Comment as seen on Hacker News (https://news.ycombinator.com/item?id=32093773):
Some Mining Pool Operators (MPO -- or simply miner) offer transaction acceleration services in Bitcoin where transaction issuers can pay miners off-chain (privately) to get their transactions included ASAP, frontrunning other transactions in the queue. Examples of such services include [3; 4; 5]
Then, after publishing the paper "Selfish & opaque transaction ordering in the Bitcoin blockchain: the case for chain neutrality,"[1 and 6], which shows that MPOs deviate from some common assumptions in Bitcoin, I decided to investigate further some of the transactions accelerated by MPOs. So I did a bunch of empirical experiments and noticed that these services work pretty well.*
I also noticed that some Bitcoin transactions include arbitrary data through the OP_Return opcode, which enables applications to be built on top of the Bitcoin blockchain. For example, the Omni Layer (https://omniexplorer.info) allows users to issue transactions that send USDT to other Omni addresses. However, on the Bitcoin blockchain, the transaction input/output is worth only a few dollars, making it hard for transaction monitoring systems to track the actual value of transactions.
For example, the following txid "ba05a61af97424b340358c2d97080236767ee5b9811d90773f86c35b1ce7f022" sends 13 million USDT through Omni but in the Bitcoin it shows an output of 27,072.48 USD where 27,058.87 USD is a change and a fee of 1 USD. We can retrieve the Omni value by parsing the OP_Return tx output from Bitcoin tx data.
The PDF file (see [2]) shows a draft of a preliminary analysis where for all 313,575,387 transactions issued and mined in Bitcoin from Jan 2018 to Dec 2020, 42,832,713 (13.66%) contain OP_Return opcode, where:
  • 17,993,300 txs belong to Omni Layer Protocol (an average of ~111 txs per block), representing 5.74% of all issued txs and 42% of all OP_Return opcode transactions. Of all Omni txs, 97.27% belong to the Tether USDT token.
  • From our set of inferred to be accelerated transactions (following methodology in [1]), we have that 1805 OP_RETURN txs were accelerated. From this set, 1740 accelerated txs belong to Omni Layer Protocol. 1739 txs belong to Tether token (USDT) and 1 to Omni token.
Question: I understand users would want to accelerate the inclusion of transactions in DeFi (e.g., in Ethereum), but why would someone want to accelerate Bitcoin transactions? I'm still struggling to find reasons why someone would want to pay for the acceleration of Bitcoin transactions.
Thank you in advance for your time and any insights you have!
References
  • These results will be available in a new paper to be submitted soon!
reply