Hidden global rank for balancing?

One problem with Empires currently is teams, which suck ass many times. Clans stacking, players joining favorite faction and sometimes just pure luck(or unluck, erm, you know what i mean) makes for pretty bad teams. Which off course leads to one team steamrolling the other in a couple of minutes and the game being over after 10 minutes. Then the game switches map and we begin all over again.

My suggestion: A hidden global rank counting K/D and score. This rank cant be seen so there is no real competition, no "OMG WE ARE GONNA LOSE MY RANK IS GONNA GO DOWN THREE POINTS, I QUIT!" or similar.

It doesn't HAVE to be hidden but whats the purpose of showing it? (except bragging rights?)

 

So if it's hidden, what does it help?

 

So..... are you saying you want the autobalance to use this rank thing to make even teams ?. I guess something like that could work... but if the server have an average of 2,something KD ratio and some dude is like 8 what team will get him?.

Also SKILLZ are not the only thing. Lately there have been quite a few games where one team is obliterated because they start with 50%+ scouts on their team. Also theres the few servers that makes you get 50-200 res per sec just having a few refs and have ultra high speed research. If thats where you play then super fast games are caused by this as heavies will be out in no time hammering away.

Also clan members are not THAT hardcore a lot of very good people are not in a clan and just have random names, but still do as much damage as any clan member. So I can rarely see it when people declare skill stack unless I see a very high amount of clan people. I have seen people declare skill stack just because of 2 people even when they had 1 of the same clan on their team. I guess a rank thing could do some good.

But there is far more to skill than a KD ratio. In some games you kill maybe 25 people and die 30+ times..... not that good right?. but if 15 of those people where in tanks then thats something that migth have saved the entire team from being raped by several rushes. I also see people scare tanks from rushing by releasing a storm of rpgs/ people sacrificing themselves to slow the enemy team down dumping mines/ good turret placement/wall/ammo box/camera and so on. Stuff that is not measured by KD ratio that have very important role and is often the stuff that allows the people with super high KD ratios to function well.

Measuring points wont show it properly either as there is a big difference between a point gotten by nade spamming a turret farm of killing a heavy tank. It might even it out a bit to ensure that one team doesnt get all the best riflemen, but we need something very advanced to get good team skill balancing thing.

 

K:D does not incorporate teamplay or time played, so it's easily flawed and above all it's restrictive.

 

Note Score, but its a bit hidden so i dont blame you.

You mean the scouts that sit and snipe in a corner and are at the bottom of the list with engis and riflemen at the top? Frequent scout players (who suck complete ass) will have less score then their more usefull counterparts.

Well more points for killing heavyer vechiles? I dunno, all the details arent ironed out but you get the general idea right? If we were to implent a hidden autobalance system and use something as a guideline to balance teams would it work?

Most Autobalance system ARE based on score and K/D. If not that, what?

 

And most autobalanced sucks.. Like the one you have in the beginning atm, I press BE, but end up with NF , vice versa.

 

Most autobalance systems are based on KD, but this is not most games. KD scores don't take anything into account in empires, I can't tell the amount of time's i've been invaluable as an engy while barely touching an enemy - think 90 points with a 3/20 KD. Snipers can have a brilliant KD, but they are a useless class, and that was Rex's point about teams losing due to scouts - not that they have bad scores, just they can't build jack and they can't destroy tanks, so more than 1 for every 8 or so people is just utterly stupid.

I would say the only class where KD may actually be taken as evidence of skill is probably grenadier - due to getting the kill when destroying tanks (the driver gives the kill)...but yet still this doesn't take into account the buildings destroyed, so even this is inaccurate.

 

so what's this suggestion for. what's the ranking for?

 

Who's going to make this system?

Also, what about a comparitive point calculation, so your points as a percentage of the players on your team in a given time frame. so if you join at 10 minutes and the game goes for an hour, your points are measured as a percentage of the total points scored by your team in the 50 minutes that you were playing.

Sadly each server would have to have an additional server for counting the [(team points / individual points) / time], but hey, money's no object.

 

In order for this systme to work globally, you'd need a master server, and masters servers need large amounts of bandwith, space, and availability, and if the server goes down, you have the problem we have now, bad autobalancing.

 

Problem is that motivation and moral of a team is just as important as skills.

 

How about we have everyone take an IQ test and autobalance by that. A team with retarded players will get some intelligent people and vice versa

Also see Solokiller's post

Also note how impossible it is to do decent autobalancing for Empires

 

The BSID server back in 1.07 used to have a rank system... a local one. This sounds like it would need to be quite complex. Maybe something simpler like hours played on the individual server ?........ or something.

But its important to note that a lot of these fast games where one team gets squished are because of servers with changed res/research. You can defend against paper on normal servers for quite a while, but modified servers result in armor/guns/engine very fast so they dont need to care much about turrets/grenadiers if they rush forward.

It would have been pretty cool to know how many vehicles/buildings/turrets/wall/people you destroyed in each version of the game. It would be nice stuff for clans picking out new members OH a TANK KILLER!. But then people would get obsessed with it.

 

That was just a simple thing eth0 installed but it would occassionally flip out and request 300% of the CPU power available so it'd crash the server. It wasn't used for balancing or anything, just a novelty.

 

Yeah, a master server with an efficient way of keeping playerinfo would be good.

 

QFE

Also CW has a stats plugin running on their server.

 

Perhaps it could be based on complex algorithms, once written and not looked at for a week, then was considered balanced since nobody could interpret it.

(bad) Joking aside, I do like the idea of local ranks. Maybe a point/death ratio, or some type of point/time ratio (with average time which stops after a certain bit of going without earning points), or perhaps less actual points for solo kills, and more for working as a team (squad system+radius, anyone?).

However, suppose if all servers had local ranks, and then they merged them all together to make one new rank list, used that, and then later merged every so often? It wouldn't require a master server, it would just require a mirror to host each list and then a program to merge the date.

But, the problem is still, who is going to code it? I think a human would die having to hand-merge the lists (since stat-tracking alone, I would think, wouldn't be that big a deal, as long as extra features like ranks weren't added).

I don't know. My post sounds kind of idiotic from my viewpoint, but I am deciding to post it anyway.

 

Well Krenzo is a coding god but I'm pretty sure he prefers to spend his time on enhancing the mod instead of making stattracking software :p (it's on his mind tho)

 

I agree ,but if there is a noob team ,and they lost few rounds (stack ,skillstack,clanstack ,w/e) and they motivation is low , they will start disconnecting. But when they see that few "pros" changed team ,and now they are with them , they will play better ,because they have a real chance of winning.

I think this is a good idea , and would make pub games more interesing.

 

I think you cant really know how good all players are but after playing some games you know who is responsible for the win.

I have invented, a not so serious 4 point system, for my personal use

1 Point = Scout that camps in base
2 Points = Normal player with no ambition to really win
3 Points = The Stackers, or the people that count to win the game
4 Points = One man army

It works for me