Loading...
  OR  Zero-K Name:    Password:   

Casual ELO & rank

50 posts, 1386 views
Post comment
Filter:    Player:  
Page of 3 (49 records)
sort
7 years ago
The matchmaker experiment does not seem to me very successful (most of the times I login I check, weblobby reports '0 in queue' and the ladder is of only 42 players which means 42 people used it ever).

In my opinion it would be great if ELO and rank (order in the ladder) would be displayed for each player in its home page. For both matchmaker and casual. At a later stage the casual could be hidden but not for now.

I think it would be good to know if any dev opposes strongly the above and how many players would like the above.
+2 / -0

7 years ago
Actually I think people fell off the competitive ladder due to inactivity, so more than 42 people have used it lifetime. It is a feature for the future.

As for the actual request I don't really have an opinion.
+1 / -0

7 years ago
For now you can still see your casual rank on the ladders, but looking forward to steam release I don't see why we should show casual ranks at all. The matchmaker makes a nice distinction, splitting the games where I mess around with coms, rushes and funny game modes from the ones where I'm actually trying to win. Why would you create a ladder for the former?

In my elo analysis I've seen that the current big team game outcomes are mostly random, which is why we would first have to establish a competitive environment before there's any point on ranking players in this game mode.

A possibility would be clanwars, but those already exist, are better for ranking clans than players and are way too rare to be considered as competitive teams right now.

So it boils down to the old conclusion again that, without a better rating system, we are simply unable to give out reasonable team ranks.

As for small teams, most of the active players have already switched to using the matchmaker for those games. I wouldn't even want those 1100+1800 vs 1600+1600 that are played outside of MM while waiting for more players to count.

For the players that don't play matchmaking and are thus depicted with a wrong rank in the forums/game it might be a good idea to display them as unranked or with their casual rank ifneedbe. We could use the grey rank icons for that and shift the other icons down.
+2 / -1
7 years ago
I can see the casual ELO only if I am among the 50 in the ladder.

When will this steam release be and why is so hard to "hide" the casual like 1 week before? You do something now for something that might happen in 1 year from now.

If it does not make sense, why is there casual ladder displayed at all? Or you do not want to upset the top players (35 out of 50 never used matchmaker)? So, is this like "you are not in the top 50 we do not care about you"?

Are team games still balanced based on casual ELO or just "random"? If there are still balanced based on casual ELO don't you think this is a "double standards"? System is good enough to use for balance but not to tell people the info.
+1 / -0
7 years ago
quote:
In my elo analysis I've seen that the current big team game outcomes are mostly random
From what do you tell that? Or more precisely, what do you mean by that? Isn't it good if games are balanced so that win chances are random? Or do you mean that our score of probability estimation ("trans log score") is too bad?
+1 / -0

7 years ago
What in my opinion hurts the system is that ranks right now do not accurately represent the skill and are very misleading. In my opinion as a substitution the rank displayed should be of max(casual,matchmaker) elo. With that said I agree that there's no point to hide casual elo if MM is not used.

(note: I'm a good example with around 500 elo difference between the queues, but I've shortened the gap recently to about 400)
+1 / -0

7 years ago
Yeah, the biggest problem with the MM/casual split elo is that now elo is often inconsistent between them. Also showing colored rank icons but not elo in deluxe player list is meh.
+2 / -0

7 years ago
seeing the elo of your teammates can impact your teams overall performance,because u know which player will need more help,therefore preventing one flank or side from beeing compleatly lost ,this is especially relevent on big maps where u cant monitor everything everywhere, cause u probably are also bogged down into micro intesive tasks on your front.

Early placement of players can give u a pretty good idea on what side is gonna be the weak spot.This gives u preemptive awarness of how u should manage these week spots.

Also the MM elo has no impact on the casual elo,which ends up in uneven distribuition of skill(in some cases) like zenfur stated.

At any rate I suggest these 2 changes:

1.You should be able to see everyones elo like b4.
2.For team games ,casual and MM elo should be combined and diveded by 2 to create a "real elo" for better distribuition and gauging of skill.

+2 / -0

7 years ago
"Also the MM elo has no impact on the casual elo,which ends up in uneven distribuition of skill"

IMO this is wrong. Since casual elo is impacted, it plays the role of MM elo for casual games. It's just that you don't see it.
+0 / -0

7 years ago
quote:
DErankBrackman
Or do you mean that our score of probability estimation ("trans log score") is too bad?
Yes. But I know this will be ignored again which is why I'm not even gonna bother with the discussion.
+0 / -0
7 years ago
CHrankAdminDeinFreund: you do not want to discuss you reduce everything to "but I proven that it is not worth it". I asked 3 (independent to the analysis you made) questions:

1) Are casual team games balanced based on casual ELO? (if yes this means it is worth "something")

2) Why ladder is displayed but players above rank 50 do not get any information (ideally rank and elo)? Current system is frustrating/discriminatory.

3) Why do a change (hiding elos) for a steam release which is (as far as I can tell), quite far in the future?

I can agree that the system is "imperfect/bad/randomish" (as you say you demonstrated) - lets not discuss that here.

Other people seem to agree there is no point in hiding the casual ELO (zenfur, aeonics, sigero, manero).
+0 / -0

7 years ago
Yes, there's no point in hiding the casual Elo, because we should just remove it.
+0 / -1
7 years ago
Fine with me to remove it completely. But if it is not removed completely now, would be more consistent to be shown on home page.

CZrankAdminLicho (asking you as I have seen some commits in which you touched mm/casual ELO): don't you think it is "nicer" to show casual ELO (on home page for player only) if top 50 casual is still kept for now ?

+0 / -0

7 years ago
By "remove" do you mean "don't show casual ELO or rank anywhere" or "remove casual elo infrastructure and balance teams at random"?

The second one scares me a lot.
+2 / -0

7 years ago
quote:
CHrankAdminDeinFreund
I would suggest getting rid of casual Elo and replacing it by readonly MM elo. Averaged among all players and games, it's more accurate to balance big games by small teams elo, even if there are individuals who never play a single small teams game.

Additionally, this will save a lot of confusion that's inherently caused by having a "hidden rating".

While for many (less active) players, the current casual Elo may be more accurate than their current competitive one, this is only because of the "wrong" initialization with 1v1/Teams values and will change over time. We could probably derive better initial values from the existing two ratings.
+0 / -0
AUrankAdminAquanim: by "remove" I mean remove the discrimination between top 50 and non-top 50 when it comes to knowing their casual ELO.

I can accept both a game that shows an "incorrect" ranking (hey, it's a game don't have high expectations :p) and a game that does not show rankings for the mode I play (same logic). The current situation is annoying me without (in my opinion) having any added value for no use-case.
+0 / -0

7 years ago
Why not just have one Ladder?
Remove Casual or MM elo and make the new and amazing JUST ELO.
And the MM should just give u more elo than normal hosted games,so thres an incentive to play MM.
+0 / -0

7 years ago
ELO is a measurement of skill, not an incentive to play.
+1 / -0

7 years ago
Yeah but MM is barely used so if u make only one Elo ladder MM should give u more elo than hosted team games,to populate the MM
+0 / -0

7 years ago
afaik big team games give less elo per game anyway, although I don't really know how elo change is calculated in teams.
+0 / -0
Page of 3 (49 records)