Why is my table aggregates incorrect but the total is correct?

[Bokazoit]

My datamodel:

The active join is between DimDato - my date table and the coloumn [Forventet afsluttet Dato]

My measure:

Count = 

VAR MinDate = CALCULATE(MIN(DimDato[Dato]),ALLSELECTED(DimDato[Dato]))

RETURN

CALCULATE(
        DISTINCTCOUNT(FactArvesager[ArvesagsKey]),
        FactArvesager[Bostarts Dato] < MinDate &&
    (FactArvesager[Realiseretarv Dato] = BLANK() || FactArvesager[Realiseretarv Dato] >= MinDate)
    )

Here is the odd part. When I build my measure everything is fine. it counts correctly and the total is correct

But if I add a date from my Date dimension (DimDato) things start to look weird:

For some reason it adds the BoNummer (MemberIds) for 62026,62027 & 62028

But the total is correct, and according to my measure, they should not be counted. If I add the MinDate to see what it calculates it does though make sense:

But the purpose of adding the variable to  my measure, this part:

VAR MinDate = CALCULATE(MIN(DimDato[Dato]),ALLSELECTED(DimDato[Dato]))

is to get the minmum date selected in the date slicer, but that seems not to be the case.

I simply can not figure it out. The key is the date variable that is suppose to be that same value for every row but apparently is not

[DataInsights]

@Bokazoit,

Try changing the four bidirectional relationships to unidirectional. Bidirectional relationships can cause numerous isues.

[Bokazoit]

Sorry but it is a questionary, so that is actually a model I got in here and will not work without