Theory behind "blind rank these 5 NBA players"-type games? What is the relevant terminology and is the probability of success known?

So a common format for sport content creators is "blind rank these 5 things." So 5 names are given one by one, and each time the you must choose a slot 1-5 for that name. You cannot rearrange the names once they are placed. So if you place a name at 1 and then Michael Jordan pops up later, you'd be forced to put MJ lower in the list and end up with a bad ranking. Framing it mathematically, say the (n,k) version of this game is to start with a list of numbers 1-n. k numbers will be drawn without replacement from 1-n and given to you one by one. For each number you are given, you must put it in a slot 1-k. You win if in the end, the numbers in the slot are in increasing order. 1) What strategy maximizes the probability of winning and what is the resulting probability in terms of n and k? I feel like a greedy approach makes sense. Given a number m, choose slot i from 1-k such that i/k is close to m/n. Once numbers are already placed, find the gap it fits in and then choose the slot that closest matches the fraction. 2) If instead the goal is to minimize the error (maybe by something like Kendall tau that counts the number of inversions), what is the optimal strategy? I'm sure this topic has been studied before, but I'm not sure what the appropriate language to search for it is.