Realistically, K:D is the most prevalent factor an individual can have in determining the win of a team and should be largest factor in determining a skill reference. [...]
Which is totally wrong, I commed games with players going 30:2 and never doing anything useful, never defending or killing a RT, never helping pushing something but whoring kills the whole game (like blinking in the marine base, killing the just spawned marine and blinking out), I rather have someone with 2:30 who killed 10 RTs and 3 phase gates on his own and frequently forces beacons, since he is way more useful then the first player.
That's laughable. Top fragging players help stop pressure which gives your other team members the ability to kill RTs and stuff. And of course everyone knows it because everyone tries to join the team with the guy who went 30-2
Normally this is true and said scenario only happened about twice in over 300h game play (I don't comm every time and as normal player don't really watch the performance of others, so maybe more often), I just wanted to give an example why saying high kd=good (team) player doesn't have to be true and therefore using kd as the main indicator is a bad idea. Btw I wouldn't call spawn killing a marine who just respawns (while there are no other aliens around killing stuff) stop pressure.
@Jekt I am talking about pubs where you normally can just walk to the RT and start biting it and at least do a lot of damage before a marine even shows up (and even if he kills/chases away the skulk won't bother welding the RT) and a lot of people tend to ignore minimap and commander when they try to get inside an alien base, even if its well defend and in the meantime marines lose half of the map.
There are a ton of posts of this exact same thing. I love the idea of balanced matches, but there are a few things to keep in mind when thinking about balance.
1. A great commander with a team that listens will win almost any game, how do you balance that?
2.Servers are community run, how do you enforce them to rank using kdr/WLR? What about different game modes?
3. When I was new to the game I was thankful for being put up against more skilled players. I got my face dragged through the mud by camo skulks and deadeye marines. But that's what got me hooked on the game!
If you have new players trying to learn together when they go against veterans they'll get smoked because no one will know the advanced tactics shared by the community. It will just lengthen the learning curve. To properly balance matches we should look to do it by balancing the game now or find a reliable way to balance the ready room.
Server Browser – Skill Based Filtering Proof of a working system?
Solid proof: a screenshot of a QL server list.
The rating system in QL doesn't work in my experience. I've been crushed 30-0 by good players, only to join their server later to spec and it says 'Skill Matched'. Me and some friends also regularly take games off of each other yet sometimes I won't be able to join because 'Your skill is too high' and other times it will be the reverse.
Player count is too low to have QL style ratings system used. There are also servers that already run KD based random teams, though they only do it based off of the previous round.
ZZZ its not intended to be a matchmaking style system - its just intended to give an average idea of the skill level of the players in servers. Ideally rookies would be 'encouraged' to join servers with lower skill ratings but beyond that i wouldnt imagine there being any restrictions - people can join any server they please. Its all about giving new/newer players a decent shot at finding a reasonably skill appropriate server.
Honestly as a player that can join most servers and completely break the game, its not fun. The few players that do enjoy 'stacking' or w.e are generally players that just are not that good, and as such look to others they know can carry them. However this system really has absolutely nothing to do with that aspect - its more about establishing a general skill level per server, nothing more IMO. Let team stacking and random based on ELO/whatever be left up to mods and/or admins.
Solid proof: a screenshot of a QL server list.
The rating system in QL doesn't work in my experience. I've been crushed 30-0 by good players, only to join their server later to spec and it says 'Skill Matched'.
See the part where I wrote...
It is not a perfect system, but it does give a general indication of rookie, average players and anything above.
FYI, QL 1v1 is a game where two even players can actually go 30-0 against each other. Map knowledge alone can attribute to this.
ZZZ its not intended to be a matchmaking style system - its just intended to give an average idea of the skill level of the players in servers.
However this system really has absolutely nothing to do with that aspect - its more about establishing a general skill level per server, nothing more IMO. Let team stacking and random based on ELO/whatever be left up to mods and/or admins.
L4D2 match making was a lobby system where a group of 4 people had to wait to be matched up against another group of players. What does that have to do with this exactly?
It's a system which was supposed to mitigate "stacking" (i.e. 4 really good people against 4 randoms).
It didn't. The novelty lasted 1 week and then it died. I haven't played L4D2 for a long time now, so I'm not sure if it's still dead, but I am willing to bet it is. When people start losing (usually after 1 or 2 "segments"), they drop out and move on to a different game (server). They're rapidly replaced by noobs which perpetuate the loss or lose even worse.
It's not the same as the proposed NS2 changes, but some aspects of it are similar, and something to think about.
Basically, the questions that come to mind are: how are you going to make people use this "system" and how are you going to make them stay for entire matches. Just because it's NS2 doesn't necessarily mean it won't suffer from the same problems as other "ranked" games to some degree.
Yea, all this does is provide additional information for people to make better server choices before entering the dreaded precache, which is a good thing! Kinda like what the server performance column was supposed to do.
My only thoughts are about how you can make this information more useful.
1) The vast majority of players are going to be low rating. If you use average, the impact of good players is vastly understated.
2) One good player has more of an impact on the pub game than 2 good players (with one on each team), yet the average rating is lower in the former.
So, i think we should really only be interested in measuring the level of higher skilled players on the server. My suggestion would be to group certain rating ranges as beginner, intermediate, expert. Then in calculating server skill, you discard all beginner ratings or apply some weighting factor to discount them.
Like (avg(b)*n(b)*x(b) + so on for intermediate and expert) / 3, where the discounting factors x(e) > x(i) > x(b). Maybe you only need to discount beginner ratings, and not intermediate + expert. I dno.
And/or you could make it so when you hover over the 'chevron' rating, it'll give proportion of beginner/intermediate/expert as additional information e.g. 50%,40%,10%
I dno how you would solve the 2nd point though, if it's even an issue. I guess that is more an in-server matchmaking concern.
Why'd this thread die? This is awesome. Got it all covered. This is exactly the kind of easy implement balancing we need. Add in a "last round kdr" based random ingame, and we got all our balance tools. Sure, they're not perfect, but they're better than nothing - by 'nothing', I mean, pondering about a super complicated totally "fair" system and then totally never getting it done.
Just have the rankings based on time or no rankings at all, but you can just left click on the server and there´s an option to see the server list with all the playtimes of players.
KDR is not very reliable.It´s good for pinpointing good players on that server in that specific game.Over time you might see, that they do well consistently, but it only starts being consistent if you are really really good.
Rankings based on score might work, but then promoting to the next rank should be made exponentially growing.I´ve seen this system work in America´s Army.Different game of course, but I do remember being able to tell if a person was good by just looking at his Honor(ranking).
Never saw this thread. Between this and the newbie servers thread are two great ideas. This, however, would be a benefit to all, not just those newbies with very low game hours. All for it!
dePARAJoin Date: 2011-04-29Member: 96321Members, Squad Five Blue
This browser wont protect rookies from pub stomping btw.
Dont get me wrong: its a great idea and i like it.
But dont expect too much from it.
For people who want to play with equal skill players its awesome.
But there always people out there who feels like godlike if there on 1st place with 50-2 on a rookie server.
This browser wont protect rookies from pub stomping btw.
Dont get me wrong: its a great idea and i like it.
But dont expect too much from it.
For people who want to play with equal skill players its awesome.
But there always people out there who feels like godlike if there on 1st place with 50-2 on a rookie server.
I think we can all accept that that's the case (like it or not). What this aims to do is to give the rest of us, the majority (?), the chance to actually avoid that situation to begin with.
Accidentally joining a server where everyone is FAR worse than you is nigh on impossible to avoid at present. Anything that empowers players to find servers of suitable skill level has to be a good thing, then it's just down to promoting that mod as much as humanly possible so that the decent players actually get to hear about it, install it, and use it. (Better still, if it works out well, UWE could implement it in the browser by default, but let's not get ahead of ourselves).
As pointed out by some; it is tempting but ultimately wrong. Each player will try to stack one way or another. While you're at it, why not giving appreciation tags like "no TeamWork", "curse, filthy mouth", "suspicious, hax". I'm afraid it would bring more problems than it would solve. On the other hand the score tab definitely need some improvement in itself.
That said. What about taking the idea upside down ?
If players are given an way to set a profile (announcing); it would be better for making a "controlled randomization", or Distribution. Of course players accept random, or admins enforce it on servers.
A profile would be submitted by a form (like an in-game option panel) that gives indications about the player way of playing: "Do you play Fade Y/N", "Commander : Yes/no i'm not ready", "Lieutenant : Are you a pistolero or a mecanic ?". Things like that. When randomizing the RR, the server have a clue about who is gonna playing what (or able play this or that).
-You could have the two commanders of the server not ending on the same side. 2 good commanders will definitely be really efficient because one knows the job of the other. So the one on the field "senses" what to do without being told, that is an F***ing asset.
-Avoiding total rookie getting in the chair (have to pass training map to unlock the option).
-Allow a better organization in team as each will probably do the part he announced he would do.
Then the time played as a role/unit can also be a parameter.
It can provide a way of doing things NorthBound or SouthBound (if i can say that). A 300hrs fade is more likely to be more efficient than a 2hrs fade. Server admins could set the server to choose for the alien the most skilled (optimizing) or on the contrary the less skilled (giving a chance to rookies). As a 300hrs fade can be an ass at welding ...
And if one wouldn't want to provide such information (i don't wanna command these days); you can uncheck it or even leaving the form blank.
How to?:
*I believe the form in the options menu is an easy job on UWE side.
*Server random algorithm would have to be improved a little but that's not difficult. Far from it.
*Players accept randomization (instead of joining A or
*Servers is set to randomize automatically
Ok there would be repetition sometimes. But that could help avoid the stacking syndrome on pubs.
So many cookies their mass will bend the time-space continuum until point A will become point B and we'll be stuck in a loop where point A will become point B and we'll be stuck in a loop where point A will become point B
ScardyBobScardyBobJoin Date: 2009-11-25Member: 69528Forum Admins, Forum Moderators, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow
If the system collects data across multiple matches for each person, then an Elo-like rating system is the way to go. I've long considered AOE3's approach to team-based Elo-ratings to be workable in NS2.
If you only want it to use data from the previous match (i.e. not track ratings) then I think using the average player score/min is the best method. K/D or kills/min are more prone to fluctuations from the relative skill of the teams (e.g. an Archaea vs Archaea match would likely have a similar K/D to a rookie vs rookie one). However, skilled players typically do many more actions per minute which is somewhat captured in their score. You'd have to gather some actual match data to get the correct values, but you could do the splits like:
<50 score/min average = Rookie
50-150 score/min average = Intermediate
>150 score/min = Pro
Can the game read the player's gaming hours played as shown in their steam profile. Hours played can be somewhat related to skill. Sure you get the odd player who is amazing at other FPS games and can jump into NS2 as a noobie and dominate with only 60 or so hours played, however it would be nice for a rookie to know that they are entering a server filled with players who are "veterans" (maybe 500+ hours?) or terribly addicted NS2 tragics (like me 1300+ hours). I think it would be nice for rookie's to know that their first experience of NS2 isn't going to be an absolute stomping, deterring them from the game and likewise for experienced players to be able to join a server with people of similar skill level for a more enjoyable game.
Also all those people shouting about being forced into matchmaking, I'm pretty sure the OP never mentioned matchmaking or "rankings" determining where you can and cannot play.
I think the few hours it would take to add "hours played" on the scoreboard or to the shuffle would have retained tens of thousands more players by now just by avoiding the obviously stacked situations
there's a huge psychological impact when you play 4 games one night and they all suck versus when 1 of the 4 ended up being fun
@UncleCrunch... Just... No. Just no. Sorry. Your whole post, no. 8-|
Let's have a look at it. I'm not taking it personally and i won't answer to you personally. See it as another point of vue.
Let's summary a little from the good post i saw here.
@james888 A first step is better than no step @HeatSurge This entire discussion is moot if whatever "system" is not forced on players, which I am personally against @male_fatalities Yes me too... Having forced random balanced teams as server variable would fix that problem though. @Zaggy So, like adding Difficulty: Beginner, Difficulty: Average, Difficulty: Veteran in front of server names?
I like the idea, but how would you enforce this? What if green players join Veteran servers, do you kick them out? Or vice versa? @male_fatalities There would be no segregation, the rank of the server is not a pre-entry requirement. The rank of the server is calculated based on the players inside the server. I am 100% against segregation.
Yes, doing nothing won't save it, but missing the target neither. We already see the RR issue coming in.
@The_Flying_Fish what about comms? @Jekt wrote
Realistically, K:D is the most prevalent factor an individual can have in determining the win of a team and should be largest factor in determining a skill reference @Sarummay I rather have someone with 2:30 who killed 10 RTs and 3 phase gates on his own and frequently forces beacons, since he is way more useful then the first player. @Nachos Top fragging players help stop pressure which gives your other team members the ability to kill RTs and stuff.
They're all right from their POV. I do believe it's a good thing to have one or two pistoleros, and also builder, and polyvalent units.
*Pistoleros block/put some pressure.
*Builders do help gain and maintain the res flow.
*Polyvalent sentry, scoot, cover and / or replace temporarily a man(or girl) down.
This brings difficulty on how to calculate "skill". If it was a CS game K/D would suffice. But it's NS2 with a whole new dimension: RTS.
@Montyp To properly balance matches we should look to do it by balancing the game now or find a reliable way to balance the ready room. @xDragon its just intended to give an average idea of the skill level of the players in servers. @HeatSurge Basically, the questions that come to mind are: how are you going to make people use this "system" and how are you going to make them stay for entire matches. Just because it's NS2 doesn't necessarily mean it won't suffer from the same problems as other "ranked" games to some degree.
The RR is spotted again.
I may add there are several other issues:
The misinterpretation of the icon can lead to bad game experience. One must know precisely what it means, which is gonna be difficult (even simple as that) because most rookie players just didn't read the NS2 manual / tutorials in the first place. Some just won't take care of it unless they get obliterated 2 or 3 times. It can be annoying for them as it can be for the server "vets residents".
It's not a guaranteed level, just an indicator. So stacking will happen one way or another... Unfortunately.
Natural Grouping of population
Another problem will occur. Depending on the impact of the indicator it will introduce a grouping syndrome that prevent mixing players. Like social groups. It's the same mechanic.
Let's say you have 3 colors for 3 different skill (chevrons).
Red(low), green and Blue(high). Noting is B&W in this world so you must have players in between. So you end up with : Red, Yellow, Green, Purple, Blue. What do you think the majority of Yellow players will do ?
They mostly go to the color they feel comfortable. Brain prefers to win easy, it's a reflex. This is how our brains (yes! even gamer brains) are wired. If it was simple as : grabbing flower, give to girl, fuck; we would do that. Because the brain always tries to find shortcuts. No matter what. You can have a good example of this in clanplay. This is always the most effective (max damage) and fastest solution/asset that is chosen. We do that in every day life not even noticing it.
The natural player behavior will ultimately create groups that will be clearly separated. At some point they will be bored to play in lower sections you would say. True, but non the less, it will take time before it happens. Brains are like that. It's better to be king in hell than slave in heaven. or even call it laziness. That won't change, "human iz hooman".
The touch and go
At some point there's another behavior that will come to happen. Players will join to see what's happening and leave if doesn't suit them. A simple look at the score board and a few second as spectator will provide a good grasp on the game at hand. If it is stacked ... the spectator leaves...
So basically we have:
*Collecting datas is one problem.
*Calculating skill level only in K/D is another or won't be accurate enough to really help.
*Natural social behavior will come in the way.
*RR is spotted as a place to start by others.
*Stacking stays.
My proposal is acting only in RR which host the real issue: stacking. I'm not saying it's the best, any solution that deals with RR is gonna be, to my opinion, better than the OP's one. Because it is simply targeting the real issue.
This kind (whatever the solution) of RR distribution shouldn't be enforced on all server. Far from it. There would be some admin who choose to use it and some who won't, i say perfect. So players can choose what style they prefer. On one hand, free and high probability of stacking, on the other, distribution. If you join one or the other it means you accept the local rules. It is of course excluding clanplay which organize on it's own.
Beside of that it would be great to have in-game stuff that works:
-Simple interface for managing a little tournament (4 teams) or PCW. Possibly with database on UWE side. Not gonna kill the budget. It just need a responsive software.
-Better vote management.
-More option to vote for.
-Record and replay tools (no FRAPS involved).
Once people can organize themselves the other issues will be less painful. It is somehow connected to this very thread.
Your suggestion is a questionnaire.
Honestly, how annoying is it going to become?
"Before the game starts, all players must complete form 101522A.1. Please ensure that you complete the form in black and type in BLOCK CAPITALS. Failure to complete this form may result in you being kicked from the server."
Your suggestion is a questionnaire.
Honestly, how annoying is it going to become?
"Before the game starts, all players must complete form 101522A.1. Please ensure that you complete the form in black and type in BLOCK CAPITALS. Failure to complete this form may result in you being kicked from the server."
Just no.
No one said it has to be obligatory. Where did you read that ?
If one player wanted to do it in order to help balancing RR distribution (whatever the system), he just fills his/her preferences (or what he can do) in some option panel. The same option panel you find bindings/audio/graphic settings. So the algorithm on server side will perform better. Otherwise it stays simple random and that's it. No prerequisite, no kicking. Where's the issue ?
For 1 minute in the option panel, one player can help make the game experience better on any server that runs the RRmod. It stays initial NS2 random on server that doesn't run it (no harm done). 5 or 6 check-box in that form will suffice. Average ppl do more clicks than that just to launch the game from steam or configuring keys. And they probably are moving hands, reading team msg and clicking more while playing isn't it ?
You can even have different behavior like :
-No auto randomizing (but only those who go random).
-Only randomize RR at the moment game start...
...Tons of possibilities.
It's simply better to use/assign people for what they know or can do (or skill they have) than the actual situation which is exquisitely derelict.
Comments
Normally this is true and said scenario only happened about twice in over 300h game play (I don't comm every time and as normal player don't really watch the performance of others, so maybe more often), I just wanted to give an example why saying high kd=good (team) player doesn't have to be true and therefore using kd as the main indicator is a bad idea. Btw I wouldn't call spawn killing a marine who just respawns (while there are no other aliens around killing stuff) stop pressure.
@Jekt I am talking about pubs where you normally can just walk to the RT and start biting it and at least do a lot of damage before a marine even shows up (and even if he kills/chases away the skulk won't bother welding the RT) and a lot of people tend to ignore minimap and commander when they try to get inside an alien base, even if its well defend and in the meantime marines lose half of the map.
1. A great commander with a team that listens will win almost any game, how do you balance that?
2.Servers are community run, how do you enforce them to rank using kdr/WLR? What about different game modes?
3. When I was new to the game I was thankful for being put up against more skilled players. I got my face dragged through the mud by camo skulks and deadeye marines. But that's what got me hooked on the game!
If you have new players trying to learn together when they go against veterans they'll get smoked because no one will know the advanced tactics shared by the community. It will just lengthen the learning curve. To properly balance matches we should look to do it by balancing the game now or find a reliable way to balance the ready room.
Solid proof: a screenshot of a QL server list.
The rating system in QL doesn't work in my experience. I've been crushed 30-0 by good players, only to join their server later to spec and it says 'Skill Matched'. Me and some friends also regularly take games off of each other yet sometimes I won't be able to join because 'Your skill is too high' and other times it will be the reverse.
Player count is too low to have QL style ratings system used. There are also servers that already run KD based random teams, though they only do it based off of the previous round.
Honestly as a player that can join most servers and completely break the game, its not fun. The few players that do enjoy 'stacking' or w.e are generally players that just are not that good, and as such look to others they know can carry them. However this system really has absolutely nothing to do with that aspect - its more about establishing a general skill level per server, nothing more IMO. Let team stacking and random based on ELO/whatever be left up to mods and/or admins.
See the part where I wrote...
It is not a perfect system, but it does give a general indication of rookie, average players and anything above.
FYI, QL 1v1 is a game where two even players can actually go 30-0 against each other. Map knowledge alone can attribute to this.
Dragon nailed it.
It's a system which was supposed to mitigate "stacking" (i.e. 4 really good people against 4 randoms).
It didn't. The novelty lasted 1 week and then it died. I haven't played L4D2 for a long time now, so I'm not sure if it's still dead, but I am willing to bet it is. When people start losing (usually after 1 or 2 "segments"), they drop out and move on to a different game (server). They're rapidly replaced by noobs which perpetuate the loss or lose even worse.
It's not the same as the proposed NS2 changes, but some aspects of it are similar, and something to think about.
Basically, the questions that come to mind are: how are you going to make people use this "system" and how are you going to make them stay for entire matches. Just because it's NS2 doesn't necessarily mean it won't suffer from the same problems as other "ranked" games to some degree.
My only thoughts are about how you can make this information more useful.
1) The vast majority of players are going to be low rating. If you use average, the impact of good players is vastly understated.
2) One good player has more of an impact on the pub game than 2 good players (with one on each team), yet the average rating is lower in the former.
So, i think we should really only be interested in measuring the level of higher skilled players on the server. My suggestion would be to group certain rating ranges as beginner, intermediate, expert. Then in calculating server skill, you discard all beginner ratings or apply some weighting factor to discount them.
Like (avg(b)*n(b)*x(b) + so on for intermediate and expert) / 3, where the discounting factors x(e) > x(i) > x(b). Maybe you only need to discount beginner ratings, and not intermediate + expert. I dno.
And/or you could make it so when you hover over the 'chevron' rating, it'll give proportion of beginner/intermediate/expert as additional information e.g. 50%,40%,10%
I dno how you would solve the 2nd point though, if it's even an issue. I guess that is more an in-server matchmaking concern.
Just have the rankings based on time or no rankings at all, but you can just left click on the server and there´s an option to see the server list with all the playtimes of players.
KDR is not very reliable.It´s good for pinpointing good players on that server in that specific game.Over time you might see, that they do well consistently, but it only starts being consistent if you are really really good.
Rankings based on score might work, but then promoting to the next rank should be made exponentially growing.I´ve seen this system work in America´s Army.Different game of course, but I do remember being able to tell if a person was good by just looking at his Honor(ranking).
The other thread (about separating newbs from pros) deserves less attention than this one. The solution is in here!
People only post when they don't agree with something
No they don't. Oh, wait...
I disagree.
Dont get me wrong: its a great idea and i like it.
But dont expect too much from it.
For people who want to play with equal skill players its awesome.
But there always people out there who feels like godlike if there on 1st place with 50-2 on a rookie server.
I think we can all accept that that's the case (like it or not). What this aims to do is to give the rest of us, the majority (?), the chance to actually avoid that situation to begin with.
Accidentally joining a server where everyone is FAR worse than you is nigh on impossible to avoid at present. Anything that empowers players to find servers of suitable skill level has to be a good thing, then it's just down to promoting that mod as much as humanly possible so that the decent players actually get to hear about it, install it, and use it. (Better still, if it works out well, UWE could implement it in the browser by default, but let's not get ahead of ourselves).
That said. What about taking the idea upside down ?
If players are given an way to set a profile (announcing); it would be better for making a "controlled randomization", or Distribution. Of course players accept random, or admins enforce it on servers.
A profile would be submitted by a form (like an in-game option panel) that gives indications about the player way of playing: "Do you play Fade Y/N", "Commander : Yes/no i'm not ready", "Lieutenant : Are you a pistolero or a mecanic ?". Things like that. When randomizing the RR, the server have a clue about who is gonna playing what (or able play this or that).
-You could have the two commanders of the server not ending on the same side. 2 good commanders will definitely be really efficient because one knows the job of the other. So the one on the field "senses" what to do without being told, that is an F***ing asset.
-Avoiding total rookie getting in the chair (have to pass training map to unlock the option).
-Allow a better organization in team as each will probably do the part he announced he would do.
Then the time played as a role/unit can also be a parameter.
It can provide a way of doing things NorthBound or SouthBound (if i can say that). A 300hrs fade is more likely to be more efficient than a 2hrs fade. Server admins could set the server to choose for the alien the most skilled (optimizing) or on the contrary the less skilled (giving a chance to rookies). As a 300hrs fade can be an ass at welding
And if one wouldn't want to provide such information (i don't wanna command these days); you can uncheck it or even leaving the form blank.
How to?:
*I believe the form in the options menu is an easy job on UWE side.
*Server random algorithm would have to be improved a little but that's not difficult. Far from it.
*Players accept randomization (instead of joining A or
*Servers is set to randomize automatically
Ok there would be repetition sometimes. But that could help avoid the stacking syndrome on pubs.
what happens to the universe if I agree?
pretty scary question
So many cookies their mass will bend the time-space continuum until point A will become point B and we'll be stuck in a loop where point A will become point B and we'll be stuck in a loop where point A will become point B
If you only want it to use data from the previous match (i.e. not track ratings) then I think using the average player score/min is the best method. K/D or kills/min are more prone to fluctuations from the relative skill of the teams (e.g. an Archaea vs Archaea match would likely have a similar K/D to a rookie vs rookie one). However, skilled players typically do many more actions per minute which is somewhat captured in their score. You'd have to gather some actual match data to get the correct values, but you could do the splits like:
<50 score/min average = Rookie
50-150 score/min average = Intermediate
>150 score/min = Pro
Also all those people shouting about being forced into matchmaking, I'm pretty sure the OP never mentioned matchmaking or "rankings" determining where you can and cannot play.
there's a huge psychological impact when you play 4 games one night and they all suck versus when 1 of the 4 ended up being fun
I agree with you... wait... never mind.
Let's have a look at it. I'm not taking it personally and i won't answer to you personally. See it as another point of vue.
Let's summary a little from the good post i saw here.
Yes, doing nothing won't save it, but missing the target neither. We already see the RR issue coming in.
They're all right from their POV. I do believe it's a good thing to have one or two pistoleros, and also builder, and polyvalent units.
*Pistoleros block/put some pressure.
*Builders do help gain and maintain the res flow.
*Polyvalent sentry, scoot, cover and / or replace temporarily a man(or girl) down.
This brings difficulty on how to calculate "skill". If it was a CS game K/D would suffice. But it's NS2 with a whole new dimension: RTS.
The RR is spotted again.
I may add there are several other issues:
The misinterpretation of the icon can lead to bad game experience. One must know precisely what it means, which is gonna be difficult (even simple as that) because most rookie players just didn't read the NS2 manual / tutorials in the first place. Some just won't take care of it unless they get obliterated 2 or 3 times. It can be annoying for them as it can be for the server "vets residents".
It's not a guaranteed level, just an indicator. So stacking will happen one way or another... Unfortunately.
Natural Grouping of population
Another problem will occur. Depending on the impact of the indicator it will introduce a grouping syndrome that prevent mixing players. Like social groups. It's the same mechanic.
Let's say you have 3 colors for 3 different skill (chevrons).
Red(low), green and Blue(high). Noting is B&W in this world so you must have players in between. So you end up with : Red, Yellow, Green, Purple, Blue. What do you think the majority of Yellow players will do ?
They mostly go to the color they feel comfortable. Brain prefers to win easy, it's a reflex. This is how our brains (yes! even gamer brains) are wired. If it was simple as : grabbing flower, give to girl, fuck; we would do that. Because the brain always tries to find shortcuts. No matter what. You can have a good example of this in clanplay. This is always the most effective (max damage) and fastest solution/asset that is chosen. We do that in every day life not even noticing it.
The natural player behavior will ultimately create groups that will be clearly separated. At some point they will be bored to play in lower sections you would say. True, but non the less, it will take time before it happens. Brains are like that. It's better to be king in hell than slave in heaven. or even call it laziness. That won't change, "human iz hooman".
The touch and go
At some point there's another behavior that will come to happen. Players will join to see what's happening and leave if doesn't suit them. A simple look at the score board and a few second as spectator will provide a good grasp on the game at hand. If it is stacked ... the spectator leaves...
So basically we have:
*Collecting datas is one problem.
*Calculating skill level only in K/D is another or won't be accurate enough to really help.
*Natural social behavior will come in the way.
*RR is spotted as a place to start by others.
*Stacking stays.
My proposal is acting only in RR which host the real issue: stacking. I'm not saying it's the best, any solution that deals with RR is gonna be, to my opinion, better than the OP's one. Because it is simply targeting the real issue.
This kind (whatever the solution) of RR distribution shouldn't be enforced on all server. Far from it. There would be some admin who choose to use it and some who won't, i say perfect. So players can choose what style they prefer. On one hand, free and high probability of stacking, on the other, distribution. If you join one or the other it means you accept the local rules. It is of course excluding clanplay which organize on it's own.
Beside of that it would be great to have in-game stuff that works:
-Simple interface for managing a little tournament (4 teams) or PCW. Possibly with database on UWE side. Not gonna kill the budget. It just need a responsive software.
-Better vote management.
-More option to vote for.
-Record and replay tools (no FRAPS involved).
Once people can organize themselves the other issues will be less painful. It is somehow connected to this very thread.
Honestly, how annoying is it going to become?
"Before the game starts, all players must complete form 101522A.1. Please ensure that you complete the form in black and type in BLOCK CAPITALS. Failure to complete this form may result in you being kicked from the server."
Just no.
No one said it has to be obligatory. Where did you read that ?
If one player wanted to do it in order to help balancing RR distribution (whatever the system), he just fills his/her preferences (or what he can do) in some option panel. The same option panel you find bindings/audio/graphic settings. So the algorithm on server side will perform better. Otherwise it stays simple random and that's it. No prerequisite, no kicking. Where's the issue ?
For 1 minute in the option panel, one player can help make the game experience better on any server that runs the RRmod. It stays initial NS2 random on server that doesn't run it (no harm done). 5 or 6 check-box in that form will suffice. Average ppl do more clicks than that just to launch the game from steam or configuring keys. And they probably are moving hands, reading team msg and clicking more while playing isn't it ?
You can even have different behavior like :
-No auto randomizing (but only those who go random).
-Only randomize RR at the moment game start...
...Tons of possibilities.
It's simply better to use/assign people for what they know or can do (or skill they have) than the actual situation which is exquisitely derelict.