Adding Is Favored Over Subtracting In Problem Solving (nature.com) 90
A series of problem-solving experiments reveal that people are more likely to consider solutions that add features than solutions that remove them, even when removing features is more efficient. Nature reports: Across a series of [...] experiments, the authors observe that people consistently consider changes that add components over those that subtract them -- a tendency that has broad implications for everyday decision-making. For example, Adams et al. and colleagues analyzed archival data and observed that, when an incoming university president requested suggestions for changes that would allow the university to better serve its students and community, only 11% of the responses involved removing an existing regulation, practice or program. Similarly, when the authors asked study participants to make a 10 x 10 grid of green and white boxes symmetrical, participants often added green boxes to the emptier half of the grid rather than removing them from the fuller half, even when doing the latter would have been more efficient.
Adams et al. demonstrated that the reason their participants offered so few subtractive solutions is not because they didn't recognize the value of those solutions, but because they failed to consider them. Indeed, when instructions explicitly mentioned the possibility of subtractive solutions, or when participants had more opportunity to think or practice, the likelihood of offering subtractive solutions increased. It thus seems that people are prone to apply a 'what can we add here?' heuristic (a default strategy to simplify and speed up decision-making). This heuristic can be overcome by exerting extra cognitive effort to consider other, less-intuitive solutions.
Adams et al. demonstrated that the reason their participants offered so few subtractive solutions is not because they didn't recognize the value of those solutions, but because they failed to consider them. Indeed, when instructions explicitly mentioned the possibility of subtractive solutions, or when participants had more opportunity to think or practice, the likelihood of offering subtractive solutions increased. It thus seems that people are prone to apply a 'what can we add here?' heuristic (a default strategy to simplify and speed up decision-making). This heuristic can be overcome by exerting extra cognitive effort to consider other, less-intuitive solutions.
That (Score:5, Funny)
adds up.
Re: That (Score:2)
Re: That (Score:4, Insightful)
And they're quite literally adding to the problem when trying to resolve it.
That's government in a nutshell.
Re: (Score:3)
Yes this should be required reading for every legislator
Re: (Score:2)
Lol. Have you read your corporate handbook recently?
Re: (Score:2)
Today' headline: Biden proposes 16% increase in federal spending
Few get points for efficiency (Score:5, Interesting)
But plenty of people get points for new initiatives.
This is totally true in government, non-profits, and academic environments.
It's less true in for-profit environments when cold, ruthless, efficiency efforts actually result in greater net profits. This is often derided as cruel and/or short-term thinking by observers in academia, government, and non-profits. Sometimes this criticism is warranted and sometimes not.
If this study focused on university types for its test subjects, all it says is that universities favor addition over efficiency.
All we can say is that there is at least one sheep in Scotland, at least one side of which is black.
Re: (Score:1)
All we can say is that there is at least one sheep in Scotland, at least one side of which is black.
Well, it appears black from this perspective some of the time, anyway.
Re:Few get points for efficiency (Score:4, Interesting)
For-profit environments are no better at cold, ruthless or efficient. For example, car and oil companies are for profit, but they are highly reluctant to change.
It's not about profits, but competition. Companies don't change what they're doing unless competition forces them to. They'd just as much rather add things (patch over holes) than to make things efficient if no other competitor can challenge them.
Re: (Score:3)
Most of the time they don't change in the face of competition either. Companies tend to find a niche to exploit, grow, stagnate, then die when the environment changes or someone with a better idea comes along to compete with them.
Re: (Score:2)
Citation needed.
Re: (Score:2)
It's true in for-profit as well, perhaps more so.
Consider: Problem, processing the $0.50 /month widget fee costs $0.65. Best solution, strike the widget fee, crow loudly to customers. Actual solution implemented: raise widget fee to $1.15.
For profits are forever spending money on re-arranging deck chairs. The reason business logic is so hard to get right and costs so much to implement in a new system is that business rules accrete over time and never get carved away.
For profit hides this by outsourcing to s
Re: Few get points for efficiency (Score:2)
Free-market capitalism would solve this. If only we had something like it. I find myself strangely sympathizing with the die-hard lefties who insist that real communism has never been tried.
Re: (Score:2)
You can go all the way to anarchy and still won't solve this. A healthy market might tend to curb this, but but we are far far away from that. And the people calling for a "Free Market" are far from willing to do what it takes to have a healthy market.
When Smith spoke of healthy markets, he was assuming that corporate charters would be very rare and closely policed, that most sellers' economic power would be within an order of magnitude of most buyers' economic power, and that for any good, most marketplace
Re: Few get points for efficiency (Score:2)
Quite right.
Re: Few get points for efficiency (Score:1)
There's a reason they call them bean counters (Score:4, Insightful)
Re: (Score:3)
This is why you have to sneak into the bean counters office really late at night, and secretly change half of their addends with their twos-compliment form.
That way they won't realize the difference.
*ducks-and-covers*
Re: (Score:2)
Re:There's a reason they call them bean counters (Score:4, Funny)
I mean, you can't really count negative beans [youtu.be] or negative lines of code [folklore.org], now can you?
Math isn't logical to begin with. If you can't figure something out it's okay to make up imaginary numbers.
Re: (Score:2)
Sigh.
Yes, We Know. (Score:4, Interesting)
Anyone who uses ANY kind of ticketing software will tell you this.
We always start off with a wonderful program that does exactly what we need, when we need it. It's designed by Engineers, for Engineers.
But Management doesn't like it. It doesn't make pie charts. You can't export the metrics that are used to make pie charts. Customers don't like it. It doesn't use terms that the customer understands, even though the customer shouldn't be seeing half the stuff in the ticket anyway.
So a software program that allowed you to do something with just three key presses, or mouse clicks, gets "upgraded" to enable making pie charts and pleasing customers.
Now it takes you seven mouse clicks to do the same job. Not key presses, no; those are gone now, because they conflicted with the system.
Oh, and the changes they made broke the database, so they had to dump it and move to a new and improved software "solution" that's exceptionally good at making pie charts for management, and providing "updates" to the customer, which make up things based on how many times you click your mouse on buttons in the ticket.
So yes.
We KNOW about this fault in the human genetic code, and dearly wish we could REM it out.
Dependency back-tracing not easy (Score:3, Insightful)
I do find that when you remove software features, you often find out that someone or something relied on them, and they complain. Dependencies, both human and in software, are often hard to track down.
Re: (Score:2)
The GNOME vs KDE wars illustrate perfectly what this story is talking about.
Re: (Score:2)
They're really easy to track down. Just remove a feature and wait for someone to complain.
Somewhat less tongue in cheek, when they complain, you often find out that they've been doing something batshit insane and, provided you can convince them to consider new ideas, can streamline what they're doing as well.
Re: (Score:1)
I don't know how to fix grouchy users and customers. If I did, I wouldn't be in IT.
Re: (Score:2)
Re: Yes, We Know. (Score:2)
And you've just proved the point about the article. You've recommended adding "REM" instead of outright deletion of the offending code.
And One More Thing (Score:2)
Let me just add...
Point demonstrated by BeauHD (Score:5, Insightful)
Adams et al. and colleagues
The term et al. is short for the Latin et alia which translates to "and others".
Adams and others and colleagues? Really?
Additive "solutions" indeed.
I propose a subtractive one: pruning the so called "editors".
Re: (Score:3)
Not sure this is what the writer intended, but usually for journal articles "et al" refers to coauthors of the paper. So the writer could have been trying to say "Adams and coauthors, and their colleagues" (i.e. non-coauthors).
Or it's just bad writing/editing...
Re: (Score:2)
TFA uses both "Adams et al" and "Adams and colleagues", but in different paragraphs.
Re: (Score:2)
Not true. (Score:3, Funny)
I used deduction to determine they are wrong.
Laying down the law (Score:2)
Our legal code is testament to this. Or tax code. Or building code. Or university departments.
Name me one politician who campaigns on and actually votes for removing laws and getting rid of departments, and I'll show you a jackelope.
Re: (Score:3)
Name me one politician who campaigns on and actually votes for removing laws and getting rid of departments
Rick Perry campaigned on eliminating three departments: The Dept of Education, Dept of Energy, and one more that he couldn't remember.
Re: (Score:2)
And then ended up helming one of the departments he wanted to eliminate (DOE)
Re: (Score:2)
Re: (Score:2)
So...he has become the very thing he swore to destroy?
To be fair, before he was named as the head of DOE, he had no idea what they do.
He thought he was going to be a "global ambassador for America's oil industry".
The DOE's main purpose is managing America's nuclear weapons.
Idiot Tasked With Maintaining America's Nukes Surprised to Learn What His Job Is [gizmodo.com]
Re: (Score:2)
Re: (Score:2)
Re: Laying down the law (Score:2)
DOE = department of energy
ED = department of education
Re: (Score:2)
Not everywhere.
Re: Laying down the law (Score:2)
Ok, well that's what the federal departments call themselves. Sometimes ED is called DoEd.
https://en.m.wikipedia.org/wik... [wikipedia.org]
https://en.m.wikipedia.org/wik... [wikipedia.org]
Re:Laying down the law (Score:4, Insightful)
My problem is the argument always seems to be for more regulation or less regulation - never optimised regulation.
Re: (Score:3)
The LOC/regulation count would fall where it does.
Re: (Score:2)
Ten Commandments vs legal code is the most extreme example of addition and shows no signs of slowing. With that being said it would have been wise for our founding fathers to have made an institution that removes laws as a counter to one that makes them.
In related news... (Score:5, Funny)
Subtractams et al strongly disagrees with the conclusions at Adams et al.
Re: (Score:1)
Well, yeah. (Score:2)
You plan to make room for adding stuff as part of most projects. Subtracting, not as much - it's a negative space, plan-wise.
Subtracting stuff tends to cause more annoying side effects... because of all the assumptions that end up presupposing the things being removed elsewhere in the project- even indirectly.
Even in very carefully modular designed systems.
And testing is the most expensive phase in most projects, by a pretty large degree. It's why game companies for instance bend over backwards to... not
Re: (Score:2)
Oh, and I should probably also mention:
This isn't an issue unique to software. Even on the most common vehicles and manufactured products, you'll frequently find wasted space - costing billions of dollars to make, ship and maintain, because those spaces USED to hold some functionality that is either long obsolete, or unused from the initial design, but never corrected. It's right under all our noses all the time, in a huge variety of products all over the place.
But more than that - it's also not unique to
Re: (Score:2)
There are key difference though between those things and the law.
In the case of manufactured goods it is a question of efficiency. (at least cost efficiency) One the price tag for caring those legacy structures hits the point where its more expensive than redesigning the product or you can not manufacture your product as cheaply as the competition does, the redesign happens or the product is discontinued.
Same thing in the case of living organisms. Their is probably more selective pressure to add some adapt
"Custom" software install (Score:3)
Back "in the day" when your "hard drive" was measured in the 100's of megabytes, it was a difficult choice regarding whether you wanted the "complete", "standard" or "custom" installation?
And if you selected "custom", which features did you want to leave out?
There, I answered TFA -- subtractive choice is hard, even for geeks.
Re: (Score:2)
My first computer was back "in the day" when you didn't have a "hard drive". As was my second. My third had a hard drive....
advice for legislators - (Score:2)
When I lived in my motorhome I learned that before I could bring a cubic foot of groceries in, I had to remove a cubic foot of something to make space. This simple principle is lost on many modern humans. More is better and there is always room for more. And that included legislation.
I propose that before any new law is passed, two old redundant laws must be repealed.
That is true for my 1800 sq ft house (Score:2)
you ignorant clod!
Obligatory quote (Score:2)
"Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." â Antoine de Saint-Exupéry
Re: Obligatory quote (Score:2)
Presumably, a valid solution to the 10x10 problem would be to remove ALL the green squares.
Re: (Score:2)
Important to teach this (Score:3)
Whenever I'm mentoring new coders, I make it a point to impress on them the attitude that few pats of the job are satisfying as removing code. Be it old cruft by someone else or stuff we just wrote together yesterday that seemed brilliant then but can now be refactored away.
I see so many attacking a new requirement by throwing more code at it, when you could really just tune or even pare down the existing paths. Most frequently by juniors and intermediates of course, but all too often by people who really should know better. Participating in cleaning up old messes as part of *adding* a modest feature tends to leave an impression, as it did on me when I finally got to learn from someone with an appreciation of focus and of style in their code.
I'm not talking about conciseness for its own sake, and certainly not at the expense of clarity, but avoiding uneccessary complexity and noise that obscures the real intent
Re: (Score:2)
That last line is absolutely vital, and far too many people stop reading before they get to it. There are enormous numbers of coders out there who think that writing "slick" one-liners is far more important than writing comprehensible and maintainable code that might be just a little bit longer.
And removing "stale" code (Score:2)
doesn't risk breaking something along a path left out of your unit test?
Re: (Score:2)
There is always risk. But adding ever more spaghetti out of fear of breaking something actually increases risk in the big picture, as well as inflating development and maintenance costs even if you do manage to avoid all the pitfalls for a while. And makes work miserable.
If the tests are lacking, you need to find and fix that just like any other bug.
A solution to make everyone happy (Score:2)
Is it related with the risk? (Score:2)
If you have a working thing and you want to improve it, it is easier to treat the original as a black box and add new features around it - this way it is easier to predict how your changes will influence the outcome and ensure that you don't break the original functionality. On the other hand, if you want to remove stuff, you need to know the inner workings of the original thing very well. For this reason adding may be chosen (consciously or subconsciously) more often because of a smaller risk (especially i
Understandable (Score:3)
When faced with a man-made thing, the natural assumption is that everything done is for some reason.
When asked to balance the grid, the natural feeling is that the previous person didn't "finish" so the inclination is to finish the incomplete job.
As far as asking for improvements, and few people asking for things to go away, again, in part you assume someone put it in place on purpose and cares about it staying. There is some presumed motivation behind the effort to have that thing exist that you are inclined to be mindful of. You either have to know and disagree with the impetus for that thing or be massively inconvenienced to raise the question of 'why is this thing here anyway?'.
Its because Humans dont aim for efficiency (Score:1)
To solve a problem, the more features you try, the faster you will come to that solution.
The the very fact that this research exists shows you that once solved, Humans tend to be be quite objective in how we improve upon solutions.
I never subtract... (Score:2)
Findings simpler solutions is much harder (Score:3)
That is well-known to anybody competent in systems design (and anybody that ever wrote a good scientific paper with a low page limit): Finding a smaller, less complex and hence more elegant and resilient solution is much, much harder than just "patching issues" where you just throw things at any new problem that crops up until the whole mess seems to function. Of course, by the second very popular approach, you create a ton of problems later.
On a more general level, the reason for this is that in order to come up with compact elegant solutions, you have to know what you are doing and you have to understand how things work. Most people cannot get there. Most people want simple _now_ (i.e. just adding stuff instead of trying to find a good solution). That results in, for example, dumping lots of trash in the environment, consuming non-renewable resources with no plan what to do when they are gone, inability to do risk-management (which always has a longer-term scope), just ignoring problems and continuing as usual and denying that there is a problem until things (often literally) blow up, etc.
Minecraft Castle (Score:2)
How do you build a Minecraft castle? Do you find a nice flat place and add every block yourself? I know personally I like to find a mountain that has a Castle locked inside it and just remove/subtract blocks until you can see it.
And in other news... (Score:2)
And in other news: multiplying by ten is preferred over tenth-roots.
It's because schools teach arithmetic wrong (Score:2)
This makes complete sense as subtraction doesn't "exist". That is to say, you can accomplish subtraction using just addition with signed numbered, i.e. +5 + -4 = 1. You are always dealing with signed numbers, it's just no one likes to think about numbers this way since the schools refuse to teach mathematics this way.
Re: (Score:1)
I don't think this is right, I'd rather believe schools teach it the best way for human brain to understand. I have never heard of someone using negative adding as mental model for subtracting.
Meanwhile, in the real world... (Score:2)
In the real world, it's not just adding or subtracting green boxes. In the real world, things are complicated. That is why adding a kludge is far easier than simplifying (or "refactoring").
Def true for meetings also (Score:2)
"We are behind schedule so we are adding a mandatory daily meeting at 7:00 to discuss the plan for the day."
"We already have a daily stand up/scrum at 10:00, a weekly status meeting Monday at 1:00 so you can do your report, outside team coordination meetings on Monday, Wednesday, and Friday, the standard sprint meetings that come around, not to mention meetings on specific issues that pop up several times a week."
"That's for you folks, this will be to get leadership tied in."
"Wouldn't it be better to let th
Subtracting is hard (Score:2)
Given that things are there for a reason, removing it requires thinking about that reason. Adding new things is far easier than figuring out why everything that is there is there.
It's not just government (Score:2)
At work our users want more and more features, even when we say this is likely to cause more problems here and here, they say we must have it to solve our current problems , we understand and are willing to take the risks. Then when the new issues crop up, that we warned them about they act like they don't remember, even though we have digitally signed docs and an email trail where they replied and accepted the risks. I think this is just human nature, I'm not sure it can ever be fixed.
Subtracting Uni. Start of Semester Registration (Score:1)
This is why government sucks (Score:1)