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

Stacking in Teams and FFA

9 posts, 751 views
Post comment
Filter:    Player:  
sort
The amount of imbalance in teams and pre-teaming in FFA is, from what I've observed, very high as of late. This has been a recurring issue for some time now in Zero-k, but as of writing this post, the scale has escalated to the point it ruins the entire elo system as a whole. While there is nothing wrong with people in a party or the same clan wanting to play in the same team, stacking a bunch of veterans in one team in order to "farm" elo from random players isn't desirable either. Take http://zero-k.info/Battles/Detail/1407565? and http://zero-k.info/Battles/Detail/1407563 for instance, the influence of putting players of the same clan/party into one team completely destroyed the balance, and the games ended up as a complete stomp. Repetition of this in the long run will result in players who are unfortunate enough to run into stacked parties to become underranked, and players who randomed get shuffled into the stacked party's team gaining free elo, which will cause further issues for the balancer the longer it goes on. Despite how bad this sounds, it isn't the worst, most terrible elo act of elo manipulation ongoing in Zero-k.

The next issue in line to this is pre-teaming in ffa, which, as you may or may not know, is a mode that gives the biggest push in casual elo, on a much more scale significant even when compared to 1v1s. For a while now, some players whose name I need not mention are found pre-teaming and essentially manipulating elo in ffa, placing their casual elo wherever they wish, whether it be higher or lower on the ladder, in a short span of time. While it is true that one shouldn't really have expectation when they enter a FFA room, and that it should be chaotic, using FFA to greatly influence the elo system is plain disgusting on top of making those FFA games less enjoyable for others who partake in it. The consequence of this will really start to show once these players return to the lobster pot, where the balancer will place them according to their casual elo. While I'm not here to address how game ruining pre-teaming in FFA is, I'm certain that @hedkeaf will be able to elaborate on the details in that regard.

With that said, I'm a firm believer that as long as the Zero-k community behave reasonably aka follow the code of conduct and the troublesome players stop taking virulent actions (which will promote more toxicity, if anything), the community and the playerbase will gradually heal, grow, and eventually reach the amount where the balancer will become much more accurate thanks to a larger sample size.
+5 / -1
20 months ago
quote:
the scale has escalated to the point it ruins the entire elo system as a whole.
I would say citation needed. Might be correct, but without any objective measure is hard to assess.

quote:
Take http://zero-k.info/Battles/Detail/1407565? and http://zero-k.info/Battles/Detail/1407563 for instance
The way I see without watching the replays, the first game is "rather balanced elo wise" (56% after battle chance generally means that before battle the chance was between 50% and 56% which is not bad). If any two players communicate out of band (meaning, if they are on voice chat) will make a huge difference which the balancer can not take into account. Of course preventing "stacking" could reduce this by an average of "50%" (average chance that 2 players will end on the same team), and I think it would be a good idea, just mentioning that I do not think will be as final solution as it might seem.

The second game probably would have been a stomp in any combination (@123vtemp is position ~250 on ladder while all others are lower than ~80), so not sure what you propose.

Regarding FFA always thought should be a separate ladder/rating... (more for fun, added bonus people can't not do things described)
+0 / -0
I think it would make sense to keep "stacking in teams" and "ffa elo" in different threads.

From previous threads on "ffa elo" it seems clear that ffa should not be rated.
Later it could get its own ladder, but for now not rating it is probably the more likely change.
+0 / -0


20 months ago
In both of the listed games, neither team contains a clan.
+0 / -0
20 months ago
I apologize for the errors on my part (mistaken a faction for a clan)
+4 / -0
20 months ago
The Balancer in teams often seems to overvalue the purple players within the team, believing they are equal to say 3-5 ~giants this in turn results in a 1v5 essentially due to the purples teammates being >giants, the fact a purple is the equivalent to multiple golds and plats is absurd and it just results in a lost game from the start, with this the more players in the pot the better it is but often times a variety of ranks culminates in a completely unbalanced team, the balancer would definitely be better if it balanced teams 1 by 1 by rank ie equal purples equal supgiants equal giants and so on this would result in a more balanced experience for everyone and not just a purple tryharding to carry an entire team and then losing elo through no fault of their own.

In terms of FFA there really is not many solutions to pre-teaming other then enforcing rules which i believe the moderation have already decided against, and even if that did happen, those who are friends and want to actively cheat and team will continue to do so, best solution is to cut out those players from your game and just try and enjoy it yourself with others, in addition to this the more you play FFA the more you notice those same people troll in anyways possible either by pre-teaming, instant resigning or rage resigning when something doesn't go their way.
In addition to this their is a large proportion of high ranked players who prefer to simply delay or completely "kill" lobbies through trolling, they do this with !maxelo commands in order to lock lobbies, join to kick and then go back to spectator or spam random votes so people get frustrated and leave this seems to be a recurring theme in early mornings or lower population games and is extremely frustrating for those trying to play.
The best solution I can suggest for those playing FFA is to
1. Host the lobby yourself and !title the lobby no talking or info sharing this can stop the teaming aspect from random players and means players have to rely on their own skill and scouting in order to team which is the heart of FFA
2. Don't play the games with those who often team(without naming names) you can discover this very quickly, this solution is however a very jagged one due to the fact FFA player pool is relatively small and has repeat customers so you are forced to play with those individuals
3.Play small teams
+3 / -3
I appreciate the desire to not publicly name names re. rating manipulation, but it is difficult for the admins to make much headway unless reports are submitted.

quote:
The Balancer in teams often seems to overvalue the purple players within the team, believing they are equal to say 3-5 ~giants this in turn results in a 1v5 essentially due to the purples teammates being >giants, the fact a purple is the equivalent to multiple golds and plats is absurd and it just results in a lost game from the start

Purple rated teams players, by definition, get and keep that colour by carrying games in which their teammates are typically weaker than the players on the other team. In some combinations of player distribution and map, this occasionally leads to pretty unwinnable games - but you might be surprised what some of the purple players are capable of carrying.
+3 / -0
There are a number of blue ranks that have highly stable performance, with a set of well executed and robust game plans that result in high army efficiency. The strength comes from, good plan, good evaluation of tactical situation, unit micro and sense for right transition timing, even though the plan itself is pretty obvious after seeing it a few times.

I don't think there is a consistent game plan that gets you to purple and it is more hit or miss. The kind of things purples do beyond blues, from what I've seen are:
1. Opening gambits that quickly collapses weaker fronts, resulting in fast advantage
2. Unreliable Gambles that saves a game from losing situation: the best athena backstabs I've seen came from purples. There are also things like sniping paladins, shield or lance balls with bombs across a thick screen. Some purples can also sniff out this and prevent it, for example thinking about weird silo locations that generally is not looked at.
3. Play effectively with large and weird comp. armies from leavers and sometimes grind out a win afterwards.
4. Effective rear guard action against heavily stacked front on maps with isolated lanes (most stable game plans require not absurd econ balance)
5. Effective play in less common map types: sea maps, hill maps, very wide front maps, larva map adaptations
6. Weird compositions that sometimes is effective against whatever it is used against, but doesn't work if someone just wants to copy it. (probably situational with most players not having enough game experience to know counter gameplay/transitions works)
7. More messages to other players on what to do. This is kinda annoying but purple automatically gives weight to suggestions. This can be decisive in things like strider rushes or superweapon/nuke/econ strategy beyond contribution from normal individual gameplay.
8. Major army position changes: entire lance army shift across maps between inaccessible terrain with charon? That kind of thing.

That said, I do suspect many casual purples come from 1v1 and FFA. Some can be identified with 1v1 style composition/gameplan as opposed to pot-type game plans, not very common sight in pots and causing their pot team to lose more often than not.

If a purple gets into an tactical fight with high ranked blue they probably not really carry, but when some lower ranks stall out blues with defensive play (for example dirtbags) maybe a way of winning can be found.
+1 / -0
We are straying here from the intended purpose of the thread, but I'm not sure there is much to say beyond:
  • (Continue to) report clear cases of deliberate rating manipulation. Don't bother making a thread about each one, just send us reports. Making some kind of a post on the replay so it stays on the server is fine, just make sure you send a report.
  • The admins are aware there is a problem with FFA and we are considering our response. We have the capability to make our response retroactive if needed.
  • Possibly there is also a problem with 1v1 in casual rating. There is certainly a problem with the no-elo 1v1 room losing its no-elo flag with some regularity. These matches should get unflagged for casual rating, but understand that this may take a while to propagate to people's actual ratings even when we get around to it.
  • It's not obvious what the solution to "friends choosing to group together with parties/clans, and having better results when they do" is. It's not even obvious that there is a problem to be solved.

quote:
If a purple gets into an tactical fight with high ranked blue they probably not really carry

Surely this depends on the purple a bit. There isn't that much difference in rating (sometimes none) between a low rated purple and a high rated blue, but the highest-rated purples have every chance of running circles around any blue (and for that matter, a lot of the purples), especially in smaller-size teams.

Putting aside the players in a league of their own, I would characterise the difference between low-to-mid purples and a lot of strong blues as "versatility". It is typically very difficult to put a purple player entirely out of their comfort zone, or to predict what they will do in a way that enables you to strongly counter it. Some strong blues are quite versatile, and often if their execution improves a bit or they have a run of good luck they become purple pretty quickly.
+0 / -0