Can you invent a balanced, non-retarded matchmaking system, /v/?

>>386128949

matchmaking sucks.

Just give us dedicated servers

>>386128949
you get matched with other players of your level and you have an even 50% chance of winning

btfo

Server browser or bust.

>>386128949
You only get matched against players with your exact winrate

Hacking is encouraged

File: latest[1].jpg (145KB, 1024x768px) Image search: [Google]

145KB, 1024x768px

Yes. It's called Halo 3.

>>386129145
Halo had a shit matchmaking system.

Get the fuck out.

File: BOOTY.jpg (59KB, 473x341px) Image search: [Google]

59KB, 473x341px

>>386128949
Dedicated servers

T H E
E N D

>>386128949
linked to ssid to discourage smurfing
All games are tracked, even casual. (secretly, so you dont get butthurt about it)
boosters flagged/banned.
cheaters placed with other cheaters.
extremely high priority to ping/latency, hardcapped at 100ms or below.
wireless players placed with other wireless players.
Habitual DCers/leavers placed with other dcer/leavers or banned.
high priority to similar skill level, increasing over time. NEVER face a player more than 2 skill brackets away from your own.
crossplay across all platforms; zero autoaim/aim assist for console players.

Blacklist and preferred player features for each player to never get stuck with the same asshole again.

>>386129228
Underage leave

>matchmaking
Why bother trying to improve a fundamentally bad system?

>>386129462

Way too much money required to maintain the server-side to be sustainable.

Especially when sales peak within the first few months and then online dies off.

>>386129462
if people on wireless connections are lagging or rubber banding, that's a problem with the game's netcode

it's fucking 2017, the wifi is shit meme needs to die already

>>386129971
wifi drops packets and adds latency.
Wire up if you want to play with real gamers, faggot.
You won't see it on your client's end, but you're probably teleporting around like a faggot on other players' screens.

>>386128949
Pretty much most fighting game matchmaking.

I never had a problem with facing far better players in shooters, everybody is such a pussy nowadays. Got fucked and now you are angry? Change the server. Boom, no external system required.

>>386130078
>fighting game matchmaking is good
>im gold
>getting matched against diamonds
>they either shit on me or arent using their actual character
>gives me +200 points for beating their joke character
>gives me -5 points for losing to their main character
Anon, FGG mm is garbage.

no matchmaking at all aside from placing cheaters with other cheaters.

>>386130057
a shitty 5 year old router, maybe

let me repeat myself, fuckwit:

it's fucking 2017, the wifi is shit meme needs to die, the packet loss is because of lazy netcode on the devs part

>>386130057
Did you ever heard of AC? Faster than gigabyte lan nerd

>>386130241
Throughput != responsiveness.
Competitive gaming requires wired connections, period.

Do you want your headshot packet dropped because you're on wireless?
probably, because you're on the receiving end, faggot

>>386130223
Kill yourself. It's gotten better since the dark ages, for sure, but it'll never be as reliable as a wired connection.
Will mommy not let you run wires to your bedroom, kid?

>>386130176
I said most and you pick the absolute worst fighting game as your example of why I'm wrong. Also if you're a gold shitter you probably couldn't beat a diamond player even if they were using their for fun character. You even gave an example of how the matchmaking is balanced as you risk losing almost no points but could potentially gain a fuck ton if you do win. Sorry this isn't as detailed as I would like but if you want I could go onto more detail

>>386130474
I have never had packet loss because I don't use the shitty modem the isp gave you, I've always had a spare router that can handle the connection without delay and interferences. Learn2networking

>>386130695
Grats dude, you're adding latency as your pc tries to grab the data from the fucking air.
You're hurting your and everyone else's gaming experience for your convenience.

You. Hate. Videogames.

>>386130559
I was a gold shitter, but they kept feeding me free abigail points.
It's not fun to play with a huge skillgap like that.
Top players have their set braindead strings they mash on, it's really not fun at all.

Other FG's might have better matchmaking, but I doubt it.
Latency makes online fightan straight garbage; instant response is essential to the genre.

>>386128949
The first ingredient is to not be a greedy corporation aiming at retaining players by forcing a 50% winrate, and not exploiting gamble psychology by eruditely distributing wins and losses to create a roller coaster of frustration and satisfaction.

>>386130883
1ms ping here, maybe you should stop using shitty routers or learn how packets are sent and received.
Maybe you should do a tracert to learn where is the biggest delay when playing a game and educate yourself at the same time.

File: 1ep4w7.png (26KB, 241x274px) Image search: [Google]

26KB, 241x274px

NOW HERES THE RETARDED THING ABOUT Matchmaking

>MM system #1: simplistic and uses only rank (RS Siege for example)
If you are gold, and you're with a party of your friends who platinum, diamond, etc,Since the majority of them are high rank and you're the only low rank, You'll see that the game has put you against other diamonds/plats

Now say you win this match (more like your team of plats won this match) and keep winning for an amount of times while your pro friends do all the work, guess what, you got carried all the way to plat without any input or effort, Practically cheating, if you get matched with a real plat he'll kick your ass.

>MM system #2: Uses lots of variables like Kd ratio, WL ratio, rank, level and all that
You'll get stuck in a limbo, literally, you'll end up waiting for the matchmaking for five hours and you'll barely find any players because Stats differ very very wildly between most players

TLDR: Matchmaking sux

>>386130883
"We're all going to help for playing fighting games online"
Losing to players better than you is literally what fighting games have always been. SFV is a terrible game in general and some of your complaints are a symptom of the game itself not the matchmaking but you should be happy to play against better players as it's a chance to learn and improve. It might sound weird but if you only think about winning as a sign of improvement then you're gonna have a bad time.

>>386128949
For a 1v1 game? Absolutely, MMR from chess is tried and tested, it's a fantastic way to measure skill and place players against similar opponents.

For a team game? Not really, go with either a server browser or a flat league system that you rank up through a set amount of wins.

>>386128949
win/loss ratio. works for literally every professional sport in the world. disconnecting from a match is automatic loss. if your w/l is that important to you, try not playing the game during a thunderstorm or tethered to your cell phone internet

>>386129010
>>386129392
>>386129121
And you fucktards will end up with battlefield 3 or CS1.6 where one MVP has 50 kills and every other player in his team has none.

>>386131150
In an ideal world we would have system 2 with a player base in the millions.

Why reinvent the fucking wheel?
Elo matchmaking works, git gud faggots

solo matchmaking in a team game is inherently flawed

1v1 matchmaking was solved long ago (any given elo-based system), and the same applies for matching premade teams

so what exactly are you asking? the best general purpose matchmaking system is probably microsoft trueskill, but no matter what you do anything which involves selecting teammates is going to suck

Number of games won/lost is the matchmaking paramater. People who have won 90 games and lost 20 will be matched with people within a +/- 5 range of that. People who've won 5 games and lost 20 games will be matched within a +/- 5 range of that. These statistics are only visible to the individual player.

>>386131150
matchmaking which attempts to use anything other than the outcome of the game is horrible.

in some games you can use the scoreline, but thats it. statistics ideas like kdr and accuracy might have a correlation with skill level, but there are also plenty of circumstances where that assumption doesnt apply.

for example, an entry fragger will naturally have less kills than his followup, and someone spamming a choke will naturally have lower accuracy that someone hiding.

weve seen this exact thing with overwatch. plays that arent easily quantifiable like mei wall placement are completely ignored whereas people pummeling shields are rewarded even if they had something better to do.

literally the only thing which represents skill is how often you win (and who against), because winning is by definition the only aim of the game.

>>386131301
w/l is useless in anything team based where you dont play with the same team all the time.

>>386128949
Make the game single player.

>>386132087
then make a better game where good players can't just carry the team if players suck, then your w/l will be accurate

File: 1470074530872.jpg (155KB, 1001x717px) Image search: [Google]

155KB, 1001x717px

>>386128949
>>386129392

>>386132040
good post

File: smug.png (728KB, 899x532px) Image search: [Google]

728KB, 899x532px

>>386131398
>Elo matchmaking works

Balance is a meme. No game will ever be sufficiently "balanced" enough for the retarded communities that multiplayer games attract.
Sometimes you're going to be matched against people who are better at the game then you. Sometimes you're going to lose. Boo hoo. If you think a particular strategy is broken, then you should either learn to counter it or start using it for yourself. Don't harass the devs into fucking around with a game that's already finished.
If you want balance, play rock paper scissors.

>>386128949
Symetric

I wrote a document once for a matchmaking system. It's untested.

https://hastebin.com/aqojapucid.vbs

>>386128949
having a balanced matchmaking system is almost impossible as there will always be flaws in it
i was preferred dedicated servers as it allowed me to play a variety of servers with players of all different skill levels and thus learn from them, plus i was able to meet some of my best friends i know through steam that i never would of been able to through matchmaking due to skill differences

>>386129462
I'm sure this will work great when fotm AAA titles like COD IW, Titafall 2 and EAfront all have less than 1k players online at this very movement.

>>386129462
Enjoy your 45 minute queue for a single match.

>>386131318
That's fine by me.

>>386136531
if you're an exploiter

start with 100 point

if you lose you get -1 point if you win you get 2 point and a happy sound
your number always stay 100 the actual score is kept in my folder only i can see it

an you just match equal numbers -/+ your choice of rank range from yours