How to balance between process and outcome?

post by Nathan Helm-Burger (nathan-helm-burger) · 2022-05-04T19:34:10.989Z · LW · GW · No comments

This is a question post.

Contents

  Answers
    2 adamzerner
None
No comments

I've been thinking recently about how to balance between process (how I get work done) and outcomes (what I achieve). I thought I'd ask the LessWrong community to see if anyone else has thoughts about this they'd like to share. I feel like both are important, but outcomes is a more long-term focus thing and process more of a daily thing. Outcomes are like long-running experiments for how you judge between different styles of process?  In cases where it's hard to get reliable outcome answers, when failing at hard things or succeeding at easy things, or timeframes are long, or uncertainty high, it can be tempting to over-update on limited evidence. Is it then better to test process types on easier examples and then extrapolate to harder ones?

Answers

answer by Adam Zerner (adamzerner) · 2022-05-04T23:17:27.589Z · LW(p) · GW(p)

This is something I've always noticed confusion about, so I'm glad you brought it up. Here is how I think about it.

Outcomes are a function of 1) process and 2) luck. For example, suppose you want to make at least $1,000/month from a side project within three months. That is an outcome. The outcome depends both on things you can control, and things that you cannot.

Intuitively, my first thought is "Why focus on things you can't control? Why not just focus on what you can?" Since outcomes depend partly on things you can't control, it doesn't make sense to me to focus on them. Instead, it makes sense to focus on what you can control: process.

However, I do recognize that focusing on outcomes might be a motivational hack. For example, if you know you have that goal of making $1,000/month from a side project, you might push yourself harder to reach it than you would if your focus was on something process-oriented like "put in a good days work every day". I'm sure there are other considerations I'm not thinking of, but to me the question seems like it's mostly about how much you want to use outcomes as a motivational hack.

↑ comment by Viliam · 2022-05-08T18:09:12.343Z · LW(p) · GW(p)

The problem with outcomes is that they depend on luck.

The problem with process is that it is unfalsifiable. You can keep doing the same thing based on mistaken assumptions, and telling yourself that you simply had bad luck.

Perhaps the right approach is to focus on the process most of the time, but once in a while check whether the outcomes match the prediction, and update accordingly.

Replies from: jmh, adamzerner, nathan-helm-burger
↑ comment by jmh · 2022-05-08T23:35:41.803Z · LW(p) · GW(p)

Perhaps your goal was to state things provocatively, or perhaps you were just being terse. If so then maybe this doesn't apply.

I struggle finding a way to agree with your suggestion if I take your claims seriously. If outcomes actually do depend on luck -- "depend" being a rather strong concept here I think -- then why spend any but the more trivial amount of effort on process? Process cannot solve bad luck or really influence good luck (if it can then I don't think we're talking about luck at all). 

Replies from: Viliam
↑ comment by Viliam · 2022-05-09T19:09:52.556Z · LW(p) · GW(p)

It is part luck and part how good the process is.

The law of large numbers suggests that in long term your luck should be about average. So if your process is solid, then a failure, or two, or even five failures in a row, are merely a delay.

(This is a simplification, of course. Sometimes a series of failures is fatal, because you run out of money or time; sometimes the opportunity is unique. But there are e.g. examples of people who started a few companies, first few of them failed, and then one succeeded and they became millionaires. Or investing in index funds: some years are bad, but the trend is upwards.)

The problem is, if your process is solid but fails three times because of bad luck, it looks similar to a process that is wrong and has failed three times completely predictably. If it's the former, you should persevere; if it's the latter, you should change your plans. How do you know which is which?

One advice is to keep learning, for example ask people who tried the same thing and succeeded, whether they see any obvious mistake in what you are doing. Check your premises. Maybe most of your process is good, only one important thing is bad or missing.

Another advice is that after failing X times, the probability of mere bad luck is f(X), which at some moment becomes smaller than the probability that your plan is just wrong. Check your excuses, how exceptional bad luck you assume. Bad luck 1:10 happens often. Bad luck 1:1000, especially happening three time in a row, is very unlikely.

In some situations, you could do some form of A/B testing: do two processes in parallel (if that is possible), measure the outcomes, then spend more time doing the more profitable one.

↑ comment by Adam Zerner (adamzerner) · 2022-05-08T21:10:49.516Z · LW(p) · GW(p)

I like that thinking. Makes sense to me.

↑ comment by gbear605 · 2022-05-06T00:42:41.159Z · LW(p) · GW(p)

One other benefit of focusing on outcomes is that it keeps you focused. If you're just trying to "put in a good days work every day" then you'll probably waste a lot of days on rabbit holes that aren't actually helping you get towards your goals. They'll feel productive, but in hindsight they probably didn't get you closer.

Replies from: adamzerner
↑ comment by Adam Zerner (adamzerner) · 2022-05-06T05:11:06.494Z · LW(p) · GW(p)

That doesn't seem true to me. You just have to find the right process-oriented things. For example, to mitigate against the risk of getting caught in rabbit holes, you can focus on process-oriented things like "spend X hours/week thinking about what the priorities are" and "give myself Y amount of slack [? · GW]". Also, outcome-oriented things suffer from Goodhart's Law [? · GW].

All of that said, I don't feel like I have a great grasp of all the pros and cons of process vs outcome oriented thinking. Those are just things that come to my mind.

No comments

Comments sorted by top scores.