Beeminding caffeine breaks (né "My latest weird Beeminder experiment")

Idea status: I just started this goal today. I’ve no idea if it will work. My weirder beeminder experiments have a very high failure rate (if you don’t count the stuff I do with tagtime, it’s basically 100%), so it probably won’t. I present this idea more for your amusement and consideration than as a thing I recommend following.

I use caffeine a fair bit. My actual daily levels are relatively low right now, but getting them down to zero is not a thing I’ve ever managed to consistently do. What instead happens is that every now and then I go cold turkey for a while and am miserable until I start taking caffeine again.

I’ve tried to use Beeminder to keep my coffee intake under control in the past and have mostly run into the fact that do less goals just do not work for me at all (psychologically, not technically), so I’m trying a new thing.

My caffeine intake on a given day is actually naturally self-limiting, in that it is based on my current tolerance levels. If I currently have no caffeine tolerance I’m not going to drink a pint of coffee, because I’ll feel useless and terrible afterwards!

So instead the goal structure I want is something that stops me from building up that tolerance. I already have the habit that does that - taking a period without caffeine - but I need a way of beeminding doing that. So the obvious starting point is to beemind days without any caffeine.

But not all average rates are created equal: Two days in a row without caffeine is much better for resetting tolerance than one day on, one day off, one day on, one day off (I don’t actually have science to support this, but it matches my experience and makes intuitive sense).

So I’m revisiting an old idea of mine: triangular beeminding (note: This didn’t work very well in its original incarnation).

The idea is that the first day without caffeine is worth one point, the second two points, etc.

Hopefully this should work very well, because as well as creating an incentive that matches the behaviour I want, it should also create a strong “Don’t break the streak” effect: I’ve been doing this for five days! I can’t stop now.


  1. I’ve got the rate set to 4 / week (I started at 7/week but then immediately decided this was too harsh)
  2. This means that if I take a 3 day break every week then I achieve my goal
  3. It also means that if I take a whole week off then I can drink caffeine for the next 3-4 weeks straight

Possible worries:

  1. The triangles haven’t really worked well for me in the past. I believe I understand why and they mostly don’t apply here (I’ve used them with do less goals in the past, and on the do more goals I’ve used them on I mostly just found them not very helpful but that was on points on a given day)
  2. The obvious failure mode is that this will cause me to feel like I have “permission” to take a lot more caffeine on the days that I’m caffeinated. I actually consider this an acceptable failure mode, both because of the self-regulating thing and also because I really mind the dependence more than I mind caffeine usage.



I recommend heavy use of retoratcheting to stop your successes building up huge buffers that let you backslide too much.

1 Like

I considered it but decided to try without. The ability to build up a buffer of a couple of weeks is very much a feature rather than a bug. I might set a three week autoratchet or something like that if I ever manage to get to the point where I can build up that much buffer! (My experience is that I cave after less than a week without caffeine)

1 Like

Progress update: Working quite well so far. I’ve done one two and one three day caffeine break. The third day was very explicitly done as a “Do I really want caffeine today more than I want it for three days in a row?” decision.

It has increased my caffeine intake on days on maybe a bit, but not to a degree that it worries me.

Still too early to really say it’s a good idea, but initial signs promising.


A follow-up update: Caffeine breaks suck. I’m on day two of a break and I hate it.

I’m going to of course respond with the natural solution to “I hate the thing” which is “Do more of the thing”. I was originally going to set the rate to 7 then went “No that’s too harsh” and reduced it to 4. I think 7 might actually be less harsh because longer breaks result in lower dependence, which makes the individual breaks suck less. I’ve now put the rate back up to 7 (i.e. one three day break gets me just under one week).

I suspect I will look back on this post as being the point at which this whole experiment went horribly wrong.


I can report so far that:

  • Being without caffeine still sucks
  • This is without any question the most effective caffeine intake management system I have ever tried

Long-term ability to stick with it still in question (See community policing thread), but I’m reasonably confident that this is going to work in the long haul.


As someone who has given up caffeine a few times the past year due to sleep issues, I think you may be designing it to maximize withdrawal time.

Even after 3-4 months without caffeine, a person can become dependent again with as little as 100 mg (about half a cup of regular coffee) two days in a row. For awhile I tried doing caffeine on the weekend, and it made Mondays and Tuesdays Horrible because I was redoing withdrawal.

Also, the standard advice is that it takes 1-2 weeks to withdraw from significant use, but it can actually take 2 months or more, depending on the person and their biology.

It’s worth thinking about your personal biology, thinking what you like about caffeine use - is there a replacement? Are you differentiating between coffee and tea? Measuring units? If your goal is reducing dependence, figuring out units may be very helpful - a half pot of black tea is equal to about one cup of regular coffee (depending on how you make both).

I have to use very little b/c after ten or so years of heavy use I get re-dependent quickly, and it dramatically affects my sleep and mood. DH can still drink it before bed, no problem.

Thanks for sharing your triangular experiments. I think they have a lot of potential.


You’re not wrong! I do spend the enforced caffeine breaks lightly caffeine withdrawn. But it’s a level of caffeine withdrawal that I would normally get somewhere around day 5 or 6 of quitting coffee. It’s not great, but it’s entirely manageable. It mostly just feels like having a bit of an off day.

My experience is that this sort of short period of low grade caffeine withdrawal is something that I can deal with much more easily than either a) The intense suffering of my normal caffeine withdrawal or b) The prospect of feeling like this stretching out indefinitely (because, as you say, I tend to keep feeling not much better than this for >= a month after quitting). The former prevents me taking coffee breaks, the latter ensures I inevitably fall off the wagon.

My vague intention is to gradually ratchet up the amount of breaks I have to take in an attempt to gradually ratchet down my level of dependence on it over time.

Anyway: Yup. It’s not an ideal system, but it’s a system that I’ve a decent chance of actually sticking to.

Hmm. That’s worth knowing, thanks. Explains in part why it still sucks after a month when I go cold turkey.

It’s not that I don’t believe you (it really isn’t!), but do you have a cite? I’ve tried to read more about this subject and I’ve really found very little good information out there. :frowning:

Not one that I can obtain legally without a prescription!

To some degree I use caffeine to self-medicate for the fact that my energy levels and sleep patterns are all over the place. This is not a problem I can get a doctor to take seriously, and most of the alternative stimulants I could use in its place are, well, not readily available to me without that.

A lot of why I’m trying to cut out my caffeine dependence is to either confirm or rule out that the caffeine dependence is the source of this problem. I don’t believe it is, but I admit that from the outside that belief looks highly unconvincing.

I’ve done estimates before and when in “Normal” mode my caffeine intake tends to be in the 200-400mg/day range. One largeish cup of coffee / day is about my normal intake, but how largeish and how strong varies over time.

In the past when I’ve tried to track units precisely to reduce caffeine intake the result has not been very good for me psychologically and resulted in my being obsessed with my exact caffeine intake 24/7, which is what this goal structure is designed to help avoid - binary allowed caffeine/not allowed caffeine seems all round better.

One thing I’ve had some success with (though am not currently doing) is to drink all my coffee decaf (fortunately I have a source of quite nice decaf coffee) and take all my caffeine in pill form so I’ve got a very precise notion of exactly how much caffeine I’m taking.

Hmm. Thinking out loud, here’s a thing that might work: If I can establish some sensible and easy to track mentally “maximum safe caffeine intake” and basically say that exceeding that costs me -10 points (or whatever) on my caffeinefree goal, that might work.

I’ll have a think about this and may implement something of this nature.


I will never be offended by someone wanting a cite =). Sadly NONE of this has actual solid data from good studies - at least not that I’ve found. That includes the standard advice that it only takes 1-2 weeks for physical dependency to wear off - lots of ‘experts’ say it, but it’s not evidence-based. Even the smithsonian mag, which usually has pretty good science reporting, quotes it at the end without a citation:

The comment box has a representative range of comments and stories abut withdrawal.

I read a lot of boards supporting people giving up caffeine, and the people motivated to visit those kinds of websites tended to have a much harder time - anywhere from 1-6 months before they felt normal.

For me, caffeine was indeed connected to insomnia, irritability, and anxiety (which is so sad - coffee is a big part of our family culture). After I gave it up I was sleepy at a normal time, and much calmer with my family and job. I did three or four trials of adding it back in, and would always have someone comment on my mood before I had made the connection.

I’m not saying caffeine is evil, as this is definitely not true for everyone, but it appears that I metabolize it really slowly. My brothers, mom, and I are all sensitive to a variety of stimulants (cold medicine, for example), so maybe it’s related to that. If you are having trouble with sleep/mood, I think it doesn’t hurt to check if caffeine is part of it, along with vitamin deficiencies, etc.

Addiction research is pretty behind, and I think public health substances are a bigger priority, so very few people are interested (or can get funding) for caffeine. Caffeine withdrawal was recently added into the DSM-V, so you can get a diagnosis (and a note for work if you need time off).

For me the replacements are more important socially and taste-wise, after I’m done withdrawing. Teecino is on my list of things to try, and I make weak tea/herbal tea to fill the hot bitter brew need.

Sorry I only have anecdotal stories - I know for me it took around 4 weeks to feel normal afterwards, so I identified with the self-selected comboxes. I’m sure there are plenty of people who were fine in five days who would never comment. The anecdotes are useful to get a sense of the range of reactions, although obviously not the frequencies.

Good luck figuring it out!


Yeah, that was my impression the last time I tried to look as well. I was hoping I was just googling the wrong keywords and that you’d be able to set me straight. Ah well. Thanks for the info!

The 1-6 months figure is quite distressing. :frowning: I was hoping 2-3 months would be enough to establish a baseline of what life without caffeine is like. Ah well.

In general I don’t think caffeine is the source of my sleep issues, because I actually sleep much better when I have a full blown caffeine addiction! The way I use caffeine I mostly frontload on it very early in the morning and then don’t drink any throughout the day. As a result come bedtime I’m feeling super sleepy because most of the caffeine has left my system. It might be responsible for the fact that I wake up feeling shitty (because I wake up caffeine withdrawn) but that doesn’t seem to be significantly reduced as my caffeine addiction ebbs.

Anyway, thanks for the info!

Incidentally, here is a variant I am considering giving a shot: As well as having caffeine break days counting up as triangles, have caffeine days counting down as negative the triangles. So the first day with caffeine is -1 point, the second day is -2, etc.

This seems like it would create a good incentive structure for avoiding the problem of maximizing caffeine withdrawal time, because as well as encouraging me to stay off caffeine for longer periods, the more days in a row I have caffeine the more I’ll be pushed into having a break.

I haven’t yet decided to do this (and definitely won’t be trying it in the next week or so because I’m on a trip where I really need the coffee) but it seems like it might be worth a shot. Thoughts?

1 Like

Actually, I have a cite! It’s not the last word in caffeine withdrawal, but it is a nice metastudy from 2004: “A critical review of caffeine withdrawal: empirical validation of symptoms and signs, incidence, severity, and associated features”. According to it, “Typically, onset of [withdrawal] symptoms occurred 12–24 h after abstinence, with peak intensity at 20–51 h, and for a duration of 2–9 days. In general, the incidence or severity of symptoms increased with increases in daily dose; abstinence from doses as low as 100 mg/day produced symptoms.” Obviously, ymmv, but that is the scientific conclusion as of a decade ago.

Here’s the link:
It’s paywalled, unfortunately, but if you don’t have access ping me and I’ll send you the pdf. It’s a great review, very useful if you want to stay on the safe side of tolerance. Short version is as little as 50mg/day for 3 days can cause dependence, which is both frightening and explains why Starbucks does so well.



Well, that’s a pretty strong argument in favour of the negative counting approach (which I just started doing, with today being my first negative points)

1 Like

Due to the comparative success of this goal, I’m trying a parallel one on the same rules regarding avoiding sugary snacks:

Will report back on how it goes.

1 Like

First instance of behaviour difference based on negative scoring: Today would have been a -3 day, so it’s a 1 day instead.

I hate this a bit, but I think that might be working as intended.


I’ve given up on the sugarfree goal for now. I’m running very low on cope right now, and having to do both this and caffeine at once seemed too stressful.

One thing I did notice in my brief experience of it is that I’m much more prone to binge on sugar on days when I’m allowed it than I am on caffeine. This isn’t intrinsically a problem (a lot of diet plans have “cheat days” where you’re explicitly allowed to do that) but does suggest the goal rate might need to be set quite high if it’s to work.