Most efficient way to query with two keys in either order? - php

Question

I have some data I receive of the following format:
{ gameId: 1, playerId: "john", score: .12 }
{ gameId: 1, playerId: "mary", score: .75 }
{ gameId: 1, playerId: "jane", score: .32 }
{ gameId: 2, playerId: "john", score: .89 }
{ gameId: 2, playerId: "mary", score: .91 }
{ gameId: 2, playerId: "jane", score: .99 }
And I want to expose these endpoints:
GET games => get a list of all games
GET games/{id} => list all the games for the gameId and the average score among all players
GET players => get a list of all games
GET players/{id} => list all the scores for the player and their average score across all games
So setting DBs aside for a moment, I thought of a hash map approach where I basically have two maps:
gameScores[gameId][playerId] = score
playerScores[playerId][gameId] = score
This way I could very efficiently return results as the following:
GET games => array_keys(gameScores)
GET games/{id} => ['average' => avg(gameScores[id]), 'games' => gameScores[id]]
GET players => array_keys(playerScores)
GET players/{id} => ['average' => avg(playerScores[id]), 'games' => playerScores[id]]
This seems like a very efficient way to return results in near O(1) time, but is it too great a drawback to be duplicating the dataset by 2? Imagine if score was some very large object instead.
I'm doing this in PHP, so not using something like a Python tuple solution here, but I feel like this problem is very generalizable (hash map with 2 keys) and I'm wondering if there's a better way to approach this rather than duplicating the hash map for both key orders.
Is using a database the only more optimal approach here? If so, would it be best to enter the data into a single table with those 3 columns, or should I be splitting the tables?

Answer 1

You only need one array and search it with
array_search($search, array_column($array, $field));