Network Economics Update - 9/21/18 - 10% Participation and MerkleMine

It’s been awhile - time to bring back the network economics updates!

The Livepeer network is currently in round 1106 which is the 145th round since the launch of the network on May 1st 2018 which was round 961. Below are the latest observations on the state of the Livepeer network economy:

Parameter Value Notes
Total LPT Supply 10,502,704 The total supply of LPT
Total LPT Generated 502,704 The amount of inflationary LPT that has been generated by calls to reward() by transcoders and distributed towards those who stake and delegate their token. This amount is in addition to to the initial supply of 10,000,000 LPT at network launch which is being slowly released over a 36 month period via the MerkleMine and vesting grants
Total LPT Staked 1,138,591 The amount of LPT that is currently staked
Participation Rate 10.84% The percentage of all LPT that is being staked. The target rate is 50% which should take at least 12-18 months to approach due to the slow release of LPT in the MerkleMine
Inflation 0.0572% The current per round inflation rate. Every round that the participation rate is below 50%, the inflation rate will increase by .0003%
Current Mintable Tokens 6005 The amount of LPT that can be minted this round and distributed towards those who contribute work on the network by staking and delegating tokens

Participation Milestone
The big semi-milestone that has recently been surpassed is the 10% participation rate of staked tokens on the network. While the target remains 50%, this milestone was actually passed ahead of predicted schedule, as a large number of token are still sitting unmined, in time-release vesting contracts, or distributed via the MerkleMine into accounts that have likely yet to discover the Livepeer project. The estimation is that approximately 60% of the liquid LPT in the hands of proactive users in the ecosystem has been staked to date, which is a great indication of the willingness to actively participate to help the network. The combination of the continuation of the MerkleMine, continued vesting release, and increasing inflationary LPT distributed to participating users will mean that the 50% participation target will likely be hit sometime in the next 12 months, at which point the network inflation will begin decreasing towards an equilibrium.

MerkleMine Update
The MerkleMine token distribution method has been going on for nearly 5 months now and has picked up steam as the incentives to submit proofs for other users grows with every passing block. The open mining period is 55 days in. Let’s take a look at some of the key numbers:

Parameter Value Notes
Original LPT in MerkleMine 6,343,700 The initial number of LPT to be distributed via MerkleMine
Remaining available LPT in MerkleMine 2,276,031 The number of LPT that are remaining to be distributed via the MerkleMine
Caller Token Allocation for 20 proofs 6.62 LPT The number of LPT a miner receives for submitting 20 proofs at this moment
Gas cost to submit 20 proofs 2,700,000 The amount of gas required to submit 20 proofs

There are a little over 900,000 proofs remaining to be submitted before the MerkleMine comes to a conclusion. With 100,000-200,000 LPT being generated per day, there are likely only a couple weeks remaining, and it’s possible that competition will increase as the remaining LPT available via the MerkleMine becomes scarce in the final days.

At times of low gas prices, we’ve observed that as much as 30% of the txns on the Ethereum network are dedicated to MerkleMining. However during times of more network congestion and higher gas prices, miners slow down and more capacity is available for whatever use cases people are paying higher gas prices for. This mechanism was designed in part to use the Ethereum network responsibly with a smooth gas price impact curve, and spread the distribution out over many months, rather than “clogging the network” for a period of days like many sale or airdrop based mechanics do. And while there is significant demand to use Ethereum for MerkleMining, the resulting usage and ebb-flow of mining in relation to gas prices, represents a nice equilibrium with alternative uses of the network.

One of the competitive techniques that has been observed in the MerkleMine is miner transaction snatching. One aspiring miner submits a bunch of proofs to the MerkleMine contract, but another is monitoring all the pending txns waiting to be confirmed on the blockchain. They observe this pending mining txn, and resubmit the exact same proofs, but with higher gas price. If this new txn is confirmed first, then they receive the LPT and the original miner spends gas but gets nothing.

A number of community members have worked on solutions for this, open sourced mining scripts that try to combat this, and have deployed a number of unique mining contracts. In the end, the mechanism really presents a race to submit the available proofs, and those who value the LPT the most will pay up in terms of gas price in order to win the race. However it’s nice that there’s a number of “consumer friendly” point+click mining tools available that aim to provide access to the network for the casual user who just wants to get a little LPT to start staking and participating in the node election.

Transcoder and Staking Update
Since Livepeer expanded the transcoder pool from 10 to 15 active transcoders, there have been a number of creative and highly proactive transcoder campaigns.

The minimum delegated stake require to break into the active set is currently 14,151 LPT. This is good, in that it provides significant economic security at stake for misbehaving active transcoders, but it’s bad in that it’s quite a high bar for a new entrant to be able to gain their way into the active set. The protocol also currently has a limitation in that LPT bonded towards inactive transcoders doesn’t earn inflationary rewards, so this makes is doubly difficult to build up a stake while not active. We have some protocol update proposals coming down the pipe to address this issue, as well as to more competitively allocate work across the network to available nodes, in the regions that broadcasters are located, at the lowest available prices. Stand by.

In the meantime, early usage of the production network for video encoding remains mostly tests, while we work on the above proposals and Ethereum scalability issues to significantly reduce the overhead costs of claiming and proving work. We’re confident in some of the early designs and will be publishing proposals for the community shortly.

One big update as it relates to economics for delegators and node operators, is that partial unbonding is now available. Nodes can unbond some of their LPT from a transcoder, and withdraw it after the unbonding period ends, without having to unbond all of their LPT. This should help users split LPT across accounts to stake towards multiple transcoders, and increase decentralization in the network.

3 Likes

And here are a couple of more detailed analysis on the MerkleMine and Transcoding/Staking economics from Coinfund and Notation Capital:


1 Like