I’ve just paid my largest ever ($90) payout for a Beeminder goal, but to @grayson rather than to Beeminder, in order to get permission to archive the goal.
On the whole I thought the goal went very well despite this and I’ve written up a post-mortem for other peoples’ consideration: http://www.drmaciver.com/2016/10/giving-up-on-giving-up-on-caffeine/
Insightful post!
I cut back on caffeine a few months ago, but didn’t eliminate it entirely for the reason you note in your post: I’m way sensitive to it. So I decided to keep a modest immunity going. My own iocane strategy, you might say. 
I’m down to one cup in the morning (plus any chocolate I eat during the day), but I’ve toyed with adding in a half-cup in the afternoon (both for the renewed boost and for all-day immunity; apparently withdrawal symptoms become evident within 12 hours, though I can’t say I’ve noticed them). That might be the perfect caffeine intake for me, I think.
I wonder if there’s a way for you to set up a Beeminder goal that keeps you in your caffeine sweet spot? (If you’ve any interest in pursuing it further.)
And thanks for the unexpected and fortuitous influx of cash. It’s been put to excellent use. 
I don’t actually find my typical caffeine intake much of a problem. It tends to gradually creep up over time as tolerance builds up, but is pretty stable on a day to day basis. On the flip side, when I’ve tried to specifically track how much I’m drinking I find that it becomes extremely stressful and causes me to become slightly caffeine obsessed, which is much less healthy than my caffeine usage.
My current intake is that I take a 50mg caffeine pill when I first wake up, have a mug of coffee when I eventually make it out to the kitchen, and then don’t have any caffeine for the rest of the day unless things are going really badly. This is more or less how my habits have historically gone too.
In particular I only have afternoon caffeine if I’m making a conscious decision that being functioning now is more important that being functioning tomorrow. My sleep isn’t great (that was the reason for the experimental caffeine break) and I’ve found both that caffeine in the afternoon makes that worse and that being slightly caffeine withdrawn by bed time makes that slightly better (because I’m sleepier).
That’s exactly what I found when I tracked my sweets-and-alcohol intake. More stress than value.
I could nix the keep-me-from-sleeping effect by regularly drinking coffee in the afternoon and evening, but so far I haven’t had a need to do that. (My reason for wanting immunity to caffeine is so that if, say, a restaurant slips up and makes my decaf a caf, I won’t be jittery for hours and unable to sleep. I solve that now by not ordering coffee beyond morning.) And I prefer other beverages in the evening, in general, so I don’t want to add a habit that requires me to drink coffee every evening or feel like crud if I don’t.
Though I suppose I could just eat chocolate for the evening caffeine…
BTW, a week on I can report that that was amongst the best £75(ish) I’ve ever spent.
My sleep isn’t great (that was the reason for the experimental caffeine break) and I’ve found both that caffeine in the afternoon makes that worse and that being slightly caffeine withdrawn by bed time makes that slightly better (because I’m sleepier).
Have you noticed any differences with exercise? I seem to find that if I exercise to the point of being seriously out of breath, at sometime during the day, I feel like sleeping at night even if I’ve had 2 or 3 cups of tea that day, up to the afternoon. Otherwise I need to stick to 1 or maybe 2 cups, and stop by about 11 am.
(Of course, n=1, and I haven’t carefully logged these results, so they’re just my impressions.)
Thanks for writing up your experience.
I don’t actually have empirical data on this (I need to start being better about recording this in some manner), but anecdotally it feels like no. I’m exercising to the point of exhaustion about twice a week at the moment and I don’t feel like I sleep better after that.
A thing that makes everything complicated and hard to diagnose things is that my problem is quality rather than quantity of sleep - I don’t actually have difficulty falling asleep more than about once a month, and I mostly get somewhere in the region of 6-8 hours sleep. I was using a Jawbone UP for a while and often it would congratulate me on having a great night’s sleep on mornings I wake up feeling terrible.
I’ve been using Sleep Cycle for a few months, and on many occasions it gauges the quality of my sleep differently than I would have rated it on waking. What I don’t know is which is more reflective of the truth; it’s made me question the relevance of my subjective point-in-time impression.
The positive (but possibly placebo) effect of this is that when the machine’s rating is high, I try to behave as though it might be true that I’m well-rested, regardless of how I might feel in the moment.
Often my self-rating of sleep is based on whether I wake up with a headache or not, so I’ve got at least one relatively objective measure of sleep quality that doesn’t seem well correlated with the UP.