Getting a Weighted Average For Selected Rows

[ssc]

I have a small data set, with a few values. I've simplified this as shown below:

Date	Event	Opponent	Set	Player	Attempts	Aces	Errors	ServeRating Total	Serve %
1/9/2022	P1 17	Energy 17	1	5	5	2	1	12	80%
1/9/2022	P1 17	Energy 17	1	45	4	2	0	13	100%
1/9/2022	P1 17	Energy 17	2	5	8	4	0	24	100%
1/9/2022	P1 17	Energy 17	2	45	3	1	0	10	100%
1/15/2022	Classic	Elevation 16 Ascent	1	5	2	0	1	1	50%
1/15/2022	Classic	Elevation 16 Ascent	1	45	2	0	0	3	100%
1/15/2022	Classic	Elevation 16 Ascent	2	5	7	3	0	19	100%
1/15/2022	Classic	Elevation 16 Ascent	2	45	3	0	1	4	67%
1/16/2022	Classic	CSA 16 Navy	1	5	7	1	0	17	100%
1/16/2022	Classic	CSA 16 Navy	1	45	2	0	0	3	100%
1/16/2022	Classic	CSA 16 Navy	2	5	5	1	1	7	80%
1/16/2022	Classic	CSA 16 Navy	2	45	3	0	0	4	100%

My dataset as calulated averages already. For player 5, we can see the first date (1/9/2022), we have 80% and 100%. If I put these in a visual and get the average, I get 90%. However, the calculation is (attempts-errors)/attempts. If I calculate this from the raw columns, I'd get 92% because the 100% value is weighted with 8 attempts to the 5 for the single error.

I have been trying various SUMX () and AverageX() combinations, but I don't quite understand how I can get a measure or column that would give me 

[v-zhangti · Accepted solution]

Hi, @ssc 

You can try the following methods.

Measure = 
VAR _table =
    SUMMARIZE ('Table','Table'[Date],'Table'[Player],
        "DIVIDE",
            DIVIDE ( SUM ( 'Table'[Attempts] ) - SUM ( 'Table'[Errors] ), SUM ( 'Table'[Attempts] ) ) )
RETURN
    SUMX ( _table, [DIVIDE] )

Is this the result you expect?

Best Regards,

Community Support Team _Charlotte

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[ssc]

That does appear to work. I'll test on a larger data set and repost, but can you explain how this works? I'm a little unsure of what is happening.

[ssc]

Lost part of my post. I am looking to understand how I can get a weighted average by date of 92% instead of 90% for a player.