there should be a separate rating that only captains can see that tell you if the player youre picking took a bong rip just before queueing

no servers

beating a dead horse but pugs just shouldnt last this long

https://i.imgur.com/GA4w9Do.png

https://i.imgur.com/uX4rcCR.png

Dolphin said that invite players wouldn't like shorter games when i asked him to change it :(

https://i.imgur.com/Ib0J1zZ.png

bro wtf

genuinely i just got fatkidded by the website in OPEN pugs 5 times in a row. FIVE pugs in a row i didn't get to play. my elo isn't even the lowest in the queue. players with lower elo got picked before me, who had ALREADY played the previous pug. i'm 110% convinced now that open pugs should just function like tf2pickup. waiting in line for no reason is so boring and the same people just play the pugs on repeat

will consider!

You're right and that never should have been the intention, should be fixed to prioritize earlier added players now!

thanks, sorry if my post sounded rude, i was mad because it was late and i wanted to play before i went to sleep lol

the site will do the "ready up" thing with 12 players queued for correct roles, but seems like it will only start when there's 13 people queued. pls fix, weird bug.

There are only two hard things in Computer Science: cache invalidation, naming things and off-by-1 errors.

pushed fix

https://cdn.discordapp.com/attachments/585291733890236419/1266207510327459983/image.png?ex=66a44f23&is=66a2fda3&hm=72e83815fe1c2f844c9963bfb399b0b467c97c8b654a5d650c74b287ea2e592c&;

Hey guys add up.

you still lose elo from winning

add duo queue to open pugs

tf2pickup.eu has duo queue as long as one of you plays medic

no more open pugs

Can we also please have more than 3 servers at some point? We're at a point where the site is reaching pretty good traction but now we're having problems of running more than 3 6s pugs or hl pugs not being able to run because there are 6s pugs currently running taking over all the current servers. If tf2center and every lobby site imaginable can get around this, then RGL should to.

tf2center lobbies have to have the server provided by the creator of the lobby

Makes sense but my point still stands, we need more servers for this site.

we gotta get rid of open pugs. this shit is unbearable. how is it that we dont pick the teams but if we play no matter what win or lose i can still lose elo despite winning. the mfin BOT PICKS THE TEAM.

https://imgur.com/a/szWf8PC

I DONT GET IT. GO BACK TO CAPTAIN PUGS.

it is a little weird how the picking algorithm sees that the favorability is like 75+% in favor of one team and still picks those teams

there is a serious problem with the open pugs. mostly in the fact that team balancing can be fucking horrible, I have had pugs where my combo scout is a new comp player and the other combo scout is an invite player. same thing can be said for demo/med, the med problem isn't as easy to solve since so few people add to the class.

here's a few egregious examples of this in action:
https://logs.tf/3684805 (p scout, demo and med diff also sniper is cringe please get a life.)
https://logs.tf/3684810 (demo, med diff)
https://logs.tf/3691366 (demo diff) https://logs.tf/3689479 (med diff)
https://logs.tf/3688725 (whole team diff HOW IS THIS BALANCED?)

I know these logs are kinda spread out and have different root causes but, some of these classes(especially p scout, demo and med) should have some sort of weighting to them in the picking so the other team doesn't get just completely rolled. to add onto this the whole favorability thing needs to be fixed, I should NOT lose elo from WINNING a pug. TY

Also, I am not sure if you can give the Europeans without RGL history ELO based on their etf2l because those Europeans start with like 1200 ELO even if they have played premiereship they weigh down their team favorability and make it harder to gain ELO against them

dingle hingle vs logan

balanced

i think invite players should go play captain pugs instead

I am kinda curious what the picking algorithm actually is. It doesn't seem to follow traditional logic of how you would implement things like this (highest elo in the pug gets picked last, etc)

the bot has to be broken, as of the time of this comment about 70% of the pugs are like 20/30- 80/70 favorability on avg...

why?
logically you just switch a couple players to other teams and its more like 55-45 even with the same 12 players

Hi Dolphin, I'm also curious on how it's currently picking teams (partition problem? max flow algo?) and calculating favorability. You're probably already looking at how predicted favorability compares to actual outcomes, but would be nice to see published stats (vs some anecdotes in the thread).

I'd be happy to help with developing the pug site / team picking algo (though it appears you have enough devs atm; applied anyhow so you have my info), or just chat if you'd like. My discord handle is Antecedent.

That hurts…

So, I think when there are more than 2 pugs running, the balancing is worst. Like a mix of fat kids needing picked and less players to choose from in the pool, and you've got a dog shit pug. Maybe some sort of high and low open pugs to try and keep the skill level somewhat fair. If one pug needs additional players people can be brought up or down. Idk how hard that would be though.