If we're going with the sporting paradigm...
Every game needs to have a unique ID. Then you have a table of customers, each also with a unique ID. Then you have a table that describes who paid for what. A table with two columns of ID's, customerID and gameID. This is what is called normalization.
So, in your joining table you might have CUSTOMER ID 001, who have paid for games 001, 003, and 005. Here is the customer table:
.------------------------------.
| customer_id | customer_name |
|------------------------------|
| 001 | SPM, Inc. |
| 002 | Stack Overflow |
'------------------------------'
Here is the game table:
.---------------------------------.
| game_id | description |
|---------------------------------|
| 001 | Giants v. Red Sox | |
| 002 | Blah v. Yada |
| 003 | Vader v. Kenobi |
| 004 | Romney v. Obama |
| 005 | Roth v. Hagar |
'---------------------------------'
Here is the table that corresponds to who paid for what:
.-----------------------------.
| customer_id | game_id |
|-----------------------------|
| 001 | 001 |
| 001 | 003 |
| 001 | 005 |
| 002 | 002 |
| 002 | 005 |
'-----------------------------'
Notice how the ID's in the last table are not unique.