Need help to calculate average resolution time which excludes top and bottom 1% of values

[Anonymous]

Hi all,
I need to calculate average of resolution time which should exclude 1% extreme values means top and bottom 1% values.
Average of resolution time= (Resolved On Date - Created on Date) divided by total number of ticket count
So if I have 1000 resolved ticket , we should exclude the 10 tickets that have the shortest resolution time and the 10 tickets that have the longest resolution time.

I have tried using RANKX function to calculate top and bottom rank on resolution time(difference between resolved and created) but not able to get it right and also, how do I combine both top and bottom in final query.

TopRANK = RANKX (table,table[ResolutionTimeInHr])

BottomRANK = RANKX (table,table[ResolutionTimeInHr],,ASC)

Top 1% =
var top_find_rank=COUNTROWS (Table) * 0.01

RETURN COUNTROWS(FILTER(Table,Table[TopRank]<=top_find_rank))

I see one more issue with Rank here is when I apply date filter then also the rank value is static. it should change dynamically as and when we choose date slicer

I had referred this link https://www.burningsuit.co.uk/blog/best-and-worst-two-approaches-finding-top-or-bottom-n-percent-usi... but I am still stuck.

It would be great if someone can help or guide on this . Please let me know if anything is not very clear .
Thanks much in advance !

[Anonymous]

@Greg_Deckler  I don't want to disturb original fact table as we have other business logic to be applied on whole data. Is it possible to achieve this without introducing another fact table ?

[Greg_Deckler]

@Anonymous It's a measure. Shouldn't introduce another fact table, just need to use it against your current table. If you post sample data, I can be more specific.

[Anonymous]

@Greg_Deckler My bad just realized its measure in your pbix file , actually I have been trying with tables/columns/measure and it is still throwing an error .
Here is the sample excel file with data  Sample file 
Its throwing an error at Value at the end . It would be great if you can help on this
Thanks much !

[Anonymous]

@Greg_Deckler Its working amazingly . Thanks a lot for all your effort and great work 🙂

[Greg_Deckler · Accepted solution]

@Anonymous TRIMMEAN - Microsoft Power BI Community

[Anonymous]

Thanks very much Greg @Greg_Deckler  !
I am going through the detailed solution and it will take a while for me to understand how the solution is working and apply the same logic in my data model. 
I see here we are using calculated table and not sure if we can get rid of this and do it using just column/measures or we have any simpler solution to this . 
Thanks again for your help !