How would you implement ELO for github users?
This is the obvious next step for AI, filtering content based on how competent the human(or bot) that produced it was.
I've got mappings from last name to country of origin so I can also give bonus points to based slavs for instance
deduck ponts four eveyry speling mistaek
code matters way more but ESL detection may be valuable
if they write good code it doesn't matter
Anyone with homosexual shit, BLM shit, Israel flags, or Ukraine flags gets a zero. The rest will sort itself out.
>hamas supporter
go back to X sweatie
Free Palestine
danes automaticaly get ELO doubled btw
swedes get 10 points taken off
What you are looking for is a metric for code quality.
Something like this
https://codeclimate.com/github/codeclimate/codeclimate
Yes but also not just code quality, for instance knowing how likely a user is to go AWOL will influence what libraries or modules i choose to implement in my software.
github users don't match up against each other in win / pier competitions so you can't use elo you dumbfuck
that's fine, i have plenty of objective metrics from linters, fuzzers, etc i can use to evaluate the quality of their code
for instance shitters who don't use typehints will be penalized.
YOU ABSOLUTE FUCKING MONGOL FOR ELO TO WORK YOU NEED TO MATCH PLAYERS OR TEAMS AGAINST EACH OTHER. USING ANY RETARDED METRIC TO RANK PEOPLE IS NOT ELO. JESUS CHRIST YOU'RE A SRUPING FUCKING CHILD
They will be matched up though, their code will compete against other code based on Objective Metrics.
could you google "how does elo work" before spewing out ill posed nonsensical thoughts
>noo you have to stick to my autistic definition you can't extend the idea using your human concepts of metaphors and innovation
the actual matchup, if needed, will be performed by using their code as modified to an code generating LLM which will then compare the results
>is completely clueless of the concept of elo
>nooooo stop being pedantic it can mean whatever I want it too
im pretty sure you're genuinely stupid and not a troll so here you go:
you want to use a computed metric to directly compare users, and then use the result to compute elo? you do realize this is completely pointless right? you already calculated a metric to rank users you're just propagating it though another metric, use what you calculated from the start retard. still they won't be matched up since they write code independently and not as a means to beat another user so elo makes no sense.
either way "objective metrics" for code quality don't exist and don't even try to disagree with this thanks.
>code as modified to an code generating LLM which will then compare the results
you want to train an ai with their code and then calculate the metrics on the code generated by the ai? that's so dumb
do some thinking and go ogling on you're own I'm not discussing this further
>you want to use a computed metric to directly compare users, and then use the result to compute elo? you do realize this is completely pointless right? you already calculated a metric to rank users you're just propagating it though another metric, use what you calculated from the start retard. still they won't be matched up since they write code independently and not as a means to beat another user so elo makes no sense.
dont care i know using the term will make people mad (like you) so i am going to do it
>either way "objective metrics" for code quality don't exist and don't even try to disagree with this thanks.
lmfao
C++: Not for trannies
Operator override is the most tranny of language concepts.
not since Bjarne Stroustrup chased out all trannies from the C++ conference through the sheer power of unrelenting disgust
If you submit a bunch of pull requests that are nothing but adding 'CONST', your account destroys itself.