From my journal. Originally posted on my personal blog. Status: quite speculative, but there's something here.



We could probably put a -5 to +5 scale of behavior together that was logarithmic about the enduring good/bad impact of various activities.

Something totally neutral — say, neutral leisure that’s not particularly recharging nor distracting — that might be 0.

Activities or time that were +1 would be very slight gains without much enduring impact, like slightly-recharging leisure. +2 might be things with slightly more impact, like doing chores or cleaning.

The thing about logarithmic scales is that they get big, fast. +3 might be good one-off business-building activities, +4 might be establishing an ongoing revenue channel that keeps outputting gains, and +5 might be an absolute game-changing type thing like getting workable protocols on a vastly untapped marketing channel like Google Adwords in its very early days (with the potential for hundreds-of-thousands to millions of dollars in profit if done right) or recruiting an absolute superstar employee or some such.

On the negative side, -1 would be lost time without much repercussion (maybe a slow start to the day kind of staring off into space), -2 would be slight ongoing negative consequences like starting the day surfing the internet, -3 would be starting to take multi-day damage with bad decisions, -4 would be doing seriously dumb stuff with very long-term implications — for instance, stubbornly training through an injury and turning it into multi-months of physical therapy required, and -5 would be multi-year damage events like a serious alcoholic or heroin addict relapsing onto booze or heroin respectively.

Logarithmic scales are notoriously hard for people to get their mind around, but actually map well to reality in some instances.

For instance, if you spent a half-hour at “-1” quality time (-30 total) and then two hours at “-2” quality time (-1200 total), but then 90 minutes on a good one-off revenue generating campaign at +3 (+9000), then you come out considerably ahead. Staring off into space or surfing the internet is of course wasteful, but a solidly inspired 90 minutes of revenue-generating is much better than that. It’s possible to get a lot of gains in 90 focused minutes.

-5: -10,000 points per minute
-4: -1,000 points per minute
-3: -100 points per minute
-2: -10 points per minute
-1: -1 point per minute
0: 0 points per minute
+1: +1 point per minute
+2: +10 points per minute
+3: +100 points per minute
+4: +1,000 points per minute
+5: +10,000 points per minute

That might be too steep in some cases, and not steep enough in other cases; pseudo-logarithmic might map to reality better, but I think something like this actually maps to the reality of “great days”, “okay days”, and “bad days.”

A day you spent the whole day cleaning and doing chores, say 10 hours of +2 time, would be “6,000 points” — you’d have been better off in some respects by doing just an hour or two of higher-tier work that day, but +3/+4/+5 time isn’t always easy to come by.

At the extremes you can see some of the potential correctness shake out. A day that you did a ton of one-off good sales and prospecting, even 10 hours at +3 time, you’d get “60,000 points” — very solid, but if that was followed up with just a few hours of destructive behavior it’d give up some of the gains. A short, way-too-hard drinking session to celebrate — to someone was a mildly irresponsible drinker but not quite a relapsing alcohol — might come in as four hours of “-3 time” and -24,000, suppressing the next day’s performance and having a slight carryover for a few days and giving up some of the day’s gains.

A day of lots of scratching and grinding and distraction (say, 10 hours split between -2 or -3 time, for -3300 total) followed by just a couple hours of serious breakthrough +4 time on systems or ops or revenue or whatever (120,000) actually comes out to a great day. Of course, it’s rare to put in breakthrough time after a totally scattered and wasteful start to a day, but it does happen — and those days, in retrospect, are like, “Huh, that’s a weird day but it went well.”

It also explains — doesn’t justify, but does explain — how people doing extremely large-scale business or culture can oscillate rapidly between massively self-destructive behavior and massively successful behavior and seem to get away with it for long stretches of time — Jim Morrison (1943-1971), the lead singer of The Doors, certainly put in the musician’s equivalent of lots of +4 and +5 behavior on making art and music and popularizing it and really connecting with people... but even short extended runs of -5 time can end your life, as it did in his case.

Of course, when young people who aren’t celebrities or business moguls emulate the more destructive parts of that scene, the -5 time doesn’t get bulwarked by +5 time in business or art, and you wind up not with the legacy of a tortured genius like Morrison, but with... well, you ruin your life.

Are there practical applications here? Well, not yet. I’m trying to find some way to capture the wildly-varying quality of how time is spent, trying to make more of a clear distinction between going well and going amazing, etc.

Days with no negative time at all and some slight positives are decent. More negative time can be absorbed at lower levels if compensated with more time in the higher-performing categories. At the high ends of the extreme, you could even maybe get away with significant -4 and -5 time if you’re doing massively-impactful positive things, but if you can get massively impactful positive time without getting stuck in high-negative time — like, say, Gene Simmons in music or Jeff Bezos in business — then you wind up building a really consistently exceptional life.

Logarithmic scales are hard to calculate intuitively; when 10 minutes of +2 time is roughly equivalent to two hours of +1 time, that’s unintuitive. A single hour or two of +5 time in some cases might be more important than everything else you do that month, so long as you don’t give up all the gains with -4 and -5 time.

This is still very much a work-in-progress, and to have any chance of viability, at least three things would need to be true —

  • (1) The weighing of logarithmic-time-quality would have to be calculated automatically in a spreadsheet or an app,
  • (2) There’d need to be a clear and easy delineation of behaviors that are very easy to classify as +1, +2, +3, etc.
  • (3) Recording and tracking these numbers would need to be as easy as possible, so it wasn’t a chore or overwhelming.

And the scales might be wrong — maybe pseudo-logarithmic, or with each tier being double the level below it, or some such. It’s certainly the case that there’s a very few activities that we can occasionally identify that are more than 1000x as impactful as normal time — think of Einstein in the Swiss patent office, compared to how most patent clerks would have spent their time — likewise, there’s a limited range of very destructive behavior that gives up all the gains in life, that’s 1000x worse than normal distraction or foolishness.

The more of life can be deployed into those highest-level tiers of ongoing positive impact, and the more the negative tiers can at least be ratcheted down to the manageable -1 and -2 time which is bad but not enduringly bad, it seems like this becomes a sort of recipe for life. A high-earning person on salary who spends only 20% of what they make and conservatively invests the difference — going the “early retirement extreme” path — has a mix of lots of +3 and +4 time, and probably doesn’t need much +5 time to hit their targets. The disruptive entrepreneur or pioneering artist or inventor, on the other hand, hunts for +4 and +5 time, often at the expense of nights of low sleep, long days, occasional thrashing around.

Rough thoughts — still a work in progress. I think there’s definitely something here, though.


2 comments, sorted by Click to highlight new comments since: Today at 11:26 PM
New Comment

"This model of the quality of time feels much closer to the truth than..."

At the ellipses is when I realized I don't have a good model of how I model the quality of time.

A lot of my personal unhappiness comes from when I thought reality was supposed to be some way other than what it turned out to be (classic example, "XYZ happened, I'm supposed to be happy now, ARGGG what gives?!"). Thinking of the quality of time more logarithmically seems like it would produce less feelings of, "Things should feel differently from how they feel :(".

Well said. Time is maybe the most fundamental thing we've got to manage and navigate to have the lives we want, and thus I find it strange how little thinking people give the topic... the scale described here obviously isn't perfect, but hopefully is useful as a jumping-off point to analyzing time usage a bit for how well it's being spent.

New to LessWrong?