Bitcoin and the threat of mining pools

Bitcoin is as decentralized as its least decentralized component. The pools, along with ASIC manufacturers, represent a weak link.

Miners and Their Pool

Miners are at the heart of Bitcoin’s mechanics. They forge consensus with gigawatts of encrypted energy: the “proof of work”.

For energy companies, their industry is a significant source of revenue. Able to set up shop anywhere, miners often consume electricity that would otherwise be wasted. They are a financial crutch to accommodate the influx of intermittent electricity sources such as wind and solar power.

As icing on the cake, miners can quickly step aside during peaks in electricity demand. They can also combat climate change by recycling methane from oil platforms too remote from civilization to be cost-effective to exploit.

This sums up the situation well:

pic.twitter.com/F9MGlgBWrv

— Nicolas Teterel #FixTheFilters (@NTeterel) January 15, 2024

To get back to pools, it’s important to remember that a miner’s job is to run electrons through ASICs that crunch using the SHA-256 algorithm. A miner generates hashes all day, and that’s all…

But there was a time when a miner also had to:

-Operate a node to receive blocks.
-Construct the blocks themselves by selecting transactions.
-Create the root of the Merkle tree for all these transactions.
-Hash to try to find a valid hash (“find a block of transactions”).

Faced with the proliferation of miners, they quickly grouped into “pools”. The first was “slush pool”, which is now Braiins.

These are cooperatives that allow miners to pool their computational power to smooth their income.

Since then, miners have been content to hash, leaving pools the responsibility of constructing blocks and selecting transactions.

A Pool to Rule Them All

Unfortunately, miners tend to cluster in a handful of pools to achieve the most regular payments possible.

However, a thousand pools would not be viable either. There is a limit. But the fact is that just two pools (Foundry USA and Antpool) today conglomerate nearly 55% of the hashrate.

Given that Antpool belonged to the manufacturer Bitmain before 2021 and that the fourth-largest pool ViaBTC (11%) is also very close to the ASIC’s manufacturer.

In summary, only a small handful of pools have the privilege of selecting transactions:

Deux pools contrôlent la sélection de 55 % des transactions #bitcoin . pic.twitter.com/ciBStxtA9i

— Nicolas Teterel #FixTheFilters (@NTeterel) January 10, 2024

From this point of view, Bitcoin – a payment network meant to be uncensorable – is at the mercy of regulatory capture.

It is much easier for OFAC to impose censorship of certain transactions on a handful of pools rather than thousands of miners scattered around the world.

Not to mention that pools can even self-censor. This was recently the case with F2Pool. Our article on the subject: F2Pool caught red-handed.

Reliance on pools to select transactions also means exposure to cheating. Some of them do not hesitate to sell block space on the sly.

For instance, block 774628 kicked off the trend of “ordinals”. Notice that the transaction fees were only $200 while the surrounding blocks earned $5,000.

In other words, the pool in question potentially deprived its miners of $4,800. Opacity reigns. Miners have no way of knowing if their pool has paid them all the fees.

The FPPS System

This opacity problem is further compounded by the operation of the pool payment system: FPPS (Full Pay Per Share). Almost all pools use this system, which is advantageous for miners. Their work is compensated no matter what.

If a pool runs out of luck and finds fewer blocks than expected, the miner still receives the agreed amount in advance. Miners benefit from very stable cash flows. But there are disadvantages.

The first is that we recently discovered collusion among pools that now use a common fund to smooth out their own revenues. This is troubling for the network’s decentralization. Our article on the subject: A pool cartel revealed.

The second disadvantage is that FPPS pools make a low estimate of the transaction fees they commit to paying. They adjust afterwards and usually keep the excess. But do they really return 100% of the money?…

It’s a real question, especially since the trend of inscribing ordinals garnered a whopping 557 bitcoins in transaction fees last year. And how many bitcoins on the sly?

These fees are becoming less and less trivial. Their share of miners’ revenues is currently only 2% to 4%, but it doubles after each halving.

The issues raised here can be addressed, at least in part, thanks to the second version of Stratum V2, which is the protocol by which miners communicate with pools.

Stratum V2

The main feature of Stratum V2 is to return to miners the privilege of selecting transactions.

Other appreciable features include the encryption of miner/pool communications and the reduction of bandwidth consumption. This is achieved partly by encoding messages in binary and partly by eliminating redundant messages. On stratumprotocol.org we can read:

“Stratum V1’s human-readable JSON-RPC protocol makes messages 2 to 3 times heavier than necessary. Stratum V2’s binary encodings minimize message size, which speeds up communications between miners and pools.”

Moreover, Stratum V1 is not suited to large installations with hundreds or thousands of machines each communicating directly with a pool, resulting in unnecessary wastage of non-negligible amounts of energy.

All in all, Stratum V2 reduces the average message size from about 100 bytes (unencrypted) to 48 bytes (encrypted).

Another advantage: the elimination of empty blocks. Their existence is due to the fact that Stratum V1 does not allow the hash to be sent of the previous block to miners. Pools therefore send a transaction-free block to communicate it quickly.

Why? Because that hash is the one piece of data absolutely necessary for the miner to start hashing. Receiving it before the transaction block offers the opportunity to hash for a few extra seconds. Blocks are sometimes found during this time gap, resulting in empty blocks.

Pools only have to communicate the hash of the last block since, with Stratum V2, it’s the miners who have the responsibility of constructing the blocks. The hash is sent via a dedicated and optimized 32-byte message.

The Ocean Pool Leads the Charge

Luke Dashjr has revived his Eligius pool, closed in 2017, with funding from Jack Dorsey.

“OCEAN solves a problem for bitcoiners that, I think, we all feel: the centralization of pools that could compromise a number of Bitcoin features that are dear to us,” Jack Dorsey, co-founder of Twitter and CEO of Block, stated.

To this end, Ocean will implement Stratum V2 in the course of the year. Miners will then be able to build their own blocks with transactions they have chosen themselves.

Another key feature in line with the ethos of Bitcoin, Ocean is NO KYC (Know Your Customer). Everything is legal as long as the pool does not hold custody of the bitcoins.

Miners are paid directly through the coinbase transaction. Here is, for example, the fifth block mined by Ocean, block 824915.

In this way, the miner is certain to receive 100% of the transaction fees, which is not necessarily the case with the opaque FPPS system.

Another advantage of the Ocean pool: filtering out the “inscriptions” of ordinals and other BRCs:

You can’t filter my inscriptions @ocean_mining! My precious ordinals! REEEEEE pic.twitter.com/Uam797XXSb

— Austin Barnhill (@bh30317) January 4, 2024

The co-founder of the pool (@GrassFedBitcoin) is convinced that “inscriptions” should be considered a ddos attack. His opinion is that it is hazardous to think that the gambling of BRC-20 and other ordinals will disappear by themselves.

“I understand that you can’t really define spam and therefore cannot eliminate it from the blockchain.

But I don’t believe either that, in any other context, anyone would tolerate miners being paid to outright exclude transactions.

Except for the one course, that of literally taking the most lucrative transactions, of course. But if these are of a strange nature that exploits an obvious vulnerability to store data in violation of the intended use case (as inherited from design choices over the years) of the bitcoin blockchain, then I think if miners can include them, it becomes as controversial as taking government money to exclude ‘naughty’ transactions.”

Ultimately, pro-spammers do not account for the fact that bitcoin collapses if we don’t force miners to actively reject the money they can make by attacking the network.

Not everything automatically adjusts just right because of basic economic principles.”

With ASIC manufacturers, pools have always been the weak point in the decentralization of the network. The arrival of a pool like Ocean is undoubtedly good news.