>>98119618>many of them are strictly vstats only for some reason which is sus.Did you read the image? I'll quote the original post:
>>84412354https://archive.palanq.win/vt/thread/84412043/#84412354I'll assume you are relatively knew for not knowing this as I think it was not posted for the longest time but, in this thread, since Oct 2021 when this tally started being compiled, there were these rules
>1) Streams that start in a day go in the day they started.>2) The day starts at midnight JST>3) The tally is tallied at 1AM JST and any peak after that is discarded >4) the highest peak out of the tally sites is the one usedThis is the original format. Then "Z"s started affecting the tally positions and a change was made
>for the top 3 the default Z filter from vrabi will be used if the Z peak would change podium positionsThen (as you see in the notice in the pic at
>>98119275 and the post at >> 84412354) two things happened:
>vrabi added the ability to fetch the Z filtered value in their API>vstats started manually flagging streamsThat allowed for two more changes
>for all streams to be corrected for Zs, not only the top 3>for streams that get manually demoted from vstats to get manually demoted in the tally as wellAnd that lands to the present day system. The peak is still taken from the highest peak from vrabi, vstats, vtstats (former holostats) and Niji-mado an, in case of Zs, the default filter from vrabi is used and, in case of manual demotion from vstats, that value is used.
It's the close thing to a "fully automated rules based well defined" system there is, except for the manual demotion which is, well, manual.
