Weasels and Barfing Cats


Ok, hopefully I won’t embarrass the person who most recently inspired this but it happens a lot. Also I completely changed all the details. So, imagine that you replied to a legit check with something like “Not legit! My cat got sick and barfed on everything in my house and I totally ran out of time to hit my do-less goal for junk food!” Maybe it’s something super reasonable that we feel like huge jerks not being sympathetic to.

(Aside: If you reply with simply “not legit” we reply back with something like “oh no, what happened?” The point being that calling not-legit should be super painless but we’re going to at least have an unautomatable interaction with you about it.)

Anyway, here’s an example of what we’d say about the cat-barfing thing!

Hey $USERNAME, we’re always super worried about breaking Beeminder’s incentives in cases like this! Any kind of “just this once” exception is a perilously slippery slope, as you can imagine. Arguably the point of Beeminder is to make sure important goals don’t get crowded out when things get busy. (We’d love to nerd out about the philosophy and behavioral economics of this kind of thing if you have more thoughts!) Can you formulate a general principle that would apply to this and all future similar cases where it would make sense to call such derailments not legit?

It’s common for people to actually decide to feel good about paying us after reading that, like they’re buying the continued credibility of Beeminder’s threat. Once in a while the person will formulate a totally sensible general principle. (A recent one was “if something happens that involves me having to wear a suit, then the derailment isn’t legit”.) And sometimes the person will kind of persist in their weaseliness and if they didn’t weaselproof themselves then we go along with that, maybe guilting them into working off the debt instead by giving us $X worth of feedback about how Beeminder could’ve held their feet to the fire better.


It sounds like this philosophy boils down to “Your excuse is orders of magnitude more reasonable if it can be characterized as an example of a reasonable, general principle.”. That notion is immediately agreeable to me, and it seems obviously true. But I wonder how widely held that opinion is.

Is this principle rooted in the same part of our brains that thinks that class List<Foo> is elegant and class FooList is ugly? Dealing with specifics is dirty and it means you don’t really understand the problem space. The more you generalize your edge cases, the easier it is to convince yourself and others that you are being reasonable. I think this way every day.

But are there people out there who don’t think like this? And if so, do they use Beeminder? And if not, do you care? I wonder!


My own tactic for handling barfing cat days is to make sure the bare minimum I have to do to accomplish my goal every day is doable basically always, no matter how ambitious the goal is normally. So, for instance “do everything on my to-do list” is a hard daily goal, but TECHNICALLY I just have to TOUCH everything on it, whether that’s “open blank email” or “make list of things to do for that task”, and THAT shouldn’t ever take me more than 10-15 minutes if I just blast through doing the bare minimum. So I can do it even when the world is on fire and burning down around me.

This probably wouldn’t work so well if I didn’t ACTUALLY want to do my beeminder goals, though, because I could always skate through doing the minimum. On the other hand, eventually if I keep doing the bare minimum, I’ll have accomplished something even if it takes ages!


@lanthala, thank you for coining “barfing cat day”! That’s totally a thing now. I.e., you have something that feels like a fair excuse but really the whole point of beeminding your goal was to make sure it got done despite such excuses.

And @drtall, yes, our minds are clearly similar but I doubt that it’s a prerequisite for beeminding. (This reminds me of The Type Bee Personality.) The anti-just-this-once principle is elegant, yes, but more importantly it’s accomplishing something highly pragmatic, namely unslipping a slope.


How does the support team feel about derailment responses like “eep, forgot to enter data today, should be all good now?” This seems like a situation where either the user actually did forget and wasn’t in a place to get zeno polled, or a stealth weasel. I tend to do this a lot, and I’ve never used it to weasel out of a payment that was due, but that seems like gray-area phrasing. Does that happen a lot? Or are people usually more transparent when they try to weasel out of something?


That’s the most common of all and it’s totally fine. I guess we have no way to know how much of it is stealth weaseling, but that’s also true for entering fake (or “IOU”) data.


Just my experience but the “general principle” feedback worked well when I cried non-legit in the past.

In one case where I did the goal but missed the deadline due to a late night of work (a rare but not unheard of event), I learned that deadlines don’t have to be at midnight and was able to adjust so that my daily deadlines had more leeway (generalized fix).

More generally this question has also helped me craft more robust goals ala @lanthala, things that get me started doing things toward my goal but are more easy to accomplish even in the face of barfing cats. Also knowing I’m going to get this question has made me self censor a few “not legit” calls before I even sent them.

On a final note, shout out to the support team, they have always been superb in all my interactions.


Maybe I’m missing something (and I want to be clear I’m not picking on this particular user at all!), but it’s not clear how a barfing cat is related to eating less junk food. The two interpretations I could see are “I was so distracted I forgot to enter data, but I didn’t eat any junk food either” OR “I was so frazzled, I couldn’t get dinner made and had to order out.” Those are very different situations!

In my mind, simple “I forgot to enter data” cases don’t need much prodding, beyond perhaps making sure the person is aware of the options for variable deadlines, reminder settings, etc. Nearly all of my non-legit derails are some variation of “I forgot”, and if anything it’s become more likely to happen as my new habits get better established.

The second case is something a bit different, and worth asking the question of a guiding general rule. Presumably, someone wants to eat less junk food because it’s become too much of a go-to option in their life for their preferences. This would be exactly the sort of situation where the rubber meets the (yellow brick) road, and a little prodding might actually be key to helping them think through exactly what they want out of this goal.


I thought that was the point—that it was an intentionally nonsensical excuse.


Yeah, that wasn’t the actual excuse; I was just being silly. :slight_smile:

I’m still going with “barfing cats” from now on to refer to something that’s kind of a reasonable excuse for not meeting a goal but ultimately in the category of excuses that you want Beeminder to make sure you stay on track in spite of. The question can be like, is this a barfing cats excuse or a genuine invocation of the SOS clause. Except not phrasing it like that to users obviously.


“Barfing cat day” is certainly a memorable handle but I had the same interpretation as @kenoubi and @gretchen that it was intentionally nonsensical (especially with “ran out of time to do my do less goal”) so I’ll probably have to rewire that synapse eventually :slight_smile:


I like the term barfing cat for any small to medium-sized phenomenon that threatens our goals or processes directly or indirectly. I think the category that we are discussing already excludes clear BS – the idea is that you can really sympathize with someone who has to deal with an actual barfing cat. It’s only when someone uses it as an excuse that it might morph into a weasel.

Putting aside whether it qualifies as a valid excuse, barfing cat —> ate junk food makes for an amusing but fictitious example, but I don’t think it’s so far fetched. It depends on all the other constraints someone might have in their life – for example mass transit delays, family and work commitments, a grocery store that closes early, a rather broad definition of “junk food”, and some food allergies. The barfing cat might just be the proverbial last straw from the narrative perspective when the rest of it were latent logistical or organizational issues.

There’s an implicit part in committing to a goal or habit: when you say, for example, “I commit to learning N words in a new language every day,” you also really want to be saying, “I want to get all my stuff together to make it possible and highly likely that I will learn an average of N words in a new language every day”.

So that leads me into a related thing I’ve been thinking about for a while: for most of the challenges in my goals and habits, the larger difficulty lies in logistical rather than motivation. That’s probably why, for a long time, I’ve been skeptical of emphasizing akrasia as the central issue. But it could be my experience is very different from others, and Beeminder already does a great job of countering akrasia once you understand how to set up your reminders and structure/quantify your goals in a meaningful way.

Regardless, I have mostly derailed on goals these days because of inadequate planning, inadequate systemization, overestimating my capability to cope with variations in logistical factors. Maybe there were some local moments in there we could call akrasia, but in the larger picture planning fallacy played a much larger role.

I think the commonly accepted advice to start goals really easy works well initially. Once things seem to be moving along for a while, you can get optimistic enough to want to increase the rate, or create more goals. Some folks, including me, have the inclination (vice) of starting goals/projects/making commitments in order to boost motivation (get the feeling that things are changing and so on.) But in general, multiple goals with overlapping constraints (unless you manage something like a waterfall) add some complexity. And I think it’s easy to model resources like time and attention in terms of spatial analogues, rather than in terms of flow quantities (as Personal Kanban puts it).

I do think Beeminder already does a lot protect against some forms of planning fallacy by forcing things to be quantifiable, and structuring progress in terms of a daily average rate.

I think us talking generally about barfing cats (barfing cat events?) as things that happen is a good next step. Do you have a system or process underlying your goal? Does that system or process and all the other constraints (including other goals) leave enough room for a barfing cat to make its way through? How large of a barfing cat can it accommodate? I think it’s good to have in the lexicon, right along with weasels.