3 ms·
This seems like an extreme example of "good in theory" overriding "good in practice". The multi algorithm outperforms single, so run multi, then greedily swap
by hedora 9y ago
This seems like an extreme example of "good in theory" overriding "good in practice".
The multi algorithm outperforms single, so run multi, then greedily swap students at random until there are none that got each other's higher ranked choice.
This obviously outperforms multi, and is fair, strategy-resistance be damned.