@raghav it seems like at a high level this approach basically is limiting the number of queries to any algorithm. Because you only start querying an algorithm after you have frozen all algos before it.