Those two yellows earlier are nice. Here's one possible explanation: you bumped up the dose from 3.4 u to 3.5 u on the morning of 10/13. This had the immediate result of giving you good numbers, so good it was a "no shot" that evening. When I get a "no shot" it generally tells me that a dose is a bit too high. Even though there are some yellows from 10/14 on, overall it looks like bouncing to me. You saw your first blue action when you got to 3.4 u, followed by higher numbers (likely bouncing) that prompted you to increase to 3.5 u. I'll go out on a limb here and say that 3.5 u is a bit too high. Maybe you could play with fat/skinny 3.4 u doses.
Bouncing makes us crazy! Increasing a dose when bouncing is happening creates a vicious cycle of dropping BG (maybe when you aren't testing so you don't see it) followed by higher BG on the rebound. The bouncing can persist several cycles and, in some cats, get them stuck up there. If there's any way you could get some tests at, say, +4,+6 and +8 or +5, +7, +9, or even a proper 12 h curve soon it might give you a better picture of what's happening - eg. we have a clearer idea of how Mr. B. did on 10/13 because of multiple tests. The other thing about bouncing is that even a drop in BG to a very reasonable number can trigger rebound. It's really complicated and almost impossible to make sense of it.
