Bitcoin Forum

I realize I may have come across as a bit of an F2pool fanboy in the past (love the PPS consistency) but I'm not one to limit my options, and alternative ZEC mining experience for me was admittedly only with Suprnova which I found terrible.

Since I had a unix box with four identical 7950's on hand (all Sapphire Dual-X HD7950, same undervolted bios, same modest clocks totalling 195h/s on claymore 10 in nofee mode) I decided to have a go at some statistical analysis.
 
Included in the study were F2Pool (of course), SuprNova, Dwarfpool and Flypool.. Unfortunately I wanted to keep all stats from the same rig or else coinmine.pl would have been a target of mine also. Maybe down the road.

Of note during my sample over 4 days this week, day 2 posed a bit of a glitch or DDoS attack on Dwarfpool's stats front end but all shares were counted and awarded.

After day 2 I switched to ssl encryption for Flypool and Suprnova since they offered it. Since I was in nofee mode this posed little to no difference in coin output.

https://s30.postimg.org/7drf44qa9/screener.png

*RESULTS AS FOLLOWS*

The basic result (as an F2Pool fanboy) was fairly surprising:

SuprNova daily average output over 4 day sample: 0.01725
F2Pool daily average output over 4 day sample:     0.018209995
Dwarfpool daily average output over 4 day sample: 0.0179318525
FlyPool daily average output over 4 day sample:    0.01968

This was for an consistent average hashrate of 195H/s

Dwarfpool roared ahead early on, took me completely off kilter but had long spells with no blocks found on day 3 and then at the end of day 4 which substantially lowered its average and allowed FlyPool to take the crown overall.

Daily output was as follows with top performers of the day in bold:


        Day1                Day2                Day3               Day4           

Supr    0.0167             0.0184            0.0169             0.017     
Fly      0.01944           0.02004           0.0202             0.01904
F2       0.01913016      0.0170826       0.01834198      0.01828524
Dwarf   0.02102721      0.02075111      0.01224441      0.01770468


Suprnova, as I have commented in previous posts.. continued to substantially under perform. This time during the whole 4 day sample period coming out on bottom by a fair margin. Shortly after switching to Suprnova's encrypted ssl server there was a 15 minute span where I was not connecting to the server for some reason, slightly lower (than the usual low) output can be noted at 68h sample period. Otherwise up time was 100% for this pool and all others.

F2pool wasen't ultimately as superior as I had thought. It was consistent however. Dwarf was a serious contender, not sure if it was plain bad luck that the DDoS and 5-6 hour periods without a block happened during this 4 day span but I did email the admin with some questions and received a prompt and reasonably helpful reply. I would certainly be willing to give Dwarfpool another try in the future.

If I were to do this over again from the start I would have used Suprnova's anonymous service instead of creating an account and logging in to check stats periodically. At this point I think I have lost enough profits with them though, and will no longer point any of my miners at those servers.

Below are samples from about every 4-9 hour ranges. The hour periods aren't quite accurate but what is 100% accurate (and most important) is that the samples were taken from all four pools at the exact (within a few minutes) same time.

Complete results below:


I captured screen shots during each sample period as available for viewing/confirmation here: https://mega.nz/#F!LJhCnbzC!OXdEiM08UK8_7FFeBQHwmw (https://mega.nz/#F!LJhCnbzC!OXdEiM08UK8_7FFeBQHwmw).

I hope this is helpful for some of the ZEC miners out there. I welcome any comments/criticism as I tried to be as non partial and scientific about this as possible.

Interesting post.  Thanks for the data.  I think a longer time horizon would be needed to adequately compare pools with PPLNS.  Also, did you compare fees?

Do the test for 30 days.

But even 30 days does not mean that much .

F2pool was crushed by Kano for months on end in btc.

Now f2pool is crushing Kano for 75 days in a row.

I kind of spilt my hash around.

LOL if you want to commission that kind of case study I may be game.. otherwise someone with some real api experience is welcome to take the reins on this one.

I think 4 days is reasonable, as the calculated average when compared to each days result can attest for.

I agree more is always better in this case but the law of diminishing returns must be considered here when attempting to achieve a decent mean between pools.

F2pool is 4% fee unfortunately and the others 1%-2%? It's not enough of a difference to make any waves but a definite oversight on my part. Thanks for shedding light on this point.

yvw all, I know I have taken/learned enough from others' posts that I'm happy to put some work into giving back for a change.

Excellent work, many thanks for taking the time to do this.

f2pool has dropped fees to 2 %


but they still take the transaction fees

are you sure about that? their payout page at https://www.f2pool.com/help (https://www.f2pool.com/help) still lists 4% and BCT thread announces no plans to change.

i guess every little bit helps but zcash being such a new coin the fees don't surmount to quite so much just yet.

BTC and ETH pools it's certainly a factor for me, ETH not so much due to age but the processing of complex contract code with blocks.

My bad they dropped BTC to  2%



https://i.imgur.com/IKbrISQ.png

Thank you very much for doing this. I was under the impression that Flypool was the worst, surprised at the results.

this is an awesome project.  Dont need API skillz, maybe report findings every week.  Of course there is a lot of variability, but its still useful.  I'd like to see other pools too !  Any volunteers to put one GPU on a pool, and report earnings ?  I could put those up in a website, I had a project like this in mind, cryptopools.online (nothing is there yet), but didnt have enough gpu's/time to do it.

Well done, but unfortunately 4 days is meaningless. A new 4 day test period would reveal completely different results.

Thanks for posting nonetheless.

Nicely done ;)

Thanks

Flypool pays way better and consistent than dwarfpool for me. I gave dwarfpool a chance because my hashrate shows up faster on it, but that does not matter it lags way behind flypool.

F2Pool is PPS ? . for ZEC ?

coinotron is one of the first pools of it's kind .
 https://coinotron.com

it also has PPS ZEC

i know i know
Antpool has PPS zec no fee now

ID rather mine PPS and don't care about the fee . and my miners what i have that are running, run for days before I reboot them manually.

i wouldn't call it meaningless at all.

gtk about antpool.

i think a community database reflecting pay per hashrate is a great idea but having reputable contributions would be difficult to authenticate.

i'll see if i'll have another go sometime down the road.

Love the post! Thanks a lot for sharing. I'm just comparing Fly to Dwarf -only 1 year later :) I'm happy with Flypool but I wanted to cross-check with some others to make sure I don't loose coins...

Luckpool (works kinda like solo pool) not many miners so no luck after 10 hours of mining -quit
Coinfoundry -having connection issues with EWBF; so quit
Sexy.pool -only 7 low hr miners -quit

https://miningpoollists.com/zcash/
https://poolstats.org/ZEC