If someone is described as "good at something", in your opinion how good is that person at this activity?

It’s always in relation to what the speaker considers average in that situation.

If someone at home is good at cooking, they could make nice meals out of fish while everyone else in that group can just make porridge. If someone at work is good at autocad, they can make technical drawings while everyone else can just barely read them.

Comments like that say far more about the person saying it than about the person being described most of the time, I'd say.

I'd need to know how good the describer is like in that area before I could make any assessment about the describee.

You are probably underestimating your abilities.
People that worry about overestimating their skills mostly underestimate their skills.

If someone says they are "good" at something, I take it to mean competency and some enthusiasm.
They might make a mistake, but they won't (or at least will rarely) make it twice.
They know how to find the solution to something within that domain of knowledge. It might not be the best solution, but it will be a solution that works.
They are also aware of what they don't know in within the domain. So, they can do C++ but know they can't do embedded programming. Or they can do C#, but know they can't do game dev.

And I would take them at their word for that, until they prove otherwise.
If they are below where they claim their skill is, I would try to help them learn (unless they show no interest in improving).
If they are above where they claim, I would tell them this.

It's always hard to judge our own skills.

Personally if someone was described as "good", I'd take it to mean they could do it at an expected level (not going to hold the team back). If someone was above average at the task then I'd expect a different adjective, e.g. great or excellent.

I assume competent. If you're "good" at it, you can do it without much more than the expected amount of help.

If you're a good runner, I'm not expecting you to win, but I expect you'll be able to finish it.

1. Have you invested time learning how to do/improve the skill?
2. Have you seen progress in your skill level?

"Good" is relative

Depends on context and the skill in question. For example, in my office, good at technology means you know how to submit a ticket to IT and you don't have a panic attack if someone mentions excel.

It's context dependent, not field dependent. "Good at [something]" usually means it's simply above average. In some occasions, it can mean it's just average as opposed to terrible. In other occasions it means it's exceptionally good.

For your specific examples I would believe they're somewhat above average, but not impressive, unless more context indicates so.

In my opinion it really depends on what they are talking about if a person says they are good at cooking I assume they are slightly above average if a person says they are good at working on cars I assume they are better than 95% of the population

In my opinion it really depends on the person saying that. There are just way too many factors like Dunning-Krueger, the interpretation of the word 'good' (Just sufficient? Above average?), or just their own ego/modesty.

I would say I'm decent with numbers. I don't like to boast, and I'm sure I'm terribly far from being a prodigy (I am quite far), but I have the feeling I can multiply double digit numbers in my head faster than the average. That doesn't even tell you much anyway.

Depends on who is talking and the context.

If I go to a WH40k night at a FLGS, and someone points to a specific guy and says they're good at 40k, I'll probably take their word for it. They're in an environment where they've been tested regularly and that guy probably is one of the better ones at the shop. Are they able to play at a major tournament and place well? Maybe, maybe not.

For a new hire for programming, depends on who is saying it. Were they in the team for the hiring process? Do they have a track record of picking up good talent?

I also have a working theory that it's not too hard to better than 75% of people who do a given thing. For example, here's a breakdown of Chess.com ELO ratings (taken from the other site, dated April 2023):

Getting to 75% would be a little over 900 ELO. Which is interesting, because not only is that fairly low, but it's also below the 1200 that you typically get assigned as a new player. Though Chess.com does assign that based on your self reported skill level at signup.

Bear in mind it can depend what country someone is from.

Traditionally, a Briton saying "I'm kind of okay at this" might mean they're one of the best in the world (mustn't blow one's own trumpet).

A Briton who says "I'm great at this" (or anything beyond "pretty good") is likely an arrogant charlatan.

In contrast, an American might say "I'm good at this" to mean "I am better than average".

I'd imagine other countries have their own tendencies for under/overstatement.

A lot of people have already pointed out that the person saying it is an important detail. If your mom says you're good at python, I'm going to get a very different idea about your skill level compared to your boss saying it.

Unless you are selling yourself in an interview context, I think it's very poor form to qualify your own skill level. Let the skill speak for itself, it's enough to say that you 'do python'. Saying you're good at something often comes off as braggadocios more than it is informative. If you must give context, it's better to talk about how much experience you have, or other objective metrics

Qualifiers are too context dependant and no matter where your skill level is at, you wont ever have enough context to know how good you actually are, because you can't know what you don't know.

Above 50-percentile in comparison to others who also participate in that subject/hobby/activity, but usually below the 90-percentile, because then they'd not be described as "good", but "great"

Related: https://en.wikipedia.org/wiki/Impostor_syndrome

Better than me.

Imposter syndrome just limits you. Nothing wrong with fake it 'til you make it. If you're not doing it, someone else is. Just base your opinion of them on what you see, and act the same regarding other's opinions of you. Actions speak louder than words.

If someone ever claims to be good at something, I would want to see a demonstration that examples their expertise. Because it is hard to take people at their word sometimes, so a demo is required.

Somewhere in the range of not actively shitty/not shitty enough to notice to moderately proficient. If they do the task no one is likely to notice something wrong.

5/10. Understands what's going. Could certainly be better, but fuck ups are uncommon.

Competent and able to handle everyday tasks with ease.

For anything considered complex, able to give a good start and potentially work a partial solution (for running, they can do a good half-marathon without too much trouble).

Not a world-beater, not someone who can handle any task trivially, but someone who can definitely contribute to even expert/Ph.D level fields, if only at at a lower level (ie doing tasks at the direction of the lead researcher/engineer).

If they're "really good", I expect all that, and that they can self-lead for complex tasks, with the potential to grow into the latter role of "lead".

I am, at my worst, "really-good" at a lot of things, but I've also met a few rare people who make me feel like I'm in that half-marathon while they're doing a 100m dash, and they just never slow down. That's rarely a question of skill or competence for me, it's more focus and dedication to a single task.

8/10