SQL Server 2022

By using db<>fiddle, you agree to license everything you submit by Creative Commons CC0.

CREATE TABLE moneys (
  game_date   DATE,
  casino_id   INT,
  game_id     INT, 
  amount      INT
)
INSERT INTO
  moneys
VALUES
  ('2024-01-01', 1, 1, 1000),
  ('2024-01-01', 1, 4, 2000),
  ('2024-01-01', 1, 2,  300),
  ('2024-01-01', 1, 3, 1000),
  ('2024-01-07', 1, 3,  500),
  ('2024-01-10', 1, 1,  500),
  ('2024-01-11', 1, 1,  100)
7 rows affected

WITH
  lagged AS
(
  SELECT
    *,
    LAG(amount, 1, 0) OVER (
      PARTITION BY casino_id, game_id
          ORDER BY game_date
    )
      AS previous_amount
  FROM
    moneys
)
SELECT
  *,
  SUM(amount - previous_amount) OVER (
    PARTITION BY casino_id
        ORDER BY game_date, game_id
  )
FROM
  lagged


game_date
casino_id
game_id
amount
previous_amount
(No column name)


2024-01-01
1
1
1000
0
1000

2024-01-01
1
2
300
0
1300

2024-01-01
1
3
1000
0
2300

2024-01-01
1
4
2000
0
4300

2024-01-07
1
3
500
1000
3800

2024-01-10
1
1
500
1000
3300

2024-01-11
1
1
100
500
2900

game_date	casino_id	game_id	amount	previous_amount	(No column name)
2024-01-01	1	1	1000	0	1000
2024-01-01	1	2	300	0	1300
2024-01-01	1	3	1000	0	2300
2024-01-01	1	4	2000	0	4300
2024-01-07	1	3	500	1000	3800
2024-01-10	1	1	500	1000	3300
2024-01-11	1	1	100	500	2900