Beeminder Forum

Why I don't feel commits.to does what I want it to, and what I would prefer instead


#1

In a recent email to @dreev, I wrote

I wish there was some good way to use Beeminder for very short term goals, the kind you want to do exactly once, but very soon. Things like “I’ll write a reply to this sometime today”. For a while I had a “must-do” type goal, but that doesn’t help with things you want to get done sooner than tomorrow. I find that if I leave something like this to the next day, I lose a lot of my motivation for doing it in the first place.

I want to clarify what I mean here, and why, having considered commits.to, it doesn’t match up with what I want.

One of the great things about Beeminder is that it takes your money. This gives failing in your goals a real and immediate sting. This is a good thing. In fact, it’s the only thing. If it weren’t for that, I wouldn’t be using Beeminder.

But commits.to is a different model entirely. Beeminder’s model allows you to make private commitments to yourself. Indeed, all my Beeminder goals are marked as private, because I know that otherwise I will feel really self conscious every single time I interact with Beeminder, knowing that I’m broadcasting my actions to (potentially) the whole world. (Even if, in practice, probably no one cares enough to look at it even if it was public.)

It seems that commits.to takes the opposite tack entirely. Its reason for existence is public commitment, which is the precise opposite of what I want here. I want to make private commitments, to myself. Long term, ongoing commitments are something I can make with Beeminder, but short term, one-off commitments are something Beeminder’s UI isn’t set up to do well (even if, in theory, Beeminder can handle it.)

Another advantage of Beeminder (as opposed to commits.to) is that it makes failing, well, not exactly OK, but something with a well-defined monetary cost. This means that I can make a reasoned decision of whether I am willing to derail or not. By contrast, commits.to, with its percentage scores, doesn’t really encourage that. I don’t want to succeed all the time. If I find that I am, I that means I definitely wasn’t aiming high enough.

But it all comes back to the first thing I mentioned, the fact that commits.to is built as a public_commitments engine. All the various things that get in the way of it being good for private commitments to myself are actually features. Many of them are really good features for if you want to measure (and show off) how well you stick to your (public) commitments.

All I want is Beeminder, really. Beeminder for one-off goals. I want to say to myself “Today sometime I’ll get around to replying to that email” or “pick up groceries on the way home” or something like that, and then, if at the end of the day I haven’t lived up to that, to pay $5 (or some other dollar amount).

That’s something I should be able to do with Beeminder, but can’t. The important factor there is that it’s a one-off event. It’s not every day that I feel that there is a specific email I have to respond to today, or groceries I need to get, or whatever other short-term goals I sometimes have.

There are ways around it. I mentioned “must-do” goals. I had one for a while, and it was fairly good. But because it involves setting a goal for the next day, it doesn’t actually help with this the way it could with goals that I want to get done today.

Perhaps there is a variation on the concept of a “must-do” goal that might work. If anyone has any ideas, I’m eager to try some out. But this post is mostly a response to @dreev about why I feel that commits.to, for all its promise, just doesn’t cut it for me. I’m not currently looking to track my public commitments (although who knows, I might get into doing that too, sometime), but rather something somewhat different.

So fine, commits.to isn’t for me. I’m sure it works well for others who have slightly different things they want out of it. But I just wish or hope that Beeminder can somehow build a feature that covers my use-case, as well. (Note that I am complaining here mostly that Beeminder doesn’t currently have a way for me to give them money in the way that I want to. (i.e. when I fail my short term goals.))


#2

I have a work hours commitment goal where I enter a negative number datapoint large enough to put me in the red plus the number of hours I’m committing to work that day. Then, when I meet the commitment, I change the value to one. I think you could do the same thing with a must-do type goal. If you want to completed it today, just set the value of the incomplete datapoint to a large-enough negative number and you’re all set.


#3

I see how that can work for something like a goal that measures a number of hours. I’m not quite sure how that could work well for multiple discrete, independent tasks.

If I’d like to be charged e.g. $5 for each task I fail to do by the end of the day, I can’t really bunch them up together in one Beeminder goal. And I don’t see how I’d spread them over multiple Beeminder goals in a way that it’s easy to use.

The reason I don’t want it to be that if I fail one of them, I fail the goal for the day, is that if I see that I am likely going to fail at one of them, I still want the motivation to work on the other ones.


#4

Have you given the GTBee app a try? It’s specifically made for one-off goals. You set a deadline and pledge money and it charges you if you don’t mark that item as done by that time using your Beeminder info.

The other thing that comes to mind would be to have a “do less” goal to penalise procrastinating 3 (4, 5) minute tasks that are not worth scheduling but should be done the instant you think about them.
Every time you catch yourself artificially delaying them you put a point into that goal.
You could draw the line there. Tasks that are more involved are free to be scheduled to the next day while tiny tasks are to be done ASAP.

Does this or GTBee sound like it might work for you?


#5

Here’s the support article on GTBee. Big bummer it’s only for iOS. I’d be interested in something like this for Android.


#6

See it like this: At least the Android users have one app that works well :wink:

(Now I feel bad for @apb, but I couldn’t resist :grimacing: )


#7

Yes, GTBee does indeed sound more or less what I am looking for… but it’s only available for iOS.

If it had an Android version or a web version (or ideally, both), I’d certainly give it a try.


#8

Looks like some work on an unofficial Android GTBee app went on here, though doesn’t really seem like it’s currently being supported.


#9

You can download several versions of GTB for Android but none of them currently charge you. I’d love this as well.

Feel free to help out!


#10

Also see this thread on the topic which has some good ideas:

In particular this procedure for making a one-off beeminder goal might work for you - it shouldn’t be too hard to automate:


#11

…and after checking out that thread and then the threads referenced in there and so on I found myself confronted with a. myriad. of. tabs.
Which ultimately shows me: You are not alone¹.

I assume that the Bee Team is quite aware that there is some sort of demand for this however high it is.
Yet the only² immediate answer to this seems to be GTBee for iOS which is… alright but you can tell it’s not the main focus of Beeminder.
Which finally brings me to the question, why is it not? There is probably a very good answer to this and it’s probably somewhere in the forum but ugh… I’m afraid of another flood of tabs. Something something rabbit hole something something.

Either way, there is plenty of ideas in this and the linked threads. I know the urge to play them through in your mind and disagreeing with all of them is strong (I’m definitely guilty of this) so I urge you to pick one right know and just give it a try for a few days. What do you have to lose? Who knows, maybe you find something that works for you.
@zzq: if I’m totally wrong here then please just ignore this :slight_smile:

¹ Yes, I’m having a TV tropes moment.
² Admittedly I skimmed the linked websites very briefly


#12

My guess is that it’s primarily a “too little time, too many priorities” reason. But I’d be interested to hear the real reason. :slight_smile:


#13

Well there was a working GTB for Android, so there was no need.

It shouldn’t be too hard to get it working again - there are just two bugs in the code that need to be fixed: the charging doesn’t work and there’s a problem with the login authentication.

Once those are fixed we should be fine.