The 90% Gravity Problem: Why We Tend to Quit Right Before the Finish Line

The final stretch is not the same as the first 90%. You don't know what you don't know at the start, you don't have formed ideas or boundaries to butt up against and so progress is easy. As you create, you have gained more knowledge and now have to consider you surroundings, the problem space becomes less open. The closer you get to completion the less space to work in, you likely have greater taste, ideas which are good enough for the first 90% are no longer good enough. So better taste and far less acceptable solutions due to increased dependencies. You also can see parts which were previously acceptable as now not so.

To top it off you have the emotional and ego side at play near the finish line. Is this good enough? This could be done better, etc.

I think that mixture of better taste, more dependent parts and ego make the last part the hardest. I also feel that the finish line being close isn't a strong of a motivator as ego is a demotivator. Whereas, at the start ego has no effect as you don't know anything, you can't be mad because you're new, it's all one big playground.

I think this pattern is quite common among certain types of people. We start things but never finish them. There are many reasons for this, but for me it’s a combination of procrastination, loss of enthusiasm, and diminishing returns. However, it's difficult to define the finish line because we often change our goals. I have a pet project: applying a genetic algorithm to optimise various problem fields. Initially, I just wanted to design a program that was generic enough for numerical optimisation. Then I wanted to experiment with different forms of genomes, and of course the question of efficiency always comes up. Even though I probably can't discover anything significant that other researchers haven't already found, it's a personal challenge that pushes me to find my own insights, which can take a long time. The next challenge is applying GA to image generation, trading algorithms, and so on. So, the project is never finished and is always in a state of flux. But I don't think this is necessarily a bad thing unless it's a paid project. :)

I am writing to add two other possibilities.

Within myself I notice that the project becomes boring when there is nothing new left to be learned from it. Depending on the project this could happen at 50% completion or 90% completion. Take scientific research for example. For me there is a lot of motivation to figure things out, to fill the gaps, to make sure everything is solid. But then there is the mundane part of putting it into text and publishing. And my energy is not in there. I already know what will go into that paper, I know getting it out will count as "success" and I know it should be shared. But my libido is not in it.

Another thing - the end of a big project signifies a big change. If you worked on something for a long time, what will you do once it's finished? Norman Finkelstein in one of his interview put it like that (paraphrasing): "I think some people genuinely don't want to end the conflict [between Israel and Palestine] because they built their whole life around it. In the past it was a problem for me as well. I have spent my whole academic career writing about this conflict. I read enough books to fill this room. Literally. If the conflict ends tomorrow - what am I going to do with my life?".

Maybe the projects weren't 90% done but more like 50% and you stopped simply because you rely on motivation rather than discipline.

Things only get more difficult as any project progresses so it's only natural that you will need more effort to reach the end. Novelty is long gone since the project started and now only hard work remains.

Plenty of people talk about this. Some people call it the valley of despair, pit of failure, etc. There's even a book called The Dip.

It is easy to put off most forms of evaluation until the project is fully done. Aside from not wanting to face evaluation or unease that something wasn't as good as anticipated it also means an evaluation on many much earlier actions one no longer feels represents their new experience and how they wish they had approached the problems in retrospect.

There are two books on the issue by a guy called Steven Pressfield, "The War of Art" and "Do the Work". I've read the latter and it's good. He talks about how resistance strengthens the closer you get to ending a project. I've found it useful in my work.

I've fallen for this trap a few times before. It's a collision of pareto principle and nirvana fallacy. Diminishing returns dominate the last 20%, causing burnout, and you're held in the tension between sunk cost fallacy and perfectionism. The former turning burning ambition into an anchor around one's neck, and the latter slowly decimating any hope of release with perpetual scope creep.

The best strategy to avert this the next time around is to vastly undershoot on goal size. The scope creep will still happen because we can't change our nature, but the project has a much better chance of reaching completion.

A few things come together I think:

20% of the effort gets you 80% of the way there. That means the last 20% is going to take 4x more effort than everything you’ve done so far.

You’re losing excitement and motivation, especially as you see the finish line recede farther into the distance.

You question whether the project was ever worth it, whether you’ll fail after all this work, whether it will be embarrassing, maybe worse than if you had never done any of it.

By this time other tasks are calling your attention, now you’re giving up real possibilities to work on this quixotic, possibly foolish project of yours.

The most-difficult parts of projects are starting them and ending them.

In my experience, the culprits are lacking a clear rubric for completion; activity inertia; fear of failure; and perfectionism (which I find is closely related to the fear of failure).

LATER ADDITION: By "rubric for completion", I mean the answer to the question When and how will I know it's done and ready to ship?

There's definitely a zone of despair on most projects I've been on - particularly greenfield projects - the whole thing seems misguided, the codebase is no longer pristine, often I come up with some horrifying realisation of an overlooked technical issue. It feels like we're moments away from failure.

So far it's never proven to be true and, yes, it generally turns out that we're incredibly close to delivery.

I think it's because we get our noses rubbed in all the small problems of the codebase at this point - before that it's not working well enough to see all of the issues, we're more focused on delivering some big chunk of functionality at all to perceive the smaller flaws. Because the small flaws outnumber the "bug chunks" stuff it feels like we're going backwards, but we're not.

It's a bit like essay writing. You get fed up with the whole thing when it's nearly finished because you've looked at it SO many times that it feels boring and all the little typos and grammatical mistakes take forever to fix up and every time you do you find more. It's hard to keep in mind that the problems are getting smaller with each iteration even though they're increasing in number.

Does that sound like the thing you're talking about?

I agree with the comments, when I read it I also had to think of the Pareto principle at first.

for me its the praise you recieve as children, from significant adults, teachers, parents etc, that set you up to follow through with your dreams.

people still manage to acheive their goals in life, but it does seem to be more difficult.

During my training as a psychotherapists, I reflected back on this exact same theme.

I had been a martial artist when I was a young man and wondered why I had only acheived a brown belt in both JKA and Aikido. I was very capable of a black belt in both.

I started looking at other areas of my life, like education and work.

I had been extremely intelligent as a child but did not go to school for the final year. So failed to take any school GCSE's.

Same thing, I had no drive to acheive anything higher than my socio-economic status allowed, pre-determined by societies expectations. how can a working class boy become part of academia and get a degree. its not easy.

I came from a very poor family, so poor in fact, I never had a coat and had holes in my shoes until I left home at a very young age of 16.

I went back to college at 36 years old. My therapist at the time supported and guided me through this period. To deal with this exact issue.

I had this overwhelming feeling of never being able to be late. So I arrived far too early for all appointments. I would arrive at least an hour early. The vacuum of unused time, waiting for an appointment, was a hell for me. it got so bad that I turned around and went home. This could be for a work interview or even a date. I missed out on so much of life.

The trick I learned, was to find something productive to fill in that vacuum, like go to a coffee shop and do the times cryptic crossword, while people watching. until the time of my appointment had arrived. such a simple application of filling in time.

Throughout my 6 years of study as an undergraduate, almost all of my peers experienced the dreaded final months of our degrees. Both as under graduates and post graduates. it never seemed to get easier. the stress, the doubt, the feelings of giving up, were a daily struggle.

Out of the 12 peers that underwent my last post graduate degree, all of us experienced this. yet we all went on to acheive our goals.

It is linked to the fear of success. but more in the realm of "not being good enough".

Society has great difficulty accepting "being good enough" as a valid way of being, its all a human being can aspire to, yet society wants us to be more and better.

is it no wonder that the youngsters of today are struggling with more depression, anxiety and other psychiatric issues, than in my younger years in the 60-70's.

They are set up to fail by constantly being fed that they are not good enough.