Two days ago somebody hacked NiceHash and robbed them of USD 60+ million worth of bitcoins. The service provided by this company was unique and designed to gently introduce hundreds of thousands of Bitcoin newbies to the world of crypto currencies with minimum risk and effort. To the best of my knowledge, no other mining marketplace has a similar business model where hash power sellers mine altcoins but are paid directly in BTC.

It seems unlikely that NiceHash will be able to recover from this disaster soon. Perhaps a new company will re-emerge from its ashes and offer the same business model, but will it find enough buyers and sellers to trust it with their money? Until we know more, I have to either sell my GPUs or start looking for an alternative to keep them busy. After briefly considering my options, I decided to keep my rig and join a different pool.

Why Give Up

Since the beginning of December, Bitcoin has been breaking record after record (as I am writing this, the price of Bitcoin briefly exceeded $17,000 on GDAX…). As a result of this, all altcoins are steadily losing value. Unless/until this trend reverses, altcoin mining would be increasingly less profitable, especially if calculating profits in BTC rather than in fiat currency. So maybe it is better to quit while I am ahead and sell my rig?

Why Keep Mining

When I sell my hardware, I will (1) terminate a revenue stream and (2) lose a connection to this corner of the crypto world. If I don’t, I will be motivated to stay interested in mining, and perhaps learn something new and interesting along the way.

Besides, NiceHash was so easy to use that I am pretty certain tens of thousands of people just installed & ran the software without bothering to understand how it worked. Now that NiceHash is gone, these lazy people will immediately ‘go out of business’, at least unless/until they start doing their research and learn how to switch to an alternative pool.

This means that at least for some time ahead, the available hash power would drop a little bit, making it more profitable for those who carry on. And while I can’t do anything about the dropping cost of altcoins against Bitcoin, I can take advantage of the reduced competition. So I said, let’s keep doing this.

Which Coin to Mine

The simplest way to start looking for a NiceHash alternative is to ‘follow the money’. Open WhatToMine.com (or a similar site), enter your details and let it calculate the most profitable altcoin, then look for the website of that coin, download some software and start mining.

While this is a possible strategy, it is rather foolhardy. First and foremost, ‘altcoin profitability’ is a rather vague term. It is derived from the market capitalization, which is itself the number of tokens in circulation multiplied by the cost of the latest price. This number can be skewed very easily, especially for less popular altcoins traded on obscure, illiquid exchanges. It caries no information about the utility of the token itself or its popularity/adoption. Some folks might not care about this and only look at the price, but I don’t want to do that.

So rather than rushing into things, I compiled a list of requirements towards the coin(s) I’d mine.

1. Popularity: it must be a coin that I’ve heard about and which has known history and community behind it
2. Diversity: it must be different in a substantial way to Bitcoin, so that I have things to learn about it & keep it interesting
3. Compatibility: it must use a mining algorithm/miner I am familiar with, so that I don’t have to waste time benchmarking and tweaking
4. Tradeability: it must be listed on reputable, liquid exchanges and I should be able to trade it for either Bitcoin or USD/EUR
5. Security: it must be supported by a hardware wallet, either a Trezor or a Nano Ledger S (ideally, both). I won’t bother with a coin that can only be stored in a web wallet, or in a single kind of wallet software by an unknown developer

Selection Process

After I composed that list, I went through it in a kinda non-linear fashion. Ethereum seems like a wise choice, but I am running Nvidia cards which have relatively poor performance per watt for ETH mining. Besides, there is something about Vitalik Buterin and the way he manages Ethereum and its community that I really don’t like (a gut feeling, nothing that I can argue rationally about). So let’s just say I’m not interested in ETH and move on.

In recent weeks I noticed that most of my NiceHash profits were coming from Equihash mining using EWBF miner. I like the way EWBF works, because I can turn down the power limit of my GPUs all the way down to 60-65% while still retaining good performance. My cards stay cool and almost noiseless. So let’s see which alts use Equihash as PoW. Here are some: Bitcoin Gold (BTG), Monero (XMR) and Zcash (ZEC).

I dismissed BTG outright, because I strongly dislike the many ways the project was f***ed up from the beginning, and how its mismanagement caused many people to lose their Bitcoins.

Monero is a token I like a lot. They have a good community and the GUI client is one of the better developed pieces of software (I have a XMR node running on my mining rig just for fun). But the main issue I have with them is that there are no hardware wallets that presently support XMR. There is a community effort to build such a product but (1) it is very far from being finished and (2) I don’t want to spend more money on new and untested hardware from an unknown manufacturer when I have two wallets already.

Like Monero, ZEC is designed with special considerations towards user privacy, and this is something I’d be willing to learn more about. Both Trezor and Ledger Nano S support ZEC. Also, I can trade it for BTC and/or fiat on Kraken, and convert it to other alts via ShapeShift.io and Exodus (beta). The last thing that tipped scales in favor of Zcash was this medium.com article which arrived in my mailbox just as I was doing my research. It sums up succinctly all important similarities and differences between Zcash and Bitcoin.

ZCash It Is, Then

All it remained for me to do was to find a mining pool. I thought this would be an easy task but I was proven wrong.

From lurking in Slack channels and internet fora I was aware of the existence of several GPU mining pools: Nanopool, Flypool, Suprnova and Minergate.

Update (10. Dec, 10:30pm) After I wrote this, I came across another pool called DwarfPool.

Update (13. Dec. 6:30pm) After testing, I do not recommend DwarfPool. Broken mining stats & charts, slow payout (2 days delay), awful and uncooperative support. Stay away.

I naïvely thought that due to market forces profitability among these would be similar if not identical, and that the differences would be strictly cosmetic. But after deciding to look into this further, I quickly discovered multiple threads on reddit and bitcointalk where people were claiming big differences in yield (on the order of 20-30%). Some folks went as far as to say that certain pools were stealing user hash rate. I am not linking to such discussions on purpose, because I am unable to confirm from personal experience if there is any truth to them.

Comparing Pool Yield

One might think that the best way to prove these claims right or wrong would be to mine briefly on each pool and compare results. But in practice this task is not simple. Testing pool performance sequentially (e.g. mining for 1 day then switching) introduces errors due to hash rate fluctuations that can easily mask yield variance. The good way to do it would be to mine on all pools in parallel but for that I’d need more hardware than I have.

There is also an inherent problem caused by the way different pools handle participation rewards. NiceHash used a simple scheme called PPS (Pay per Share) which is ‘short-term fair’ — as long as your GPU sends shares, you get paid. This makes sense in a situation where you are getting paid for the raw hashing power you’re providing.

However most pools that mine a given currency use a different scheme called PPLNS (Pay per Last N Shares). Your income in a PPLNS mining pool depends on the overall luck of the forum and takes into account your past and present participation. For that reason, short trial runs are not very practical, because you will not be able to enter ‘steady state’ where you achieve maximum profit.

I decided to try Nanopool first, and liked the user interface a lot. There is also a very, very beautiful and functional monitoring app for Android called Nanopool Monitor by Amber Labs. For approx. 24 hours, I was able to mine a total of 0.0215 ZEC.

Because many people claimed that Flypool was more profitable than Nanopool, I then switched over. Sadly, I think the Flypool Zcash website had inferior interface, and while Amber Labs also have a Flypool Zcash Monitor app with identical functions, I found the interface and color scheme much less appealing.

As I am writing this, I am still ~12 hours into my Flypool experiment. So far I see I have earned less ZEC than what I got with Nanopool by the same time, but that may be an artifact of the different PPLNS implementation. To be honest, I will be happy to find sufficient reasons so switch back to Nanopool.

Update (10. Dec, 10:30pm) Flypool seems substantially more profitable. Nanopool was paying me ~0.01 ZEC every ~13 hours, whereas Flypool is paying ~0.01 ZEC every ~9 hours. I also tried DwarfPool for 12 hours. Its website is really unintuitive, charts are downright broken. It employs a different sharing scheme called HBPPS (hour based pay per share) which is supposed to be more fair but distribution of profits is strange and the pool pays twice as little, six times as slow as Flypool (probably because hash power is low and enough to find only 2-3 blocks per day). Stay away from DwarfPool.

I briefly tested Minergate, but the reported speed and profit projections made no sense to me and I shut it down after about 2 hours. I may or may not have time to also test Suprnova.cc; if I reach satisfactory results with Flypool or Nanopool, I plan to stay on one of these for the time being. But wait… what IS a satisfactory result?

On the Small Matter of Profitability

I’ve written this elsewhere on this blog — I honestly can’t figure out the profit model of people who pay with Bitcoins to mine altcoins. Especially during periods of strong Bitcoin growth (so basically every day since mid-September) Bitcoin-denominated yield from altcoins strongly declines. On the other hand, the rising price of Bitcoin also seems to float the boats of altcoins, and their price in USD also rises. These two forces somehow balance each other, and my fiat-denominated yield seems more or less regular.

But is a steady profit in fiat good in times when fiat sharply depreciates against Bitcoin? Even if I don’t convert ZEC to fiat, ZEC still has steadily depreciated by about 4 times against Bitcoin since mid-September (see chart). This depreciation goes so much faster than the BTC yield depreciation from my NiceHash payouts that I used to complain about in my earnings reports!

Sadly, I am not intellectually equipped to understand this and select the best course. For the short term, I have decided to keep the ZEC I’ve mined. If it starts to quickly depreciate, I will sell it for fiat. If by any chance it retains value during an inevitable Bitcoin price correction, I will use it to buy BTC.