Paying Is Not Punishment

Wednesday, February 9, 2022
By dreeves

Bee in stocks with sign that says 'i derailed on beeminder'

UPDATE: See follow-up post announcing No-Excuses Mode.

An under-appreciated fact about Beeminder is that it doesn’t force you to do anything. It just puts prices on things and you continue to do whatever you feel like doing, factoring in those prices. Just like you might buy a box of cookies if the price is right. [1] As long as what you end up paying Beeminder is low relative to how much more awesome you are because of Beeminder, it’s a good deal.

That is the philosophy of Paying Is Not Punishment.

We’ve been making related points for a long time. Bee’s classic “Be Nice To Yourself” post argues that not only is paying not punishment — it’s literally a treat. Treating yourself to a derailment once in a while is a perfectly legitimate choice for spending your money. As with the box of cookies, you just want the price to be high enough that you don’t buy them every day.

A bit more recently on the blog, we explained how Derailing Is Not Failing. Paying Is Not Punishment turns that up to eleven. To recap, derailing on a Beeminder goal just means a little kick in the pants and some respite and renewal of the commitment. That is not remotely like failure. Failure is setting your goal so unambitiously that you never derail, and don’t accomplish any more than you would have without Beeminder. In fact, it’s empirically true that, overall, the more you’re derailing the more awesomeness Beeminder is inducing.

Yes, that is a very self-serving thing for the creators of Beeminder to emphasize, since those derailments are how we make money. But that’s our point: Beeminder is not as perverse as it seems. The better Beeminder makes your life — the more words you write, steps you take, junk food you forego — the more you tend to pay us and it’s all very win-win.

Which brings us to what this post is really advocating for, namely, a no-excuses mindset, for at least some people and some Beeminder goals. Or to flip that around, it’s advocating against a mindset of “I shouldn’t pay for this derailment because it wasn’t my fault”. There’s no fault. You just voluntarily enter into a contract in which you pay money for allowing your datapoints to cross a bright red line. If you cross it, you owe the money — exactly as agreed and thus perfectly fair. The fact that you want to minimize those payments is what induces all the awesomeness you’re achieving by beeminding.

The universe may conspire against you but a derailment is a derailment. Paying is not a punishment that you deserve or don’t deserve. It’s just how the commitment contract is set up. If you zoom out from the trees that are the particular circumstances of this derailment, the forest is that you’re accomplishing all these things over the course of years with Beeminder-boosted motivation and the more you tamper with that motivation, like by questioning whether a given derailment was deserved or not, the less power it has. You want to maintain the nice clean property that crossing this red line is inherently costly. No exceptions or excuses. Then the incentives are pristine. At least that’s the hardcore extreme version.

For me, personally, the extreme version even works with things like illness. Maybe I’m sick enough that I’d rather pay this $X pledge than write this TPS report. Or maybe I prefer to suck it up and get it written. It’s a well-defined choice I can make. Beeminder just consistently exerts its motivational pressure.

I know some people really hate that way of looking at it though. Like, “if the universe can randomly conspire against me and make me derail then I lose motivation”. Or “I actively want to conceive of derailments as punishment and if they don’t feel fair then Beeminder is like a capriciously punitive jerk and I want to rebel.”

Again, my extremery does not work for everyone, especially not when it comes to illness! Just think about consistent, pre-decided rules and heuristics of your own that are as simple as possible (but not simpler).

What about flexibility?

Not everything needs to be hardcore and extreme and I know that some of our most amazing users could not afford to do that on all their goals. The flexibility Beeminder offers in terms of allowing for extenuating circumstances can be insanely valuable (also kind of expensive for us, of course, but it works out fairly). But this less flexible framing can be valuable and powerful for many of us.

Of course really there’s a whole continuum here. A middle ground might be something like, as our Support Czar once put it, would you make this excuse to your supervisor and feel like a professional? If not, then maybe don’t make it to Beeminder either.

Wherever you put yourself on that continuum, Beeminder is amazing at creating maximally flexible commitment contracts. Namely, Beeminder’s one-week akrasia horizon allows you to lock yourself in only for the upcoming week, with the option to change your commitment at any time, taking effect a week from now. That feature alone puts Beeminder contracts miles ahead of everything else in terms of flexibility. That’s true even in the case that you allow yourself zero flexibility in terms of excuses! But an infinitely flexible commitment isn’t a commitment.

If your (meta) goal is to commit to doing X but only when X is the optimal thing to do, that’s just never going to be perfectly achievable (for those of us who need commitment devices in the first place). If you can costlessly drop a commitment when you deem dropping it to be the optimal choice then you’ll end up making that choice suboptimally.

​In general, the more severe one’s akrasia, the more costly commitment devices will be but also the more valuable they will be. Maybe you need $270 at stake for each little task that keeps your life from falling apart. Even if you never derail at that amount, you’re paying dearly in attention and opportunity costs. But the benefit is your life not falling apart, so that cost is worth it.

In short, the greater the cost of the commitment device, the greater the net benefit. And commitment devices always inherently have some cost.

A pet peeve of mine when doing Beeminder support (not that this has happened in a long time that I’ve seen), is when a user argues that a derailment is not legit because circumstances became such that derailing was the objectively optimal decision. To make up an example: “I was super in the zone with my dissertation so it made sense not to disrupt my flow with teeth-brushing”. And it’s not that they’re necessarily wrong about what’s objectively optimal. Just that the power of commitment devices comes with a cost and it’s worth paying it. That could mean literally paying for the derailment or it could mean dispatching the beemergency at the expense of some productivity.

The brightest bright lines

By setting up a fully automated commitment mechanism I don’t touch any slippery slopes like deciding after the fact what counts as a legit derailment. If I’m derailing too often, I adjust the commitment, subject to the akrasia horizon. But the commitment — in this framing — is always the simplest, most bright-line-iest possible thing: “keep your datapoints above this literal bright line”. Sometimes I’ll derail blamelessly in a sense, but who cares? It just means the commitment mechanism isn’t perfectly efficient. No commitment mechanism ever is. Trying to eke out the additional efficiency of “derailing doesn’t count if it wasn’t my fault” is playing with fire.

And sure, I can pre-decide in my fine print about circumstances for calling non-legit but the slope is still a little slippery that way. Ambiguities can creep in. I can be tempted to look for ways to apply the fine print instead of focusing on keeping my datapoints above the line. (Also it just tends to feel more hard core and high-integrity to accept any and all derailments, but I’m highly biased here.)

This isn’t the only philosophy. As we concluded at the end of Derailing Is Not Failing, people improve their lives in a lot of different ways with Beeminder! My sense is the no-excuses, non-punitive mindset tends to correlate with greater user-awesomeness, but sometimes it’s the opposite.

Background and Related Reading

This post is the distillation of years of hand-wringing and debate and the gradual evolution of some community norms. Rather than narrate that history, we’re throwing in a list of links! (Hover over them for commentary/excerpts/synopses.)


 

Footnotes

[1] Is the price of cookies right? When deciding, you should probably mentally add a tax for their impact on your health. Beeminder can make that literal. It’s like a personal Pigouvian tax.


 

Thanks to Support Czar Nicky, our Queen Bee, Mary Renaud, Katja Grace, Eugenio Bruno, Luke Barone-Adesi, Michael Merchant, Clarissa Littler, @poisson, Nathan Arthur, Robert Perce, and Yassine Meskhout for helping to formulate the ideas in this post.

Image credit: Faire Soule-Reeves

Tags: , , , ,