One man's brilliance is another man's "who wrote this crap?" It is indeed quite likely that most of us, possibly even all of us, will fall below the median of someone's measure.
They're making a math joke. It's impossible for more than half of the sample to be below the median since the median is constructed so that half fall on each side.
Having 60% of a sample below the median means someone calculated the median wrong.
> It's impossible for more than half of the sample to be below the median
No. It is very much possible when the measure is undefined, as was the condition setup before the bet was made.
If we take it to the logical end where your chosen measure is simply the inverse of my chosen measure, then those above the median by my measure fall below the median by your measure and vice versa, leaving the entire set of those outside of the median to be seen below the median (and also above the median).
As it pertains to the larger discussion, how the quality of a developer is measured has not been established, which means everyone is likely to measure differently and often in conflicting ways, and so most of us are bound to be below the median according to someone.
So, yes, this was originally pointed out in a humorous manner. The bet made on the basis of that joke was, indeed, unfortunate.
> No. It is very much possible when the measure is undefined, as was the condition setup before the bet was made.
I promise you, it's not, because that's not how medians work. This is the Wikipedia definition of a median:
> In statistics and probability theory, the median is the value separating the higher half from the lower half of a data sample, a population, or a probability distribution.
There have to be half on each side or it's not a median. The mean doesn't have to be distributed like that, but the median does. It doesn't matter if the measure is undefined; the median divides the upper half from the lower half. There are always 50% on each side (and one on the median, if there are an odd number of samples).
> If we take it to the logical end where your chosen measure is simply the inverse of my chosen measure, then those above the median by my measure fall below the median by your measure and vice versa, leaving the entire set of those outside of the median to be seen below the median (and also above the median).
Sure, but that's not a median. That's the intersection of people above 2 separate medians of 2 separate measures. You can't combine the median of separate measures and have it still be median; you can combine the measures into a single measure and that has a median where half are above and half are below.
> As it pertains to the larger discussion, how the quality of a developer is measured has not been established, which means everyone is likely to measure differently and often in conflicting ways, and so most of us are bound to be below the median according to someone.
Totally agree there, I'm just beating a pedantic, dead horse that that's not a measure (it's a bunch), and it doesn't have "a" median (it has a bunch). You're talking about the intersection of people above the medians of a bunch of measurements, which doesn't share any of the properties of a median or being above/below it.
> Sure, but that's not a median. That's the intersection of people above 2 separate medians of 2 separate measures.
And? That's a given. The comment we are discussing made it clear that the are multiple measures, and that most developers will fall below the median (in at least one of those measures). Not that difficult to understand.
> You can't combine the median of separate measures
Sure. Which is why nobody did. What are you trying to add with this strawman? If you are simply trying to refresh your memory of your kindergarten-level math classes, I'll chuckle for a minute over the foolishness of me thinking you were trying to have a discussion and leave you to it.
If you're arguing for some quantum metric that could make me lose it, you're for some reason not providing the word "us" with the same leverage which puts us into a situation we should just default to the literal definition of median since there's literally no way to prove (or disprove) your point.
But sure, you've proven you can buzzkill anything, so you win!
As was said by the parent commenter (drewcoo) just before you placed your bet, because no measure is established, each evaluator is bound to have a different opinion of how to measure a developer, and as a result most developers are likely to fall below the median (according to someone's measure).
There is no mathematical contradiction. Perhaps a misunderstanding on your part? Either way, the bet was lost. Time for drewcoo to reap the rewards.
Or rather: I’ll mortgage my house and put my life savings on the line against that bet.