Everyday Systems: Podcast : Episode 96

Greenishness

Do you use Duolingo, the language learning app, or fitbit, or some other biometric tracker? If so, you’re probably familiar with the concept of “streaks.” It’s when you track how many consecutive days you’ve hit some daily goal, like how many days in a row did you do a language lesson, or did you get at least 10,000 steps, or, say, how many days did you practice the No S Diet without any failures. While you’re on a streak, it can be very motivating. Especially as it gets longer, you don’t want to break it. It’s like this necklace you’re building that you add another pearl to every day. It becomes more and more precious, more and more worth keeping intact every day you do it.

The problem is when inevitably you fail and the streak breaks. Then the necklace gets torn off and the pearls are scattered all over, gone.

And unless you are a psychopath, you will fail. One day you will simply collapse from exhaustion right before getting your 10000 steps, because you stopped paying attention, because you figured you were close enough that it was inevitable and you could just lie down and watch TV. That’s what I did last night, just 22 steps short, breaking a 144 day streak. Or you get sick. Or you have your self-optimization efforts put into proper perspective by the death of someone close to you. Or you just run into a twinkie that you can’t face down.

Then the streak mentality that was carrying you along suddenly jerks you back. Your screwup today cost you not just today but also the last 20, 50 or 144 days or however long it’s been. The more days you’ve succeeded, the harder it is to recover from. It’s like a black hole of loss sucking you back. Psychologically, all your successes were just setting you up for a bigger and bigger, more traumatic failure. You feel not just a sense of loss then, but foolishness and rage for having set yourself up like this. And then you might think, “What’s the point? my streak is gone, I might as well wallow on the floor. I might as well at least take a few days off. Or a week. Or forever. What the hell.” You might even think you’re sticking it to someone by wallowing, sticking it to that little fat green, annoying owl, for example.

It’s the good old “what the hell” effect I’ve spoken about before. It appears in many guises but streaks really roll out the welcome mat for it. Duoloingo tries to mitigate this with “streak freezes” that let you preserve your streak by (essentially) paying them. It’s a little like the fitness equivalent of selling indulgences in the pre-reformation Catholic church. It’s a kind of solution, but then you start to wonder, what does a streak even mean anymore if I can buy/bullshit my way out of it? Do I really even have a streak? Did I ever? Now even the motivational upside of streaks is compromised. So “streak freezes” are a kind of solution, but an expensive one – and I’m not even talking about whatever money or points you paid for it.

So what do you do instead? Or at least, additionally? How do you track and motivate long term success in a way that doesn’t have this terrible self-sabotaging side effect of making future failures progressively more traumatic and harder to recover from the longer you succeed? In a way that motivates you to keep going both when you are succeeding AND inspires you to get right back up on your feet again when you fail? And, ideally, scales so that you can apply it to the many habits you are presumably working on?

Do not despair! Such a way exists. I call it Greenishness.

The name only came to me a few months ago, but I’ve been practicing the substance of it for many years, at least since my first Lifelog in 2016. And its antecedents go back even farther than that.

Longtime listeners of this podcast will be familiar with the Habit Traffic Light metaphor for tracking and motivating daily progress. I first podcasted about that in 2006, so almost 20 years ago now. Now, these days, it’s all over the place, but then it was at least a little novel. “Greenishness” builds on it, it’s like an expansion of the Habit Traffic Light for longer periods of time, conceptually.

The Habit Traffic Light, to quickly remind you, is mapping the colors of the traffic light to daily habit compliance: green is success, red is failure, yellow is exempt, like an S-day in the No S Diet. Many habit trackers, including my ancient but still functional habitcal tool, provide the ability to mark off days in colors like this, or you could use a spreadsheet, as I currently do in my “life log,” or even a physical, paper calendar. The Habit Traffic Light is great at the level of individual days and even for a month at a time it paints a very clear picture. If there’s any red, it jumps out. If it’s mostly green with some yellow, that’s obviously good. People often post screenshots of how their month looked in whatever habit tracker app they are using to the No S Diet facebook group and it works beautifully at that level. It gives a sense of accomplishment or, conversely, a sense of “yeah, I need to work on that.”

But longer term, beyond a month, it gets harder to read. And if you want to compare month to month progress or see how you are progressing over time, it’s harder to tell.

Enter greenishness. Or the next level of greenishness. Like everything else these days, it seems, greenishness is a spectrum. You could consider the habit traffic light itself to be the first level of greenishness, marking individual days green or not. Although, as I’ve mentioned, there are many ways to implement the habit traffic light, the more advanced greenishness techniques I’ll be describing benefit heavily from using a spreadsheet, so I’ll just mention quickly how to do level one greenishness in a spreadsheet too. Every row is a date. And you have a column giving the date, right? That’s first column. Then you have a column for every habit you are tracking. Success days you mark with a 1 and add conditional formatting to make them green. Failure days you enter as 0, and you add conditional formatting to make them red. Exempt days you just leave blank but you manually set the background color of those cells [to yellow] just to distinguish them from no data days. Right? Like, it didn’t happen yet, or I forgot to enter it. I like to do one tracking spreadsheet per calendar year, for a fresh start and so it doesn’t get too unwieldy, too big, too much data, though as I’ll explain in a bit, I calculate and carry over certain summary yearly averages from yearly spreadsheet to yearly spreadsheet.

The next level of greenishness (this is level two) is to move from individual days to focusing on a multiday average of green-ness, either a moving average of some number of days, or chunked averages for periods of calendar time like months or years. Now you’re focusing on the percent of green success days in a given period. I recommend using calendar chunks like months and years, because, one, I always like working with the calendar, if possible. It’s this great social convention to hang your own habits on. And because it just lends it more drama. It’s like every month and year is a new game or contest you want to do well in vs. the last, and sort of a bit of a rest, right? Also there are few extra cells of data to add. It’s not like you have to add another cell for every cell that you’re averaging, like you do with the moving average, day to day. If the app you’re currently using doesn’t let you do this, it’s super easy on a spreadsheet. I don’t advise distinguishing between red failure days and yellow exempt days at this level, just focus on green days, both for mathematical simplicity and because it’s motivationally helpful to give a bit of a nudge to keep the S-day count reasonable. So at this level, yellow days and red days are kind of the same, we’re just counting the number of green days over that period.

Here’s a simple example of how to calculate this, and even if your brain turns off whenever you hear math you should be able to follow this: let’s say I was tracking my No S Diet compliance for a month. For simplicity let’s say a February with exactly 4 weeks, 28 days, starting on a Monday, every day equally represented. I had one red day failure and two non-weekend “special” S-day, and all the normal weekends as S-days. That would be 28 total days minus 8 weekend S-days minus 2 special days minus 1 red day = 17 green days or 61% rounded to the nearest integer. 61% would be my green day percentage for that month, the number that I’m focusing on. A fine number, incidentally, and close to my historical average over the many years I’ve been tracking – a touch over, even.

So that’s level 2 of greenishness, calculating your green day percentage over chunks of time. Doing that lets you simply and meaningfully capture a sense of how you are doing over time and compare it to other chunks of time. You can see longer term progress.

Level 3 greenishness (and this is where we stop, level three is the highest level) is about making longer term progress not only visible, but in your face visible, and motivating visible. A spreadsheet is helpful for level 2 greenishness, calculating the averages, but it’s almost necessary for level 3. So I hope I can persuade the spreadsheet hesitant among you to overcome your fear and aversion and give it a shot. I’ll link to a simple, but real working example from the transcript that you can copy and use as a template for creating your own.

It works like this, once you’ve got those averages calculated, say, several calendar months, then you apply “color scale conditional formatting” to them. It sounds more complicated than it is. It involves selecting a column of data in a spreadsheet and having it automatically format the values on a color scale (and you do this all on menus where you just click on stuff) from (say) red to white to green. Good values are greener (and that’s either high or low values, depending on what you are measuring) bad values are redder and values at the middle are white. So shades of green to white to shades of red. It’s super easy. If you can calculate the averages, this part is even easier. You just select the column and choose the menu option and select what colors you want to apply to the minimum, middle, and maximum values.

Basically what you’re doing is turning your spreadsheet into a heatmap. The numbers are still there, but now they have this color dimension. It’s an elegant way to have a table and a chart at the same time.

Then instead of focusing on your streak you focus on how green (and not red!) the averages in your spreadsheet are looking, and if they are looking greener over time, over the months and years.

You can define explicitly what your bounding values should be when you set the color formatting, or you can just have it automatically take the highest and lowest values. To start, when you don’t have much data, you might want to manually set the bounding values. Otherwise, with just one cell, like one month of data, the color coding logic is going to have nothing to compare it to so it’ll be meaningless. But once you’ve got a few cells of averages to work with (and you could just wait until you have this) it’s good to let your actual performance provide the range to provide a built in reality check. Instead of hubristically setting some unattainable goal and living in a hell of red, your actual worst is the worst, your actual best is the best.

Why is this technique so powerful?

First of all, it’s powerful because it provides a failure-proof motivation system. With level 3 “greenishness,” a failure still hurts, still has an impact, but it isn’t catastrophic, it doesn’t completely derail you, and you have visual incentive to immediately shake it off and get back on track, to keep your month or year greenish or at least not (shudder) reddish.

But it’s also motivating when things are going well. You want to keep your green. Maybe even make it greener.

And – this is a subtle but important point – it’s also helpful in moderating your hubris. If things are getting too green, remember that now you’re raising the bar for what green looks like. Don’t raise it too high. You want to be able to keep on getting green. Not suddenly get so unsustainably far ahead of yourself so you’ll never be able to achieve it again. I find this particularly helpful to keep in mind with exercise, where it ties in brilliantly with Zeno’s Paradox Exercise Progress Plan of stretching out the feeling of progress for as long as possible. I don’t want to run so many miles one month, for example, that I can’t keep running that far, or farther, in the future. Though if you do, don’t despair, you could always hard code the bounds so that maximum green is some more reasonable number, you can crank it down. I personally experienced this a few years ago when I developed a bad case of runner's knee and had to drastically cut my running mileage, probably permanently. To keep me motivated, to keep me chasing greenishness, I reduced the top of the color range to something attainable in my current state. I lowered my goals in order to keep me motivated and doing anything.

You could say, why bother with the color? Isn’t the raw data, the number, the average enough? If you see that your average for his month is better or worse than last month, isn’t that rewarding (or chastening) enough? Maybe, if you’re a Vulcan. But remember the Habit Traffic Light: humans respond more strongly to approval and disapproval than to cold facts and that’s what the color conveys, a heat, a moral charge to the data. It evaluates the data. To riff on Hamlet: numbers aren’t good or bad, but color makes them so. And of all colors, green and red work best because of the powerful moral associations with them. I know they’re bad from an accessibility perspective, if you were designing a system for many users, because of red-green color blindness, but if you personally can distinguish them, use them for yourself, don’t throw away all that powerful psychological baggage that comes with them. Use them in your spreadsheet. If you can’t distinguish them, find the best stand-ins that make sense to you.

I tell myself: “Greenishness,” (sort of a tongue twister), “Greenishness shall be thy Motivation.” If I’m doing bad, if I’m seeing a lot of red, I want to get less red, closer to white, closer to green. If I’m doing well, if I’m green, I want to get even greener, but not too green, not too fast, so I can stretch out that feeling of progress for as long as possible. Ideally I want to get just ever so slightly greener.

“Greenishness” is kind of a mushy word. And the mushiness is essential to it. We want to be gently reminding ourselves that what we are after isn’t perfection or optimization or superhumanity but “good enough.” Greenishness conveys that. Couching it in quasi-biblical language is a way of giving it some more umph without compromising this essential mushiness (and also little dose of Everyday Systems comic pragmatism). “Let Greenishness be thy motivation.” A divine “good enough.”

Another benefit of “greenishness” is that it scales well. You can track many habits and metrics at once and immediately get a sense of how you are doing over time on all of them. And you get a sense of the gradations as well, a somewhat complex (in a good way) picture of reality with all those shades of green and red. It conveys an incredible amount of information rapidly and intuitively. And it works for a variety of habits and metrics. You can track and evaluate No S Diet compliance, step counts, meditation minutes, weight, time spent on creative activities or serious reading all in the same spreadsheet. For metrics like steps and weight with a wide range of values, not just 1=success and 0=failure, you can use color scale conditional formatting even on the daily level for immediate “chromatic” feedback on how you’re doing.

A final benefit of this approach, and this is really a benefit of using a spreadsheet at all rather than an app, is that you can upload it to ChatGPT or another AI and ask it to do further analysis – ask it to look for seasonality or correlations in the data that might help you identify situations you need to explicitly work on.

I imagine you could even use AI to generate the initial spreadsheet for you, customized for your particular habits and metrics. I personally wasn’t able to get it to do this without it being slightly but irritatingly off, but I imagine a better ChatGPT whisperer probably could, and probably anyone will be able to if we just wait a few months for it to gain a few more IQ points.

I’ll keep experimenting and let you know if I come up with something that works reliably. I considered trying to produce a lifelog generator tool, an online tool that would ask you what habits and metrics you want to track and for what time period and then spit out a customized spreadsheet with all the formulas and conditional formatting in place, but I think it makes more sense just to wait until your run of the mill chatbot can do it. My sense is that specialized apps of all types are on the way out now that we’ve got this generalized technology that can increasingly do everything. It’s like investing in a waffle maker when a star trek style bit replicator is right around the corner.

In the meantime, in the transcript for this episode, as I mentioned, I’ll link to a simple, partially filled-in 2025 lifelog spreadsheet that you can use as the basis for your own. I used some of my own actual data so you can get a realistic sense of what it looks like once there’s something there. Just overwrite it with yours. It’s got two tabs: daily, and summary. Drag the monthly formulas down to have them hit future months in the summary tab. As I mentioned, I do a new Life Log spreadsheet every calendar year so I copy the yearly values over manually every time I make a new one, in the same columns as the month ones. A nice side effect of having several yearly values to carry over is that they provide starter bounds for color ranges in the new spreadsheet when I have no daily or monthly data yet.

Spreadsheets – I know. They can seem scary if you’re not used to them. So if even my simplified example spreadsheet seems too complex, just start by tracking your raw data, your daily numbers. AI has already made it a lot easier, and will likely be able to do all kinds of magical things with it soon, even for the spreadsheet-phobic – but it won’t have anything to work with if you don’t start recording the numbers. It’s basically going to be a dream app factory, for whatever digital functionality you want and can at least stumblingly articulate, a Star Trek replicator app oven, but the one thing it will never be able to dream or cook up is a time machine to go back and get your missing data.

Well, that’s all for today. In sum: Beware the streak. fear ye not the spreadsheet, and let greenishness be thy motivation. Thanks for listening.

By Reinhard Engels