The runtime between two submissions can vary so wildly, it can put you between 25% and 97% of the solutions ladder. The statistics is useless unless you fix it by running the submission, say, a thousand times and taking the average of it.