You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
So we should treat all 3 as same authority (the 3rd one being the average between some electrum servers, which is the only source we have in master branch now). But instead of doing an average of the 3, we should probably do an average of the 2 closest ones, to prevent outliers. Example for this algorithm is here: c5530e2 (in particular, c5530e2#diff-334c23b77817534a6745af372d0f1f2b4512ef053d679909a5e8471ae278c301R1096 )
The text was updated successfully, but these errors were encountered:
I noticed this in many occassions. An example was that I was given about 200 sat/vB while mempool.space was suggesting a range around [50-70] at most.
Maybe we have wrong a units conversion issue? In case we don't, we should probably mitigate this issue by having 2 sources more of estimation:
So we should treat all 3 as same authority (the 3rd one being the average between some electrum servers, which is the only source we have in master branch now). But instead of doing an average of the 3, we should probably do an average of the 2 closest ones, to prevent outliers. Example for this algorithm is here: c5530e2 (in particular, c5530e2#diff-334c23b77817534a6745af372d0f1f2b4512ef053d679909a5e8471ae278c301R1096 )
The text was updated successfully, but these errors were encountered: