question on update function in proc sql

[Xamius32]

Lets say I want to update table A based on a column in table B. Table B will have a list of games (like A, B, C, D) and if those games are in table A, I want to decrease a certain value. if I do

proc sql ;

  update A set value=value-1

  where game in (select game from B)

  ;

quit;

if dataset B has a game listed more than once, will the value decrease more than once? IE if its listed twice, the value in dataset A will decrease by 2.

[Xamius32]

and one more question, if I wanted to do the same thing, but using simply another column in table A, is that possible? When I try it by changing B to A, it says sql cant use the table it is updating to update.

[Tom · Accepted solution]

You could make it execute by adding the UPDATE_POLICY=NONE but why?  If you leave it coded the way it is the update will happen for every row as every value of GAME in dataset A is by definition in dataset A.

[Xamius32]

I realize by you saying that I am looking at this in the wrong way in terms of my goals. Anyways, thx for the help.

[Tom]

No.  It will process each row in A and decide whether that GAME exists in B, if it does it will execute the update.