Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

KPI for arb efficiency #213

Open
hardyjosh opened this issue Sep 21, 2024 · 4 comments
Open

KPI for arb efficiency #213

hardyjosh opened this issue Sep 21, 2024 · 4 comments

Comments

@hardyjosh
Copy link
Contributor

hardyjosh commented Sep 21, 2024

Per an idea by @thedavidmeister, we believe a decent proxy to track how efficient the bot is by looking at the bounty relative to the size of the trade.

Smaller bounties should mean more efficient.

Edit:

effecieny dashboard has been created on hyperdx: https://www.hyperdx.io/dashboards/66eef48472f83c063402ee21
where it shows % of tx gas cost/total bounty, the charts can show any selected timeframe/granularity

@thedavidmeister
Copy link
Contributor

specifically we should be able to calculate the breakeven for the gas for any arb by looking at the transactions after the fact, and compare it to how much revenue the bot made from that arb

if the revenue is much larger than the breakeven on the arb then it's a good chance that the arb was simply inefficient somehow and the bot could have cleared it more efficiently

@thedavidmeister
Copy link
Contributor

this was put into review without it being clear what is being reviewed, what the acceptance criteria are or who is doing it

there's the efficiency dash linked in the OP edit, but it doesn't show anything for me, i think i'm doing it wrong somehow 😅

Screenshot 2024-11-29 at 18 00 10

this is a nice start assuming it works, but we did say we want a separate process that can calculate the KPIs from onchain data so that it can be run independantly of the bots being operational or their logs needing to be accessed directly and stored indefinitely for historical data

@thedavidmeister
Copy link
Contributor

@epappas FYI you should be across this i think

@rouzwelt
Copy link
Collaborator

rouzwelt commented Dec 3, 2024

this was put into review without it being clear what is being reviewed, what the acceptance criteria are or who is doing it

there's the efficiency dash linked in the OP edit, but it doesn't show anything for me, i think i'm doing it wrong somehow 😅

Screenshot 2024-11-29 at 18 00 10 this is a nice start assuming it works, but we did say we want a separate process that can calculate the KPIs from onchain data so that it can be run independantly of the bots being operational or their logs needing to be accessed directly and stored indefinitely for historical data

@thedavidmeister you need to select the time frame and granularity on top right corner of the page, the default timeframe and granularity when you enter the dashboard is last hour and auto granularity, however you need to set it to longer timeframe to get a better picture, granularity is like sample rate, the lower you select, the more point there will be on the chart, the data for each granularity level is avg of all the data in that granularity level, for example granularity of 12 hours, means the value it shows is the avg of all the data in that 12 hours period , like below:
Screenshot 2024-12-03 at 9 09 18 PM
here it is the data from past 7 days and granularity is set to 12 hours

Screenshot 2024-12-03 at 9 11 45 PM here it is the data from past 7 days and granularity is set to 1 hour,

this is percentage of tx actual gas cost / total bounty, total bounty meaning net profit + tx actual gas cost.
when you hover over each data point, it will show the percentage and also when you click it will take you to the eaxct log for that data point, ofc as it is clear, the higher the percentage, the better, meaning there were lower bounty for bot and the gas cost covered more percenatge of the total bounty, and lower means less effiecient

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants