The Ballmer Peak: An Empirical Search

by slooonz • 2 years ago

88 24 2 years ago

Y_Y• 2 years ago

Of course the finding from this paper of the optimality of slightly less than two drinks has long been known to certain elite segments of society[0].

[0] https://www.youtube.com/watch?v=VTSCppeFzX4

willis9362• 2 years ago

It's nice to have statistically meaningful data to back it up. I recognize that hallway from MIB. Is it some famous NYC location?

rob742• 2 years ago

That being a BBC production, it would be a bit unusual for them to travel to NYC for filming this. But who knows, maybe they decided it's a good idea after slightly less than two drinks...

yatac42• 2 years ago

Note that they're actually citing "Mitchell, Webb et al." in the paper.

mannykannot• 2 years ago

Also footnote 5: "We originally used Microsoft Word for this but it kept putting the entire document into 8pt Times New Roman whether we wanted it to or not."

rob742• 2 years ago

"My God, you're the Inebreati!" - I have to remember that...

josh-sematic• 2 years ago

The xkcd comic about the Ballmer Peak: https://xkcd.com/323/

Havoc2• 2 years ago

I suspect it’s somewhat dependent on the starting mental state. ie might benefit someone who is tense more than someone that is relaxed.

I could totally see the relaxing effect outweighing the impairment for someone tense

kijin• 2 years ago

You can certainly code faster when you have higher confidence and lower inhibitions. Stop worrying about correctness in edge cases, security, or long-term maintainability!

Havoc2• 2 years ago

Maybe, but I've got a feeling there are some cases where it is actually objectively better not just "having a beer & doing some happy path coding" type better

..admittedly I think that window is vanishingly small...but I do think its there

jlawer• 2 years ago

Guess we need to do a follow on study, what a way to use the R&D budget. I don't think I will be short of volunteers.

I look forward to publishing the differing effects of XXXX Gold vs Stone & Wood in order for us to optimise the department budget. However I will be waiting for further research before starting trials of Bundy Rum. I am concerned the development efficiency will be offset by repairing punched screens.

430958942689• 2 years ago

In the same spirit, as mentioned in the abstract of the paper:

https://www.imdb.com/title/tt10288566/

mobeets• 2 years ago

This paper was published in this year’s issue of sigbovik, the best satirical academic journal you (may) have never heard of: https://sigbovik.org/2024/

progbits• 2 years ago

I hope that means a new Tom7 video soon.

Edit for context: https://news.ycombinator.com/item?id=36846820

hprotagonist• 2 years ago

he had a paper, but possibly no video this year.

pcthrowaway• 2 years ago

Oh wow, they finally solved the tabs vs spaces indentation debate in pg 102 of https://www.sigbovik.org/2023/proceedings.pdf

"Maximizing Code Readability Using Semicolon Indentation"

Though I guess end-of-line semicolons in Javascript is the next open question

;;;;;;;;5.2;;JavaScript
;;;;;;;;;;;;With JavaScript, the practice of omitting semicolons from the ends of lines is one that may be divisive when implementing these principles. One of the primary controversies within JavaScript circles is whether to use semicolons or not, as JavaScript has optional semicolons.
;;;;;;;;;;;;This proposal does not specify whether semicolons should be used at the end of lines, so as to not be controversial in this space. With this in mind, we can see this proposal in practice with a snippet of the elevator.js library[3], with and without end-of-line semicolons.

    // With end-of-line semicolons
    ;;;;element.attachEvent("onclick", function() {
    ;;;;;;updateEndPosition();
    ;;;;;;document.documentElement.scrollTop = endPosition;
    ;;;;;;document.body.scrollTop = endPosition;
    ;;;;;;window.scroll(0, endPosition);
    ;;;;});
    // Without end-of-line semicolons
    ;;;;element.attachEvent("onclick", function() {
    ;;;;;;updateEndPosition()
    ;;;;;;document.documentElement.scrollTop = endPosition
    ;;;;;;document.body.scrollTop = endPosition
    ;;;;;;win

DonaldFisk• 2 years ago

Thanks. That cheered me up, and I've only got as far as https://www.sigbovik.org/2023/proceedings.pdf

Eddy_Viscosity2• 2 years ago

The biochemistry suggests there might be a thing. Alcohol slows down brain activity. This may seem like it would make coding harder but if your brain is racing and thinking of too many things at once then this slowing mean the brain starts to, with a task manager analogy, stop a bunch of extraneous tasks. This means you can focus more. The trade off is that you also have less CPU power for that task. But, many coding activities aren't that hard and so the increased focus more than counteracts the loss of power. The trick is staying in that very narrow band of blood alchohol where this holds. Too much more and the loss of brain power means the coding tasks will not go well at all.

k__• 2 years ago

We had a cocktail afternoon at the last company I worked for.

I had some brilliant ideas after drinking, but looking at the code sober next day revealed, I wasn't brilliant.

hardlianotion• 2 years ago

Incredibly self-serving paper.

hnfong• 2 years ago

The relationship between alcohol and programming performance is interesting enough that I actually skimmed through the paper and tried to figure out how legit it was.

Sadly there's only one test subject.

So basically this is a report of somebody drinking a couple glasses of alcohol and then went on to practice leetcode questions...

hnthrowaway0328• 2 years ago

A little drink always temporarily removes my sour temper and everything else gets improved.

The Ballmer Peak: An Empirical Search

Related Stories