pool reward

Lexicon

in my last job i was mainly a data analyst working with loads of SQL data and i do this in my current role as well but far more development. with this in mind i have tried myself to do this with a spreadsheet but in general its pot luck to get any sort of figures you can match.
for example if i base it off the past 3 days transactions i get the following.

trying to work this out off shares is also pretty hard as both pools display the info to you differently. its like trying to predict the weather or comparing apples grown in 2 different cliamtes. however looking at data like this you need to take into consideration how long you were mining at the pool for. and whether you were lucky enough to get good deadlines on that day. i can go days where i get nothing but bad deadlines others where i get really good ones. also you need to take note that about a week ago the block reward dropped by 5%

also there are day where 2000 blocks are mined and other days where 1200 blocks (an example not actual figures) are mined due to deadlines not being constant.

there are to many factors to even try to work this out via a spreadsheet. going by transactions alone.

i'm interested to see what way you have tried working this out. it might help shed some light on things.

the only way i can think of making it like for like is if i do what ninja does and remove the all rounds box and only show the historic shares on pre worked out values then hide all the working out.

Cihlida

I think you have good results. I have not yet reached such.

RichBC

@Lexicon The difference that I think there is between the uray and ninja pools has nothing to do with long term rewards or luck, it's merely an observation, perhaps incorrect, on the short term and how the All Rounds / Historic share is built up and decays either after a Block win or when you leave the Pool.

You understand better than I do how this happens on the uray pool so I will concentrate my description on what I see on the ninja Pool.

I stress that I am not certain about the detail as my understanding is based on observation.

So the way I think it works is that on each Round your Current Round Share is added to your Round Total and this Total divided by 50 is your Historic Share. The Round Total grows for 50 rounds after which the value for the oldest Round is dropped, so it then becomes a rolling average of the last 50 Rounds.

You can see this as it takes some time to build up your Historic Share and also some time for it to Decay if you leave the Pool or have a run of bad Deadlines,

If a Block is found your Historic Share is added to your Current Share and becomes the payout that you will receive 3 Rounds later.

So what's "good" about this relative to uray? For me the main positives are:-

that it removes that element of short term luck that exists with uray that you could have just had a few good deadlines on the trot and then if there is a block you get a big payout.
Putting 1 the other way round on uray you could have a very long run of good deadlines followed by a few rounds of bad deadlines and you reward can be quite small.
on uray after a Block is hit all the counters are zero'd and everyone starts again from scratch. On ninja you carry on from where you left off in terms of the Historic Share and so continue to be rewarded from a Historic Share point of view based on the 50 rounds.

So maybe it's me but I just find it a fairer system removing the element of very short term luck that exists on uray because of the importance of the last few rounds.

Rich

Lexicon

i might have to add in a new pay mode into mine but ill probably not change it in my pool and enable it. mainly because no one else has made these observations. talking to tross he was saying that people were moving from ninja to his pool because his pays better.

i did some more analysis today on your transactions and found that you were infact earning more burst at tross's pool that you were on ninja pool. info below

from observing ninja pool last night this sentence is wrong. if the current shares were added to the historic shares then devided by 50 the block winner would never get anywhere near half of the reward, meaning hasri in the currentshares of the screenshot below wouldnt get 1049. instead it would be much lower as it has the good deadlines of the 50 rounds prior to contend with.

your Current Round Share is added to your Round Total and this Total divided by 50 is your Historic Share. The Round Total grows for 50 rounds after which the value for the oldest Round is dropped, so it then becomes a rolling average of the last 50 Rounds.

the way i see it working is 60% of the block reward is payed out on the current block see fig a.
fig a

then the rest is divided amongst the historic shares.
fig b

what this says to me is that it could be worse at ninja because when someone leaves the pool they will still be getting paid until all 50 blocks of historic shares have gone. meaning other people on the pool are paying for miners who are no longer mining at that pool and paying for blocks that have been already paid out on.

i can see where your coming from but looking at the data you were getting 23.49% more per hour at the other pool compared to ninja.

RichBC

@Lexicon Thanks for the reply, a lot to digest, so I will just make a couple of comments for the moment, and then come back when I have fully read and understood.

My comments are nothing to do with relative long term earnings from different Pools. I am quite happy to accept that be it uray, ninja or solo the swings and roundabouts of the different systems will given time equal out.
I think we could all take a short snapshot of time and look at earnings and come to any conclusion we wanted to :-) Also you are missing a very key point on the Tross earnings, that is the 3K burst giveaway's (Thank You Tross ) without which the numbers would look very different.
I think you have misunderstood / I have badly explained the system that I thing ninja uses for Historic Shares. I will do a better job and come back on that later.

More later stay tuned.

Rich

Lexicon

ive already factored the 3k in to my next set of results and the 3k is an increase of 16% of 2500 which leaves you better off on tross pool by 7% more per hour when you deduct this from the total.
however this also deducts 16% off any actual block-wins during this period meaning even when 16% is taken off actual block wins at 2500 your still earning a significant amount more when blocks are won. the percentage of this could be anywhere between 7%-16% more (rough guess as i need to focus on getting this code ready for release to my vps for testing)

Focus

Please base your findings on a longer sample period, say 1 week, because small samples may be skewed because of various factors as you were describing, difficulty, time, number of blocks in that day etc...

Lexicon

@Focus this is where it gets difficult. as i need a miner to mine on one of urays pools for a week followed by ninja pool for another week. with no additional plots. unfortunately i dont have anyone who has done this so unless someone wants to try mining on both for a week without plotting more drives and let me know then ill be able to base this on a bigger data-set.

this will also need to be done in the middle of the month to avoid the 5% decrease in block reward month on month and even when thats done the only other thing that could skew the results is the amount of miners on each pool. and there capacity's

there isn't a simple way to retrieve the amount of blocks a pool has mined on a day so that's something i cannot factor into any calculations. i've checked through the Wallet Api and couldn't find anything that would do this

i would do this myself but currently i have people trying to brute force their way into my vps via ssh. and im also making tons of improvements to the pool code. along with a day job and various other projects. its just something i dont have time to do.

with the giveaway i did the other day to attract more miners(managed to get an extra 8 miners) I've now spent much more money and time into burst than I've actually got back out of it. i dont know if i am the only one who's actually spent money trying to make burst better. but it shows what im trying to achieve for the community.

RichBC

@Lexicon OK just looking at one point, the overall / relative return on the various Pools.

My position on this is unchanged and is that in the long run there is no difference, swings and roundabouts and all sorts of short term ups and down yes, but long term including solo mining you will come out about even.

But if we must do short term comparisons between my time on Tross & ninja then let's at least use the data I posted for 29 Days on ninja for that side of the comparison. https://forums.burst-team.us/topic/2274/a-month-mining-on-ninja

In summary
Total Burst for the 29 days 15,5290
Burst / Day 5355
Burst / Hour 223
Burst / Hour corrected for block reward reduction 212

Your latest figures for my few Hours on Tross 160

Like I said if you take short periods of time you can prove anything :-)

I will come back to what I wanted to discuss. How I think ninja calculates Historic Share and why I prefer it to uray when I have found a way to express it more clearly.

Rich

Burstde

This was great reading. Nice work everyone. Less than 15 months of mining left.

RichBC

@Burstde Why only 15 Months of mining left?

Rich

Burstde

@RichBC the rewards keep on getting reduced every month.

2634-2504 = 130 x 15 = 1950 so not much time left.

Reward get reduced 5%
2634 x 5% = 125 x 15 = 1875 okay how about less than 2 years of mining left.

RichBC

@Burstde In 15 Months the reward will reduce to 1160 and in 24 Months it will be down to 731. So yes a substantial reduction, however mining hopefully not dead as long as we see increasing transaction fees and most important Burst strengthen against the Dollar.

But yes if conversly we continue to see a weakening of Burst exchange rates it could happen....

Rich

Lexicon

i agree as i previously stated it was hard to get any sort of reliable info by doing any sort of comparisons long or short term. all i did was do one to prove how hard it is. as last month the block reward was higher. there was probably also less miners and since then people have plotted more drives. and difficulty changes along with various other factors.

also burst has about 533,747,063 coins left to be mined, burst has a limit of 2.1 billion coins after that no more can be generated and it will be transaction fees that you mine.

https://forums.burst-team.us/topic/1297/is-the-party-going-to-end-soon/3
this one was taken today

and heres a snapshot from 2 months ago

nixxda

I'm up for the challenge! or better comparison. In fact I already started!
I have set up clean logs and I'm going to mine one week on burst.lexitoshi.uk and one on burst.ninja.
But, the one thing most interesting to me isn't the actual amount paid out, (well, of course that to !) but rather the the pools performance!
And I'm sorry for bringing this up in this thread!-/
What I am talking about is the pools ability to confirm my Deadlines!
Now, this may not be the pools fault but rather my geolocations, clogged internet in general and the amount of miners (lemmings you are !-) in those ninja pools.
I am located in the middle of Europe with an rather nice ISP and I still cant get all my DL's confirmed from those clogged US servers! This annoys me quite a bit!
So for me, this is the main factor for choosing a pool. Not the actual payout scheme. Because I strongly suspect them to be the same in the end.
I would evan go so far as to say, if the pool your mining on has at least about 100Tb overall, it doesn't matter anymore! Your payout is always going to be the same overtime! As long as you can get your DL's confirmed!-)

some Numbers:

random day at lexitoshi

random day at ninja

yes, I know, bad picture! "connectionTimeout" is set to 20'000ms and i have tried increasing it. but to no avail!

I'll be back! With real numbers for a payout comparison.

edit: I realized that I have to clarify that statement because it might be misleading!
first of, 11 "not confirmed" Deadlines is a small percentage of the 303 "confirmed!" I got in the same timeframe!
second, of those 11 an evan smaller percentage might have won something! (hard to check afterwards) most probably non!
but, its still annoying not to know! And for me, it gets worse on pool.burst-team.us!
If it works for you, by all means go ahead mine there!-)

Burstde

What is pool.burst.de connection timeout?

nixxda

@Burstde said in pool reward:

What is pool.burst.de connection timeout?

I dont know!-) you set the "connection timeout" in jminer.properties.
And, if I understood that right, its the time jminer waits for a confirmation of a deadline.
Quite often you will see that your dl made it through anyway and you just did not get the confirmation back.
And some other times it doesn't matter because it wasn't your best one in that round anyway.
But ever so often I happen to see one which didn't go through. And it may have been the one!-)

RichBC

@nixxda said in pool reward:

But, the one thing most interesting to me isn't the actual amount paid out, (well, of course that to !) but rather the the pools performance!
And I'm sorry for bringing this up in this thread!-/
What I am talking about is the pools ability to confirm my Deadlines!

You will find no Pool better than ninja for rapid and reliable confirmation of deadlines, relatively few miners and because of the low Deadline the number of nonces submitted each round is less than 2x the number of miners.. I have a poor internet connection and it's one of the main reasons I mine there.

Rich

nixxda

@RichBC Well,... no! But as I said, if it works for you and your happy with it, good! You are probably mining from the US. But, you know there are other people out there, like "the rest of the world"!-)

I am aware that both of these pool servers we are talking about are located in the US. Ninja in Lenoir and Lexitoshi in Dallas as it seams. And some basic testing of their response time while in "stand still" shows an equal result for me. But still, I do get some "not confirmed"s from ninja and non from lexitoshi.

Back on topic, what I'm trying to suggest is, maybe it would be better and worth it to spread out to smaller pools! If only to to avoid the annoying "not to know" factor. As in "not to know if those not confirmed Dl's would have won something"!
If it's the same on both pools payment wise, I dont know jet! But i will! In about two weeks from now!-)

RichBC

@nixxda I am in the UK, with a very poor internet connection (1.5Mb/s Down and 0.4Mb/s UP on a good Day often worse than this) and have tried most of the pools and ninja is definitely the best i have found. But as you say what matters is what works best for you.

Rich