed to account for variability and tell you what an expected average would be if you instead sampled 1,000, 10,000, etc. instead.

The problem is that most people don’t do that and just use raw averages to compare instead. There may be a way to analyze the data in this post, but I haven’t really looked at it too much to say much about it yet.

 

[M Rufen Power]

You have an idea with 100 drops, but in my experience there is too much variance in such few loot events in this game to make any real conclusion.
I wouldn't trust resource distributions until at least a sample size of 1000, and it's actually one of the biggest gripes I have with these threads.

That said, Mark's sample size here should be more than enough.

I think I done about 1679 dual drops on Lv2, is alot if you think about it. Again, I mined in a concentrated area for Igni. I honestly believe if I hadn't found the Igni size 9 claim, I'd probably be around 7% distribution. I think I got lucky tbh on that run, not just with the huge TT surplus, but also Igni claim sizes.

Well, must be a coincidence then that my unamped Igni return is also around 7% after the the last (unpublished) runs, where every run is 100 drops, even when we both where mining on dif planets.
Mark did the test cause ingame we talk and because we had the same results, he tried the amp 2 and unamped..

Did you read this test ?

Call for miners: join the mining train hit rate testing

A bunch of us have been discussing whether other people mining in the same area before you affects your hit rate, TT, etc., but there really isn't any good data on this. There are a few ways to test this, but the first test to science the answer out if pretty straightforward if we have a few...

www.planetcalypsoforum.com

I'm now starting to get the feeling that this game is very avatar based, not just in the digital sense of the word but knowledge of game mechanics as well. I've met some really lucky avatars in my time in this game, these individuals tend to do things very differently in terms of playstyle to most of the playerbase.

Having said all that, I think me having a break from mining in the short term is an ideal situation for me, and will be back to mining, hopefully in the next month, whilst focusing on hunting a litlte bit. Then I will try my luck with level 3 amps and at that point in time try analysing the distribution %'s further.

So I can't for 100% certainty know if MA released a silent mining distribution patch, or if it's planet specific to Caly etc too early for me disclose this as fact. So based on the claim sizes I received, I reckon I got higher than avg, and lucked out, thats just my personal take on it, intuition nothing more at this stage.

 

[Necy]

100-200 can be more than plenty if you are doing actual statistical tests and have a a valid experimental design setup for those tests. Even a sample of 30 is good enough in many situations. There are examples of that in my signature. The short of it is that those tests are designed to account for variability and tell you what an expected average would be if you instead sampled 1,000, 10,000, etc. instead.

The problem is that most people don’t do that and just use raw averages to compare instead. There may be a way to analyze the data in this post, but I haven’t really looked at it too much to say much about it yet.

But this also assumes that the distribution of resource is always the same.
Which there is no evidence to suggest. On the contrary, I have noticed very fluctuating distributions between runs of ~500 drops.

There are multiple things to account for here:

• Possibly loot table distribution switching over time. We already know that areas can switch their entire loot table to another set of resources, so why not an in-between where just the distribution changes?
• Possibly some resources having reduced/increased hit rate in timed windows. We see this happening in hunting all the time. Why not also for rarer resources in mining? This would mean that any shorter run has a risk of missing the window of the rarer resources.

These two mechanics would make any short run invalid for determining resource distribution.

Another topic is resources with very low hit rate. For example Redulite. Or let's even say rare resources while amped. Anything that sits at the 1% or less hitrate mark would be too much of a subject to variance to get a real idea after 100 drops. And it's especially unfortunate to make assumptions on this data if most of your MU depends on that variance.

 

[Necy]

Well, must be a coincidence then that my unamped Igni return is also around 7% after the the last (unpublished) runs, where every run is 100 drops, even when we both where mining on dif planets.

It most definitely is a coincidence if you're talking about a single run. If you're talking about the Igni return for all of your runs combined, that's likely to be accurate because you would have much more than 100 samples
If you do 10 runs with 100 drops with the same finder, I guarantee you that you will see a variance of distributions between all the runs. Your Igni % will likely vary quite a bit from run to run.
I would be very happy to trust the resource distribution of those ten runs combined, but not a single run.
Unless there is some mechanic in the game that strictly enforces that you get 7% Ignisium. But that would be more tricky to implement as a developer, and per-avatar data intensive, than just rolling the resource based on a parameter.

 

[Leeloo Miner]

Unless there is some mechanic in the game that strictly enforces that you get 7% Ignisium. But that would be more tricky to implement as a developer, and per-avatar data intensive, than just rolling the resource based on a parameter.

That is exacly what is happening now after the last "quiet" maintenance, but you could see that on auction also cause before MM started, Igni was becoming very rare and even hit 300%. Now it is back ... (capped @ 7% unamped) and it dropped down to 220% + there is another factor about distribution but to soon to tell about that.

 

[Necy]

That is exacly what is happening now after the last "quiet" maintenance, but you could see that on auction also cause before MM started, Igni was becoming ve(ry rare and even hit 300%. Now it is back ... (capped @ 7% unmped) and it dropped down to 220% + there is another factor but cant say that for sure at this time.

Those are two different things. I think we can all agree that the hitrate of Ignisium was much lower for a period of time, but I find it hard to believe that any mechanic is strictly enforcing a 7% return of Ignisium other than statistical averages of long enough runs.
Also, "quet maintenance" is called a server side change.

 

[kingofaces]

But this also assumes that the distribution of resource is always the same.
Which there is no evidence to suggest. On the contrary, I have noticed very fluctuating distributions between runs of ~500 drops.

There are multiple things to account for here:

• Possibly loot table distribution switching over time. We already know that areas can switch their entire loot table to another set of resources, so why not an in-between where just the distribution changes?
• Possibly some resources having reduced/increased hit rate in timed windows. We see this happening in hunting all the time. Why not also for rarer resources in mining? This would mean that any shorter run has a risk of missing the window of the rarer resources.

These two mechanics would make any short run invalid for determining resource distribution.

Another topic is resources with very low hit rate. For example Redulite. Or let's even say rare resources while amped. Anything that sits at the 1% or less hitrate mark would be too much of a subject to variance to get a real idea after 100 drops. And it's especially unfortunate to make assumptions on this data if most of your MU depends on that variance.

If there’s variation over time, the experimental design should account for that. If you’re testing two different finders or amps to see how a resource rate may vary, you should be alternating between “treatments” each find. That way, if the the underlying resource percentage changes, it’s accounted for while focusing just on the underlying mechanism of what you’re really trying to test. That’s why it’s best not to try to test everything at once but have very focused comparisons like you see in the posts I mentioned.

 

[Gopher]

I think I done about 1679 dual drops on Lv2, is alot if you think about it. Again, I mined in a concentrated area for Igni. I honestly believe if I hadn't found the Igni size 9 claim, I'd probably be around 7% distribution. I think I got lucky tbh on that run, not just with the huge TT surplus, but also Igni claim sizes.


I'm now starting to get the feeling that this game is very avatar based, not just in the digital sense of the word but knowledge of game mechanics as well. I've met some really lucky avatars in my time in this game, these individuals tend to do things very differently in terms of playstyle to most of the playerbase.

Having said all that, I think me having a break from mining in the short term is an ideal situation for me, and will be back to mining, hopefully in the next month, whilst focusing on hunting a litlte bit. Then I will try my luck with level 3 amps and at that point in time try analysing the distribution %'s further.

So I can't for 100% certainty know if MA released a silent mining distribution patch, or if it's planet specific to Caly etc too early for me disclose this as fact. So based on the claim sizes I received, I reckon I got higher than avg, and lucked out, thats just my personal take on it, intuition nothing more at this stage.

You could have been the only one in that area, I find bigger claims of wave-based materials when theres no-one there or only like 1 other person.

 

[M Rufen Power]

You could have been the only one in that area, I find bigger claims of wave-based materials when theres no-one there or only like 1 other person.

I 100% agree. I never saw anyone on my radar the whole time I was out mining in the concentrated area, so it's definitely something I can confirm.