Don't Fear Failure

post by atucker · 2011-04-03T22:52:32.694Z · score: 30 (31 votes) · LW · GW · Legacy · 12 comments


  Overestimation of Damages: "Its not that big a deal"
  Rationality and Failure: "Don't panic"

Last post, I talked about how trying things out yourself is a good way to learn about them. This post, I'm going to talk about ideas that helped me overcome one of my major obstacles to trying something -- fear of failure.

Overestimation of Damages: "Its not that big a deal"

In most cases, failure really isn't that big of a deal. Really. The difference between failure and a null action is the attempt. If the attempt isn't damaging, failure isn't damaging.

A few cases:

There's lots of things where an attempt is actually worse than doing nothing. Jumping halfway to the other side of the ledge, for instance. Or only removing most of the toxic part of a pufferfish. But for a lot of potentially high-value things, a failed attempt doesn't really do much, so you might as well try them.

Rationality and Failure: "Don't panic"

Some people I know basically buckle under failure. A common failure mode seems to be to do something badly, establish an ugh field around that area, and then continue in a downward spiral. Getting a B on a math test turns into "Ugh, math", turns into "well I was never really good at that anyway", turns into a complete lack of effort. Here a little failure becomes a huge problem. A failure isn't catastrophic on its own, but giving up is.

Rather than focusing on what you didn't accomplish, try to figure out what happened insofar as understanding that helps you fix it. Toyota apparently emphasized this as their  5 Whys system, and it has become widespread in industry.

Another good reaction to failure from my life, at a robotics competition:

Just  learn from your mistakes -- figure out what happened and what can be done to fix them. Then do whatever is needed. Try things you're interested in and learn enough to carry them to completion, or just be the wiser for having started them. 

quote paraphrased from  here. His username on LW is lionhearted.



Comments sorted by top scores.

comment by meta_ark · 2011-04-04T03:37:44.395Z · score: 7 (7 votes) · LW(p) · GW(p)

Very true. Since starting university I've had many more opportunities to try new things, many of which I didn't think I'd like, but had a small enough failure cost that it seemed worth trying.

I'm now head writer of a university show, in a Madrigal choir and can dance salsa. I'm having a lot of fun.

In social situations, where you fear embarrassment and a status loss if you fail, you could try to cultivate a reputation for giving anything a try. Most people have the same fears, and trying something scary will usually get you respect because you're facing fears others couldn't. Trying and failing at something can actually be a status-raising move.

comment by atucker · 2011-04-04T23:15:01.561Z · score: 1 (1 votes) · LW(p) · GW(p)


Sebastian Marshall talks about doing things that have a high upside and a low downside. I'm trying to do that, but have currently been running into issues with previous commitments (which, luckily, seem to largely run out by the end of April).

comment by Fyrius · 2011-04-06T18:56:41.305Z · score: 2 (2 votes) · LW(p) · GW(p)

1 quote paraphrased from here. His username on LW is lionhearted.

(Is it me or does this footnote not refer back to anything?)

comment by Zvi · 2011-04-04T12:15:43.326Z · score: 2 (2 votes) · LW(p) · GW(p)

Understanding what happened is important, but I find five iterations to usually be far too many. You have to be careful when recursively asking "why?" because otherwise you end up with this.

comment by atucker · 2011-04-04T23:13:00.378Z · score: 0 (0 votes) · LW(p) · GW(p)

Point. That's just what they do at Toyota, where I imagine that they work with systems complicated enough to actually merit 5 "why?"s.

I personally find 2-3 to work in most cases.

comment by PhilGoetz · 2011-04-04T00:43:17.426Z · score: 1 (1 votes) · LW(p) · GW(p)

I really like the pottery story. Not sure if it's saying the same thing as the rest of the article, which contains some good advice, but felt long.

comment by atucker · 2011-04-04T01:14:03.685Z · score: 0 (0 votes) · LW(p) · GW(p)

Fair point.

I split them up now. Not sure if I was supposed to do that, but you can just reply here about that.

comment by rabidchicken · 2011-04-04T13:24:34.356Z · score: 1 (1 votes) · LW(p) · GW(p)

When I saw the names of the articles and that you had posted twice in a row, I thought you were delibrately invoking your advice to try quantity over quality. It is rather amusing that you ended up doing this accidentally.

comment by atucker · 2011-04-04T15:57:14.531Z · score: 1 (1 votes) · LW(p) · GW(p)


Actually, I thought that the quality would be improved by splitting them, which happened to increase quantity.

comment by atucker · 2011-04-04T01:05:54.771Z · score: 0 (0 votes) · LW(p) · GW(p)

Would it be okay to split the article after I submitted it?

It seems like the pottery story and the other two sections could be separated into separate posts.

comment by Dorikka · 2011-04-03T22:52:29.239Z · score: 0 (0 votes) · LW(p) · GW(p)

Often, it makes it

I think you lost some words right before the "Rationality and Failure: "Don't panic" section.

comment by atucker · 2011-04-03T22:55:00.526Z · score: 0 (0 votes) · LW(p) · GW(p)

Thanks. I forgot what went there, but will go add it if/when I remember.