Forum Discussion

Anonymous's avatar
Anonymous
Not applicable
4 years ago
Solved

Predictions based on past performance

I have a data set containing player performance for the last year, which looks like below. In this sport, assume there are 3 types of matches, so the "Type" indicates which type it is. I've taken X, ...
Need Help
Show and Tell
Tips and Tricks
  • Greg_Deckler's avatar
    Greg_Deckler
    4 years ago

    Anonymous Yeah, you'll have syntax errors when you wing it, there was an offending comma. Try:

     

    Measure Total Points =
  VAR __Player = MAX('Table'[Player])
  VAR __Team = MAX('Table'[Team])
  VAR __Table = 
    ADDCOLUMNS(
      ADDCOLUMNS(
        SUMMARIZE('Table',[Type]),
        "__Team",__Team,
        "__PPM",[Points/Match]),
        "__FutureGames", MAXX(FILTER('MatchesTableNextYear',[Team]=__Team && [Attribute]=[Type]),[Value])
      ),
      "__Points",[__PPM] * [__FutureGames]
    )
RETURN
  SUMX(__Table,[__Points])

     

    It's that last comma just above the first line you get the squilly red line under.

Greg_Deckler's avatar
Greg_Deckler
Icon for Community Champion rankCommunity Champion
4 years ago

Anonymous Assuming that you have that as a table visualization? First, you will need to unpivot your type columns (X, Y and Z) in your future matches table and then maybe a measure like:

Measure Total Points =
  VAR __Team = MAX('Table'[Team])
  VAR __Type = MAX('Table'[Type])
  VAR __PointsPerMatch = [Points/Match] //assumes a measure
  VAR __TotalMatches = MAXX(FILTER('MatchesTableNextYear',[Team]=__Team && [Attribute]=__Type),[Value])
RETURN
  __PointsPerMatch * __TotalMatches

 

Anonymous's avatar
Anonymous
Not applicable
4 years ago

Greg_Deckler , thanks for the reply.

 

This seem to work if I then display this on a table with 'Type' as a column:

 

The calculated values are fine.

However, if I take out 'Type' column (ultimately I want to know each player's total predicted points), then the summation seems to give me an incorrect value:

 

For instance PlayerA's total should be 500 + 112.5 + 225 = 837.50.

The resulting value, 252.40, doesn't even seem to be the average (which should be 837.5 / 3 = 279.16), so wonder what's happening here.

 

 

 

  • Greg_Deckler's avatar
    Greg_Deckler
    Icon for Community Champion rankCommunity Champion
    4 years ago

    Anonymous So what about this?

    Measure Total Points =
  VAR __Player = MAX('Table'[Player])
  VAR __Team = MAX('Table'[Team])
  VAR __Table = 
    ADDCOLUMNS(
      ADDCOLUMNS(
        SUMMARIZE('Table',[Type]),
        "__Team",__Team,
        "__PPM",[Points/Match]),
        "__FutureGames", MAXX(FILTER('MatchesTableNextYear',[Team]=__Team && [Attribute]=[Type]),[Value]),
      ),
      "__Points",[__PPM] * [__FutureGames]
    )
RETURN
  SUMX(__Table,[__Points])
    • Anonymous's avatar
      Anonymous
      Not applicable
      4 years ago

      Greg_Deckler  This doesn't seem to be a valid formula, starting at Line 11.

       

      • Greg_Deckler's avatar
        Greg_Deckler
        Icon for Community Champion rankCommunity Champion
        4 years ago

        Anonymous Yeah, you'll have syntax errors when you wing it, there was an offending comma. Try:

         

        Measure Total Points =
  VAR __Player = MAX('Table'[Player])
  VAR __Team = MAX('Table'[Team])
  VAR __Table = 
    ADDCOLUMNS(
      ADDCOLUMNS(
        SUMMARIZE('Table',[Type]),
        "__Team",__Team,
        "__PPM",[Points/Match]),
        "__FutureGames", MAXX(FILTER('MatchesTableNextYear',[Team]=__Team && [Attribute]=[Type]),[Value])
      ),
      "__Points",[__PPM] * [__FutureGames]
    )
RETURN
  SUMX(__Table,[__Points])

         

        It's that last comma just above the first line you get the squilly red line under.