No set of data in this sort of discussion is going to be perfect, I selected the closest possible. For example, comparing him to the other opening bowlers rather than the change bowlers, and selecting 4-7 because it seemed more relevant to the actual discussion. 3-7 would have worked too but considering how often early wickets fall and no.3 comes out against the new ball I considered it best to be left out. And not including no.7, as was done in the initial analysis, is a clear error in sampling IMO. Unless a team plays 5 genuine bowlers then no.7 is always a decent bat.
I just don't think there's any evidence that Starc is worse against the middle-order than his contemporaries. I understand all the reasoning you've put forward to why you think so, but I'm just not sure it's actually based in fact. I don't completely disagree though. He's a swing bowler, he uses swing (regular and reverse) more than a Cummins-type bowler. This is just common sense. If you're just trying to say that he's less effective when the ball isn't swinging, that's great, so are most bowlers.
As I said before as well, a better stat would be his averages against each position (ie. runs scored divided by wickets taken) rather than pure percentage of wickets. I think that would show better his utility against the tail compared to other bowlers, which was the initial point of the discussion.
edit: and if we're going to continue this discussion we should try keep responses shorter. I'm not going to lie, I didn't read your whole post. It was massive.
