Jump to content

Sunday Pro Games League Fixture


HMS-Surprise
 Share

Ratings for 4v4 League  

16 members have voted

  1. 1. What are the most important things to calculate ratings in 4v4 match?

    • Total Score
    • Military Score
    • Economy Score
    • KD Ratio
    • Total Kills
    • Other: Share your ideas by replying this topic.


Recommended Posts

6 hours ago, stanislas69 said:

@HMS-Surprise Would it possible for you guys to run the svn version of the game ? This would provide us with accurate feedback and recent commands.txt @wraitii is working on the pathfinder, and will be needing real life samples to make sure nothing get unintentionally broken

Sure. We can do everything to make the game better. Let's discuss the details over an instant messaging app that you prefer.

Link to comment
Share on other sites

My idea was to weight not by a constant but rather a variable reflecting the value of the resource at the exact time of each event. This factor can be roughly based on the game time, but it could better be formed of the amount of the given resource available to the team at that moment, perhaps even considering the possibility of bartering.

If we want to include tributes, we should probably also subtract the equivalent score at the receiving side to prevent easy score boosting between two players.

I'd tune the constant factor in a way that the resources given to a player give the sending party a similar score to that that a player gets when killing a unit that could be made for those resources. It's hard to say the exact value because the other parts of the score should also be weighted variably.

Link to comment
Share on other sites

2 hours ago, itrelles said:

cc down, barracks down, castle down, and making someone surrender should be the most important for me. speed to reach phases maybe 2

scores dont mean nothing to me, scores are full of lies

These points came from your team game. You did a good job maybe, but individual success is not important here like 1v1 matches. If your team wins, everybody in the winning team gets 2 points. If you are in the losing team you get 1 point. If you see your skill score it's 0.56. You can't even compare with winning or losing. I think individual skills matters but not enough to win a team game because all the players are really good actually. 

 

Sometimes rushing can help your team to win the game, sometimes it's a defence tactic. You can play really good in a match but that doesn't mean you win and your teammates lose. You win all, or you lose all. Being in the winning team is the most important thing here. Your individual skills affect less. 

  • Like 1
Link to comment
Share on other sites

Link to comment
Share on other sites

On 5/6/2019 at 11:32 AM, Boudica said:

We could probably extract the time of a resource share event from the replay file easily, but it could be more accurate to weight by the average economy strength or value of each resource at the given time. To get the whole context, it's probably better to just run the replay with some event hooks that update the score.

The metadata.json contains a record of all scores (and kills etc.)  for about every 30 seconds of the game. So no need to rerun the replay. ;)

  • Like 1
Link to comment
Share on other sites

Link to comment
Share on other sites

I already talked about the resign time resource exchange problem above, @ffffffff. There shouldn't even be a need to address that case specifically with good weighting. As a simplified example, imagine just counting the area below the resources-sent curve in the graph. Increasing the total near the end wouldn't affect the score because the game ends at that moment. I think that when it's a team game, the score should stop counting when the first player of the losing team resigns. That could help eliminate score-hunting style of playing.

I've been working on that idea a bit more, but maybe we should do it properly and improve the system that is present in the game instead. There are some parts I don't know how to address yet. When someone is rushed, they take damage for their team, but it's currently not reflected. Such player might seem to play worse but it could be the exact opposite (they get rushed because they are good).

I still think that the system used in this thread is pretty decent. It could eventually be replaced with something that is much more complex (rather than fine-tuning), but that new system is going to take more time to make. For now, I'd just suggest to sort by the average rating (as someone already said), because it makes me appear better. ;)

Link to comment
Share on other sites

  • 2 weeks later...

Hi all,

4v4 Sunday Pro Games event is finished last week. Thank you all to be a part of it. Special thanks to @psypherium, @Feldfeld, @Unknown_Player. Stay tuned for upcoming events. Here is the final league fixture:

0AD_League_Fixture_-_Players.thumb.jpg.396533186ecb0dc760d3f367d05d8e17.jpg

@psypherium @Feldfeld @ValihrAnt @camel @Boudica @ffffffff @JC (naval supremacist) @Unknown_Player @borg- @chrstgtr @Lefo @Pudim @Philip the Swaggerless @Stockfish @PhyZik @itrelles @nani @Hannibal_Barca @LeGenDz @phoenixdesk @LANDLORD @kizitom

*** If you want to be in the notification list, or not, send a PM so I can add/remove your name.

 

Edited by HMS-Surprise
Link to comment
Share on other sites

  • 3 years later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...