Excluding a value, it allows me to know what it will be the result without it so with this alternative result, I have a new perspective and from here, I can plan different actions to improve the performance. Alternatively, this article Power BI: exclude the filter selection from the calculation explains another way to do it. I have this table:

I will create a measure to calculate in percentage the valid tickets:

Now what will be the result without the specific group, for that, I will create a new measure:

CALCULATE(CALCULATE(COUNT('table1'[argument1]),'table1'[argument2]="yes")/COUNT('table1'[argument1]),
'table1'[argument3]<>SELECTEDVALUE('table1'[argument3]))

NOTE: if I have another table like that, replace SELECTEDVALUE('table1'[argument3]) by SELECTEDVALUE('table2'[argument3]) and importantly, it should have no relationship with the other table

I will create a visual table:

I can see that if I exclude “db”, the result is better and in the contrary, by excluding “unix”, the result is worst so based on that, I know that I need to take actions in the “db” group to improve my performance.

To have a better view, I will create a slicer, a clustered bar chart and a card but before, I will create another measure:

Since I don’t care about negative values because those are groups that don’t need any actions for improvement, I will remove them and I will active the “data labels”:

The data showing in the chart, it is not an useful information because it is showing me the difference and I prefer to have the alternative result. In “data labels -> value”, click on the arrow to select “measure 2”:

The final result:

And if I select for instance “db” in the filter: