[Request] Simple ingame ranking system

[JohannesH]

Scoreboard would not be shown in the lobby. It would only be shown on the website.

What is the player at the top has more losses than wins? Would you still refuse to play him? Please, none of that is a concern.

Its not about me dude... I'll play vs anyone and be confident in winning IF I try my best, but that's beside the point.

I just don't get whats the point behind showing such stat to public, for the already stated reason.

For a scoreboard to be relevant it should be only recording select games where both players agree to try their best to act as a measure of skill, and/or be used for some balancing or matchmaking system.

[Forboding Angel]

So you take issue with your cs or l4d stats being shown to other steam users?

Competitive play drives competitive play. Simple fact is, compared to commercial games out there, your claims are pretty much bogus.

If you buy sc2 are you going to only fight noobs? Besides if you care THAT much, just fight an AI repeatedly with a 100% bonus.

It's just a simple scoreboard, not an all in one stats reporting thing. If you don't want to know, you could always just not look at the page...

[JohannesH]

And do those steam stats encourage competitive play? I dont know how they work exactly but I would guess they're just clutter nobody with a brain really cares about.

In SC2 the point of the ladder is matchmaking, to find a random oponent of your skill. If I do buy it I'll surely play games outside the ladder too. That's the point of springie elo rank too for example, its not some e-peen only list derived from random games but tries to give well balanced pub games, it has a purpose

Well most likely nobodys gonna take your scoreboard seriously/competitively if it records all games. But if people do care about that ranking, well that's probably bad for everyone involved

If you want to encourage competitiveness, have a proper ladder and tourneys

[malric]

Uh oh, just change it on website? Its open source, you can modify it to count elo for each game independently. (Even retroactively). It just has list of games in database..

Where is this source ? I looked on the wiki, git, springlobby and can't find it . I'm just curious to have a look...

[Licho]

http://code.google.com/p/summerspring/s ... ngie%20web

[Hobo Joe]

And do those steam stats encourage competitive play? I dont know how they work exactly but I would guess they're just clutter nobody with a brain really cares about.

Yeah I can say first-hand how true this is. Anyone who cares about stats on Steam is a brainless pubstar who totally sucks. I've played comp on Steam games with stats and anyone who actually plays the game has horrible stats. Your accuracy is shit from blind wall shots, your win/loss is shit from dicking around or because you actually play people of equal or greater skill, etc. Only people who care about stats are the people who can go into a game and be like "I HAVE 100% ACHIEVEMENTS AND A 4.0 W/L RATIO I'M BETTER THAN YOU".


Basically stats are pointless and the only time where they are at all relevant is a w/l ratio in a game that is chosen to be ranked such as a ladder game.


Besides that, in games with small communities (such as spring) anyone who wants to know who the good players are only has to ask or spec a few games that aren't 8v8 dsd. The idea to use something like this for balancing games is a nice thought but would never work in practice. Most people who play Spring stick to huge games, which means their contributions are padded by 6 or 7 other people, which makes their wins totally pointless for balancing purposes. Not to mention smurfing.

[Forboding Angel]

Perhaps none of this is aimed at the spring crowd...

You guys think too small

[Pako]

I have made a shitty enough ranking for this purpose.

Code: Select all

function widget:TeamDied(teamID)

local stat = players[teams[teamID]].stats["Enemy kill ratio"]
stat[1] = stat[1] + teamsDied
stat[2] = stat[2] + teamSize

teamsDied = teamsDied + 1

end

I already have stats from several hundred games but I forgot to save player names(only Hash) so I can currently only see stats from players ingame but there seems to be some correlation to skill. Average ratio is 65%, over 100% is quite rare.

[Licho]

Winrate ratio would be useless for player of non-BA mods who play on springie. It balances by skill (ELO) and over time each player's winrate converges to 50%

No matter how good you are, it will just give you sucky nub allies to win 50% of time :)

[Pako]

Winrate ratio would be useless for player of non-BA mods who play on springie. It balances by skill (ELO) and over time each player's winrate converges to 50%

No matter how good you are, it will just give you sucky nub allies to win 50% of time :)

My algo doesn't suffer from this. It just counts how many teams die before the player. (at endgame alive teams are credited)

[Licho]

Does it work in real games? I mean normally nobody leaves until whole team dies. And Im often the one who quit first when im losing.