Network Economics Update - 5/6/18


#1

Livepeer protocol has now been live for 5 days/rounds. Let’s take a look at some of the economics of the network after a few days. It may be helpful to review the initial parameters at launch from last week’s update.

Param Value Comments
Number of Active Transcoders 9 There are currently 10 slots for active transcoders. And 20 candidate slots for visible transcoders. Still room for more candidates - view campaigns in the transcoder campaign category.
Current inflation 155/1000000 round 5.66% annual. This is up by .000003/round over 6 rounds, resulting in 1550 token/round being generated and allocated amongst those who stake LPT per round.
Delegated Stake 16173 LPT This is how much token is staked. The participation target is 50%, or a little over 5M token, but it will take a long time to achieve that through the full MerkleMine distribution + slow token release.
Reward Calls 13 for 17 Active Transcoders have successfully called Reward() 13 out of 17 possible times. Delegators should look for 100% reliability in their transcoder invoking reward, since this is how new LPT gets generated and reallocated, and if the transcoder misses a round, they miss the opportunity, and it doesn’t come back.

Key Takeaway
The protocol is doing what it is supposed to be doing. Incentivizing participation through allocating LPT to those who are staked and invoking the txns on the network. No disasters or surprises during week 1 is a good thing.

Observations

  • One of the big tests during the alpha is whether the protocol can achieve greater than 90% reward calls in the first month. If delegators elect transcoders who are reliable at this task (which is just 1 txn/day), then there is no reason this shouldn’t be achievable. It will be a priority to make this more visible in the UX for delegating.
  • Gas costs remain a challenge. Delegators aren’t required to take any time sensitive actions, so they are encouraged to use low gas costs when MerkleMining, delegating, and claiming, as even if their txn takes a few hours to get confirmed, it’s ok. If a delegator is happy with the transcoder they’ve selected, they only need to average 1 txn every 20 days (and you don’t even have to do this on time - you can do 3 txns after 60 days for example), so if you’re concerned about gas costs, don’t get overactive with the transactions.

Very common usability requests have come in related to delegating. Right now each account can only delegate to one transcoder at a time, and when you are bonded you can not unbond only some of your token - you have to unbond all. As such, right now users can split their token across multiple accounts if they’d like to manage the above, but it will be a high priority for the protocol and UX to support these actions!


#2

Great work Doug, thank you. I don’t target 90% but 100% for my node! :smiley: Couldn’t you implement a “reward score” inside of https://explorer.livepeer.org/transcoders/ ? That would make it easier for users to choose a transcoder to delegate to.


#3

Yes! Check this issue: https://github.com/livepeer/livepeerjs/issues/66

Suggestions as to how to best encapsulate and display this are welcome. Keep in mind the explorer is currently a DApp reading from the blockchain, and not a database, so aggregating past history over many many blocks can be challenging.