The only 3 factors in this scenario are $ cost of the player, their projected points, and their position.
Starting a little more high level, let’s ignore the position piece. A good way to figure out what players to get would be to create a coefficient that would most likely resemble Projected Points / $. You can then sort all players based on this coefficient (according to your APIs) you have the best buys for that week.
Integrating position becomes a little tricky and I don’t have enough time this instant to delve into it
closed - thanks