In this comparison, I will have a look at BitCoin Mining Pools, which you can use to create or mint fresh BitCoins. I had a look at BTC Guild by eleuthria, Horrible Horrendous Terrible Tremendous by Fireduck and last but not least BitMinter by DrHaribo.
Hashing Rate and Difficulty
In order to conduct the test I tried to re-create the same set of parameters with each pool. The 24h testing period, the same amount of hashing rate, the same network difficulty and I even tried to factor in individual payout features of the pools.
PPLNS vs PPS
As variance is higher in slower Pay Per Last N Shares (PPLNS) minting pools, because they solve blocks at a very slower pace than truly huge pools, I gave such smaller pools a bit more time, for the last n shares to get accounted for. e.g. 27h for BitMinter.
Concentration of Power
in fact, most BitCoin pools are comparable to each other, except BTC Guild because it is so huge and harnesses over 37% of the BitCoin's network's hashing power, at least at the time of writing.
Problems that arise from this do not only affect comparable parameters but do also pose a thread to the independence of the BitCoin Blockchain. The Blockchain a public and openly verified reference of past BitCoin transactions. People speak of such a threat as a 51% attack, but as one BitCoin forum member put it correctly:
An attacker with only 40% of the hashpower can reverse six confirmations with a success rate of ~50%.
You can calculate the probability of a successful 40% attack at xiph.org: https://people.xiph.org/~greg/attack_success.html
I went into the Pools with 1.8 GigaHash per sec (GH/s) and these are the results of mining with such computing power over the course of 24 to 30 hours:
BitMinter in 24h (+3h) of pure PPLNS mining ฿ 0.1253 + ฿ 0.016 in NameCoins
BTC Guild in 24h of combined PPS and PPLNS mining ฿ 0.0957 + ฿ 0.01 in NameCoins
HHTT in 24h of pure PPLNS mining ฿ 0.0073
I let my miners hash away in each pool 12 hours longer after the initial 24 hours in order to let the PPLNS shares settle in.
The results now were even more clearly visible:
BitMinter ฿ 0.1616 (excl. NameCoins)
BTC Guild ฿ 0.1127 (excl. NameCoins)
HHTT ฿ 0.0073
Being the biggest pool of them all, the almighty BTC Guild takes a 2-5% mining fee. I chose to mine combined with their Pay Per Share and PPLNS system. This would eredicate variance as much as possible for the purpose of this test. After I stopped my Bitcoin miners, I waited for 10 hours and re-visited my stats dashboard in the pool's user front end. It displayed ฿ 0.1227 including gains from integrated NameCoin mining.
The results of the BitMinter Bitcoin mining pool, which took 0% fees at the time of testing, reached an excellent score of ฿ 0.1253 - already after 24 hours of testing time. (please note: I think BitMinter now has introduced a small mining fee of 1%)
I looked at the results again 10 hours after stopping my miners to make sure all PPLNS were accounted for and results were even more exciting.
A total in gains of ฿ 0.18 including Namecoin. That is almost 50% more than what the Guild yielded.
Of course this can and probably is most probably owed to luck. Luck is finding the correct 0-bit hashes in a block, but it is a remarkable result nonetheless.
NameCoin production was also higher than at the Guild. May it be owed to the fact that I could not connect all miners to merged mining with the Guild - probably because its pool owner is phasing out NameCoin mining. The Guild is still a heck of a pool with a great contributer to the scene.
The last of the three pools I tested was the Horrible Horrendous TerribleTremendous Mining Pool.
At the time of testing, it took a fee of at least 2% up to 10%. FireDuck switched his BitCoin Mining Pool to PPLNS just the other day. The first test results were throtteling in; and very meagerly so: only ฿ 0.007 after 12 hours of mining.
Furthermore I had no idea how much more coinage were on its way to me, due to missing statistics and no transparency. I did not know which shares are receiving payment in what shifts or whether shifts are being used at all.
Intransparency led to the abortion of the test after 12 hours. I just couldn't justify 1,5 Kilowatt of power usage into a big void of unknown. After another wait of 15 hours there had been no change, my hhtt account still showed only 0.007 btc. Hence this pool really sucked. Sorry Fireduck, nothing personal.
It certainly is not always the biggest BitCoin pool which is the best.
I wouldn't call any BitCoin Mining Pool a clear winner, because variance and other factors were too unpredictable randomisers in my test.
In my opinion, the correct conclusion to draw from these tests would be that there is no compelling reason on why not to mine into a smaller pool, especially if you do plan to stay at a pool over a longer period of time. It certainly is not always the biggest BitCoin pool which is the best and you would do BitCoin Mining a service by diversifying hashing power and block-finding chances between BitCoin pools.
Out of the three tested bitcoin pools I would rate BitMinter the highest and give it a special recommendation for its truly excellent Java Client, which even surpasses the excellent GUIMiner, as well as for the outstanding transparency regarding shifts, block and luck charting.
If you liked this review and plan to use a great BitCoin exchange or BitCoin advertising network, why don't show your appreciation by using my affiliate link for CoinURL or even BitFinex or as of 2016, even more recommended BitMex.