A statistical analysis of the work of Bob Ross (2014) | Hacker News Reader

A statistical analysis of the work of Bob Ross (2014)

fivethirtyeight.com

72 pointsfrumpish22 comments

Growing up in Alaska I had always taken it for granted that when Bob Ross painted outdoor scenes they looked just like the outdoors I was used to. It's funny how that never stood out enough for me to see it as peculiar, given that absolutely no other media showed landscapes that were similar to the one I grew up in.

Having grown up east of the Mississippi River, I always assumed they were fantastical, imaginary landscapes.

Ha! I love it.

I was surprised that deciduous trees edged-out conifers. My (completely uninformed) mental image of Alaska is "Mostly pine, with other conifers sprinkled in." Can you confirm that deciduous trees are prevalent in Alaska?

Lots of them mixed in. We get a ton of alders and birches, but in the winter it looks like everything is coniferous b/c the leaves are off the others. Where I was it was lots of spruce, but I grew up in the coastal rainforest in Prince William Sound and Alaska is a big, big place.

Thanks for answering!

So 10 years ago this author manually examined all the photos and tagged them to find the conditional probabilities and all the neat stats-blog content.

I'd love to see a redo of this post using today's AI tools. Like, I know midjourney can spit out images "in the style of bob ross", but I mean how do you use these tools to analyze features across the bob ross corpus and ask "How many pictures contain happy little clouds?" And can these tools answer more abstract questions like "what kinds of bob ross paintings tend to have happy little clouds vs paintings with no clouds?"

Oh yes we can! Its already a component of these generative tools:

https://huggingface.co/spaces/pharma/CLIP-Interrogator

You might want to finetune it to reliably recognize "happy little clouds" with some samples of what you consider to be happy little clouds, as well as not happy and not little clouds for regularization.

Also, see LLaVA, aka CLIP hooked up to an LLM, if you are looking for more verbose and logical analysis of images: https://llava-vl.github.io/

Funny thought: this write-up will probably be instrumental for calculating future valuations of Bob Ross paintings should his estate ever consider selling them.

I wonder if that's why they agreed to it?

Right? Suddenly I want one of the two with people in them, or the one and only chimney!

Love Bob Ross, but that's a sad little chimney. Actually a vent pipe.

Remember that just because you know that 19% of the painting have snow and 7% have a beach, you do not have a good estimate of how many paintings have both snow and a beach.

(There is one which might qualify, but upon viewing, I don't think it actually has a beach in it.)

I recently saw a color breakdown of Bob Ross videos on reddit: [OC] Color Frequency in Bob Ross’ The Joy of Painting https://www.reddit.com/r/dataisbeautiful/comments/1330smg/oc...

And then when looking for some other items found a Tableau report of all of the elements: https://public.tableau.com/app/profile/calloni/viz/TheJoyofP... (which has the additional neat feature of linking to the corresponding videos that make use of that element)

> “The majority of people who watch Bob Ross have no interest in painting,” she said. “Mostly it’s his calming voice.”

There must be a correlation when creating a calming narrative with statistical features of the paintings: Painting two or more of a thing allows you to talk about friends, not being alone etc.

But somehow the friend effect only applies to natural objects. Charming little cabins don't seem to need a friend (or a chimney).

Charming little cabins (in the woods) are where friends go to hang out under the happy little clouds next to the happy little bushes while viewing the waterfalls

Perhaps part of the calming effect was being alone with Bob Ross. Adding human characters to the scene might invite people to project their stressful relationships or lack of relationships.

Not quite the same thing, but I do have vague memories of him sometimes adding wildlife--a squirrel here, a fawn there. Am I misremembering? Or maybe he talked about how an animal might enjoy a space, but didn't often actually paint them?

My first impression when i saw that grid of paintings was that they could have been made by DALL-E.

Bob Ross wasn't the most skilled painter, and he would probably be the first to admit to that. He had a certain skill set that lent itself well to certain styles of paintings. Now, he developed that skill set because he liked pictures like that, so it's not any knock against the man at all.

To my knowledge, he didn't to portraiture or urban landscapes or people in general. Few, if any, animals.

Mountains, trees, waterfalls, sunsets, sunrises, moonlight reflecting off a lake. Occasionally a house or barn or shed of some sort. All very nature oriented and pastoral. He wanted to bring what he saw as the peace of the Alaskan wilderness to people.

But it does lend itself well to the sort of generative content created by DALL-E. Bob Ross had a method for painting mountains. A lot of his mountains are fairly interchangeable. He had a set of trees. He had techniques for river banks and flowers, etc. Bob Ross was in a very real sense, sort of a human DALL-E.

Isn’t the point of Bob Ross the journey of painting, less so than the result? In that sense, it shouldn’t be surprising that originality isn’t a variable he’s optimizing for with his output or television programme.

> Isn’t the point of Bob Ross the journey of painting

The joy of it, even.

The godfather of ASMR.

Reply on news.ycombinator.com