Big differences comparing MKR votes of a poll between the dashboard and vote.makerdao

[cmalfesi]

I have detected an issue comparing MKR votes of a poll between the dashboard and vote.makerdao.
This are the link compared:
1- https://mkr-gov.netlify.app/poll/184 > MKR 37082.35
2- https://vote.makerdao.com/polling-proposal/qmsv7qsmqafardbhun2f9s65nge3xf5oe4we3dswzwbtfy > MKR 26,555.63
You can see that there is a big difference.
Lisandro has checked manually all values of every voter and here I send you a spreadsheet where you can see that our calculation looks good, it matches with the Dashboard.
https://docs.google.com/spreadsheets/d/1y-2RME38AML2gazlGsMDWHC_cRjLDKz_NthYTgyZ8Ec/edit?usp=sharing

[cmalfesi]

Answer from Derek Flossman:
Hi cristian.malfesi and lisandro.corbalan , we identified this to be the result of an edge case that's always existed. What happened was that a whale with 10,526 MKR created a vote proxy with MKR in his cold wallet, but didn't move the MKR into the vote proxy until after the vote ended. https://etherscan.io/address/0x8dd0dc203090591183b7b26203a911e6a220b3bd#tokentxns

I’m meeting with the Devs on Monday to walk-through some solutions, define better messaging and honestly to better understand the details of what happened. (it will either involve us adjusting polling vote power expectations to not include the mkr balance of the cold wallet attached to a vote proxy or some new calculations to account for the above case), I’ll follow up with you on our direction and to help us align on showing the same data. Thanks for reaching out guys.

[cmalfesi]

[cmalfesi]

The last status is: Maker team is checking how to calculate in the right way the MKR in a vote. We are waiting that they inform us if we need to improve or change the formulas.

[cmalfesi]

[cmalfesi]

[cmalfesi]

by derek.flossman:
Attaching a doc describing Ranked Choice in case you guys find it useful for your implementation

https://app.zenhub.com/files/204923920/ee519234-508e-4681-9538-7a559dd2bccd/download