Leaderboard

That was my thinking as well. The only problem i see is that the scanner is done for everyone using the same default settings. I can't run the scanner with all the combinations, it's too time consuming.

I was thinking about this, and came to realize that even though the sky-high values appear to be anomalies, in reality, they were not binary events. That is, given that those high values happened, and that they progressively decreased towards the normal-times lower values with each successive cycle, it would make sense to actually leave them in the mean table. That would seem to provide a more relevant mean trend. As for the median chart, I agreee that it gets skewed by those cycles. However, rather than remove them from consideration, there should be some mathemagical method to integrate that information.

That idea would cause some significant problems for how I use the data, so my request would be to please not do it. Automatically excluding cycles within certain time frames would screw up a lot of calculations I make that rely on all of the presented data being available. If you start automatically cutting out a cycle or cycles from every underlying then different parts of the data will be removed for different underlyings making them much harder to compare to each other. I understand that it is a question of either way somebody has to go and manually select or deselect elements of the dataset they want to see or hide, but the default should remain that all the available data is presented unless the user selects otherwise.