Actually, it clarified what you were saying considerably. I agree that if the low overnight sensor reading is causing your delivery to undershoot your basal needs irrespective of DP, then it is going to make your BG rise, again irrespective of DP. I don’t even see that as particularly controversial: same thing would happen if the same sensor problem were occurring in the afternoon–or whenever. In that sense it’s really just a semantic wrangle.
The only thing I’d question is this:
The thing is, the ICR lets you set different ratios for different times of day–practically the only thing you can do in auto that feels like you have some actual control over it–so if you do one for, say, 3am-5am or whatever, it shouldn’t affect your wizard setting for eating breakfast at 6:30 or other meals later in the day.
That’s what I was experimenting with, setting a different, super-aggressive ratio in the wee hours, and I thought I was seeing some progress, but then one morning it started to go the other way, and I sat there looking at another higher-than-I-want fasting BG, scratching my head. The problem was that I couldn’t tell what to conclude from the result. Maybe the official word I got was correct, that it wouldn’t make any difference, and I’d been seeing some improvement for other reasons. Or maybe it did work, but there’s that whole business of new settings not getting integrated into the algorithm for 6 days or whatever it is, so maybe it hasn’t integrated yet and it’s not doing anything yet. Or my infusion site was losing absorption… or the sensor was registering lower than before… or… or… or… And I finally said to myself, Ya know what? I wasn’t having this problem on manual. What exactly is the benefit I’m getting in return for all this struggle? Heck with ICR ratio, what about AVB–Aggravation Versus Benefit–ratio? That’s when I switched the thing off.
I think the chief problem that we’re all running into is that Mdt is so closed-mouthed about how the algorithm actually works that we’re left at the mercy of user-community lore, some of which may be fact based, some not, and nobody actually knows except by empirical fiddling. Experimentation, in a word. There’s nothing hidden about how a manual pump works, but with auto mode you’re continually reduced to trying to deduce what’s in that blackbox.
As I’ve said in other posts, technologies embed ideologies, and this one veers uncomfortably toward one in which patients are supposed to shut up and do what the Dr says. The upside is supposed to be that it takes all the fuss and bother out of dealing with this dangerous drug, insulin, that we have to dose ourselves with–something unique to this disease that the medical profession has never been particularly comfortable with. But that easily shades over into the old “We’re the medical professionals, you don’t know enough to have an opinion” attitude toward D-care that I thought we had long since moved away from.