buidl-labs / livepeer-pricing-tool

A tool for enhancing the price visibility in livepeer ecosystem.
http://livepeer-pricingtool.surge.sh/
4 stars 2 forks source link

Fixes + Updates for Orchestrator Statistics Table #15

Open Saumay-Agrawal opened 4 years ago

Saumay-Agrawal commented 4 years ago
prastut commented 4 years ago

Problem Statement

image

Saumay-Agrawal commented 4 years ago

@prastut Here are the columns I think we can keep to make tool pricing oriented as per the problem statement:

  1. Orchestrator Address
  2. Price per pixel
  3. Total Fees earned [to be added]

Do you agree on this? Or any other column should also be retained from the current UI (http://35.223.32.189:3000/)?

prastut commented 4 years ago

@Saumay-Agrawal These are all the columns present on the website:

  1. Address
  2. Activation Round
  3. Last Reward Round
  4. Delegated Stake (LPT/LPTU)
  5. Reward Cut (%)
  6. Fee Share (%)
  7. Price Per Pixel (wei/pixel)
  8. Active
  9. Status

Columns that I want to see:

1. Address:

Important.

What I would more like to see is who the person behind the orchestrator is.

How is Livepeer explorer fetching the metadata for an orchestrator? Eg: https://explorer.livepeer.org/accounts/0xe9e284277648fcdb09b8efc1832c73c09b5ecf59/campaign

Metadata that I would like:

  1. Name (minimum) (wherever possible)
  2. Bio.
  3. Website.

Only the name is supposed to come on the landing page. Rest can be moved to the profile page.

2. Price per Pixel:

This is the crux of the app. What I want to see is whether we can add a tooltip explaining what price per pixel actually means for an end-user. The north star I want to design is that the copy (copywriting) of the app can be read by somebody who is noob to Livepeer but knows enough about markets and video streaming.

3. Total fees earned:

This is my speculation that total fees earned = more people trusting you to transcode = trust in you.

Hence super important.

On that note, I would want to trust the orchestrator who has:

  1. Made the most wealth (if that can be ascertained by total fees)
  2. And is charging the lowest price per pixel for their services.

4. Delegated Stake (LPT/LPTU):

I feel this is important from a network perspective. I have a feeling it represents how much skin in the game this orchestrator has and how many people are staking their money it.

As long as it can represent skin in the game, I want to see that. That decides how long term invested the entity is in the network.

5. Percentage metrics:

  1. Reward Cut (%)
  2. Fee Share (%)

Here I am again out of my Livepeer knowledge comfort zone. These are the metrics I would be analyzing when I analyzing the pricing data:

  1. The most wealth it's made on the network.
  2. The least price I would need to pay.
  3. There should be some reward that they should be charging. Only the big players in the game can possible run a system at 0% commission and those players usually don't care about individual people. I would want to invest in somebody who is charging and is charging a fair amount.

6. Status indicators:

  1. Active
  2. Status

As long as there are some orchestrators that have "non-active" and "non-registered" status. If all of the values are same, probably we can ditch these.

Columns I don't care about:

  1. Activation Round
  2. Last Reward Round

Lense to see the pricing tool's landing page with:

Imagine the times you have bought something and tried to haggle in a marketplace. What all do you need to make a decision? This is the same lense a user will use our tool with.