That idea of yours wouldn't really help out new players, as practically all servers would exceed their skill level. It also wouldn't prevent team stacking (be it unintentional or not) on those servers.
Its a double idea. There is the servers ranked by average/median hour played of the players on them and there are the rookie only servers where rookies can play against each other
The only thing hours works on is pure rookies and "experienced" (not necessarily good) players, and the number of hours separating the two varies greatly. It does NOT separate "experienced" from good/great/pro players. Thing is, rookie servers NEED experienced players, so separating by hours makes them worthless to begin with.
What we need is a complicated metric akin to elo. Something that takes into account score possibly split by individual action, kills and deaths dependant on the other players metric and possibly that of any assists, and wins/losses taking into account game duration, percentage of the game actually played in, and of course the composition of both teams.
With a good system, we don't even NEED to separate servers. Rankings will help add a sense of progression to the overall game, retaining players for longer, and the ratings gained/lost by playing with or against people outside of your own level will cause people who care (typically higher end players) to gravitate away from just noobstomping. All we really need is an aggregate number applied to the server based on the current players so people know what they are getting into. Another aggregate applied to teams could help balancing those as well by choice rather than force.
Of course, all of that is very difficult to do, let alone do right. It's also not very viable in a heavily modded game (hint hint).
My other suggestion is around team stacking and the 'Vote Random' option. Most games I play start with 'Vote Random', however the teams continue to seem stacked. Would be good if NS2 recorded in the background the following for each player:
Points per round / total number of rounds played = Player Skill Level
'Vote Random' could then use this figure for each player to work out evenly matched teams.
Random is broken, slightly, a lot of people who want to stack have figured out that if you have already joined a team before random mode is activated, you stay on the team you already joined. The stackers are usually the ones who leap on teams as soon as possible so that they may avoid getting random'd
The best way around this is to start a random vote BEFORE the round is over, quite amusingly these players almost always drop if they are forced to random.
When using K/D as a major balance factor, you must assume all players care about deaths and attempt to preserve a good K/D ratio.
Amusingly enough, if the game didn't show or keep track of the K/D ratio half of the stacking complaints would probably disappear. Although I'm wanting to say I remember this was being discussed during beta...
I doubt the majority of people who join before the random vote are actually deliberately stacking, and instead just joining the team they want to play (not in terms of the people in the team, but the team itself i.e. aliens or marines). Whilst some people do join early to bypass the vote and stack, I think the majority are merely just want to play aliens/marines (perhaps cause they've played the other side several rounds in a row, or perhaps they just prefer one side, or perhaps merely feel like playing aliens/marines).
Having said that, it does throw out the balancing conerns (that random is supposed to address) out the window if people do do this (even if people aren't joining to skill-stack). So I would prefer that randomize readyroom is replaced with randomize all.
One of the issues with the randomize readyroom vote is that, after a map change, people will load the game at different rates. An early randomize readyroom vote (when people haven't already joined teams) will most likely fail, as not everybody has loaded the new map at this point and thus can't vote. Then when a random vote is done later, most people have already joined teams.
In any case, I think people need to get used to the fact that not every game is going to be even. Nobody has an accurate knowledge of every single player in the servers skill level. And there are far too many other variables that may throw a game off (perhaps some good players may have an off game, or some mediocre players do unpredictably well). Maybe some players end up playing gorge, or a lifeform they aren't proficient with, or a strategy they aren't well practiced with. Maybe a team is good, but has bad teamwork or a bad commander.
It's been a long long time since I've been on a server that has been skill-stacked so blatently that it must be deliberate. Sure, I've had my fair share of uneven games (ranging from only slightly uneven, to "there's no nope" levels of uneven), but I've long since come to terms with the fact that this is a symptom of most computer games, and that ns2 is no better or worse in this regard. You have a whole bunch of amazing players, good players, mediocre players, and bad players all thrown into one big melting pot, and it doesn't always seperate out like you'd hope (and even if it does, still doesn't guarantee a particular match will be even).
The one thing that seperates out ns2 from other games is that it has a higher learning curve, and thus brand new players will have a greater effect on balance as they can't just jump into the game like any traditional FPS. Even then, even if you make sure each side has the same amount of greens, there are still a lot of variables. Some new players are less willing or able to learn than others, and some might have a lot FPS experience whilst others might not, thus one green player may have 10x the impact that another does. Also, maybe there are people on one team trying to teach and coordinate the newbies and none on the other. New players also tend to flock to marines, and be more individually proficient as most people know how to aim and shoot. This typically results in newbies going marines most of the time (throwing out the balance if there's a sale and thus a lot on). However, on the other side of the coin, in the average pub game, marines require more teamwork. So said players may do better on aliens.
Anyway, it's a complicated issue. I'd like to see some sort of matchmaking/balancing system, but I really don't think that it will have the impact everybody would hope for - even if it's masterly designed and we have enough of a playerbase to make it work.
At the very least, lets change randomize readyroom to randomize all. Maybe spectators shouldn't be subject to this, but then it'd just be another way to bypass the random vote (i.e. jump in spectate after the vote starts, then join the team you want after the vote passes).
My other suggestion is around team stacking and the 'Vote Random' option. Most games I play start with 'Vote Random', however the teams continue to seem stacked. Would be good if NS2 recorded in the background the following for each player:
Points per round / total number of rounds played = Player Skill Level
'Vote Random' could then use this figure for each player to work out evenly matched teams.
Random is broken, slightly, a lot of people who want to stack have figured out that if you have already joined a team before random mode is activated, you stay on the team you already joined. The stackers are usually the ones who leap on teams as soon as possible so that they may avoid getting random'd
The best way around this is to start a random vote BEFORE the round is over, quite amusingly these players almost always drop if they are forced to random.
at least some versions of the vote random reset the game when they are used, so using them at the end of the round means the round ends prematurely (before concede).
And yeah, I usually want to play a particular side at a particular time for whatever reasons, so I'ff often make use of my binds to join the instant the round ends. Kind of hard to say someone is stacking merely by joining a side though, especially when there are 0-1 players on there by that point and you aren't even looking at their names.
It doesn't help that the vote system uses F1 and F2, so if there is space on the marine team and I vote to randomise, I will be placed on the marine team as I bound F1 to join marines and F2 to join aliens, like a lot of ex-NS1 players did...
It doesn't help that the vote system uses F1 and F2, so if there is space on the marine team and I vote to randomise, I will be placed on the marine team as I bound F1 to join marines and F2 to join aliens, like a lot of ex-NS1 players did...
Or just half way intelligent players who know about the bind function, since that's pretty much the most natural place to put it. Of course, you could always move it somewhere else.
It doesn't help that the vote system uses F1 and F2, so if there is space on the marine team and I vote to randomise, I will be placed on the marine team as I bound F1 to join marines and F2 to join aliens, like a lot of ex-NS1 players did...
Or just half way intelligent players who know about the bind function, since that's pretty much the most natural place to put it. Of course, you could always move it somewhere else.
I could, but being a RooBear of very little brain, there's no way I'd remember where I put it
1. Use something that ranks player skill such as gametime, K/D, Win/Loss, ELO or TrueSkill (from worst to best).
2. Allow cvars that prohibit too good players joining the game.
3. Possibly implement a cvar to enforce players to be of equal skill on both teams, either before joining or auto-team'ed
4. Experiment with server admins to see how this works out in practice
I think in NS1, at least sometimes, there was kind of an honor code for better players not to spawncamp or do something retarded that made the game horrible for everyone. I think too few players followed to suit, but yes there was this. And there were some who counter-stacked the best players to prevent a walkover. When you are part of some community (YO-clan hello?), you feel bit of a moral obligation to contribute to the gameplay experience, when its not some "random server". Competitive players had a bad habbit of ruining everyone's game just because they could.
You could also look into ways to make the game mechanics have more exponential skill curve. And limit the amount of skill variance a single player will have (a bad player shouldn't be able to stomp bad players 10-1 on a good day).
There isn't enough players to do any sort of balancing.
The problem is lack of players.
This.
Also, for team balance, for crying out loud, just TRY KDR. I know it's not perfect, but it's easy as hell to make, works to a degree GUARANTEED, gives you a skill level in one game, ...just try it already.
And even though KDR isn't the best balance factor, IT'S THE ONE PEOPLE CARE MOST ABOUT. They don't care about Harry Harvester who is the bane of all alien harvesters killing them in droves, they care about [xXxSNipEz] Superpwner with the 25/2 KDR. If they see Superpwner dropped into a team full of suck, they don't care about losing the game to Harry Harvester. ...that's a really good name.
If they did implement KDR balancing, we would never see xenocide used again. ( I've had many games where the bulk of my deaths were from being a skulk bomb. It's goofy that those deaths count against you anyway, but that's for another thread I suppose.
Hokay, so there's ALOT of focus based off of balancing teams, strength vs weakness judged by KDR. For anyone who has spent more than 50 hours in the game I would expect you to have figured out this is NOT CoD, KDR means nothing!
I have over 1000 hours logged in the game majority as commanding, and when I'm not in the chair especially after build 250 was released I prefer battle gorge over skulk. I make it my soul mission to support my team by building structures, destroying res towers and stoping enemy advances. My KDR is normally 3/0 while everyone else is in double digit kills. Does this make my contribution any less important? When I can tie up 3-5 marines with a few clogs/hydras and only 1 skulk while busing forward structures I'd say that those contributions as a support role often mean a lot more than going Rambo and racking up a High KDR.
The game rewards score points for being a good team player, if you want better balanced games I would recommend balancing off of score-per-minute especially when you play on a server with regulars.
TacticalGamer servers have been using this technique for a long time now, for the most part it works pretty well. This puts the best team players on opposite teams. Personally I'd rather have 4 people who can follow orders and can communicate with eachother over 4 Rambos who get a 10-2 KDR but just run around doing their own thing.
The biggest change we can make (other than being able to have frequent updates to the 250 build because that's part of the balance issue) is to have commanders on all servers step up and be a leader for your team. I've played on a lot of servers where commanders don't speak. No orders, no strategy, advice to newbies especially on newbie friendly servers. If this changed the people coming to this game for the first time would see that it grisly is something that breaks the mould of the standard archetypes.
If you're still looking for an external fix, 250 needs to be rebalanced because for newbies playing against the experienced its BRUTAL. If you want to balance teams I say go by score per minute
dePARAJoin Date: 2011-04-29Member: 96321Members, Squad Five Blue
Distracting the enemy is part of the teamplay.
If one "Rambo" is killing harvesters alone constantly, he get all my attention and help if im com. This "rambo" helping the team more than 5 marines camping a hallway against 1 gorge and die against 2 skulks.
Teamplay isnt running around in groups of 3-4 all the time. Every "role" is important.
If one skulk is killing RTs all the time noone would say he is an rambo. They would say "Hey, thats a good skulk. Biting RTs is part of the skulks role"
If 1 marine is doing the same, he is an rambo and doing nothing usefull running around alone.
That makes no sense.
Btw, you can have an high score by sitting around doing nothing usefull all the time with 3 suicide bombing runs in the endgame.
Balancing by score isnt the holy grale like some mentioned here.
You can get an insane score just by sitting in base and spamming babbler eggs and healing them to hatch. For some reason if you healspray a babbler egg and build it to full you get 5 points every time. You also get 5 for gorge building a cyst, although that's actually a useful thing to do.
Comments
Its a double idea. There is the servers ranked by average/median hour played of the players on them and there are the rookie only servers where rookies can play against each other
What we need is a complicated metric akin to elo. Something that takes into account score possibly split by individual action, kills and deaths dependant on the other players metric and possibly that of any assists, and wins/losses taking into account game duration, percentage of the game actually played in, and of course the composition of both teams.
With a good system, we don't even NEED to separate servers. Rankings will help add a sense of progression to the overall game, retaining players for longer, and the ratings gained/lost by playing with or against people outside of your own level will cause people who care (typically higher end players) to gravitate away from just noobstomping. All we really need is an aggregate number applied to the server based on the current players so people know what they are getting into. Another aggregate applied to teams could help balancing those as well by choice rather than force.
Of course, all of that is very difficult to do, let alone do right. It's also not very viable in a heavily modded game (hint hint).
Random is broken, slightly, a lot of people who want to stack have figured out that if you have already joined a team before random mode is activated, you stay on the team you already joined. The stackers are usually the ones who leap on teams as soon as possible so that they may avoid getting random'd
The best way around this is to start a random vote BEFORE the round is over, quite amusingly these players almost always drop if they are forced to random.
Amusingly enough, if the game didn't show or keep track of the K/D ratio half of the stacking complaints would probably disappear. Although I'm wanting to say I remember this was being discussed during beta...
Having said that, it does throw out the balancing conerns (that random is supposed to address) out the window if people do do this (even if people aren't joining to skill-stack). So I would prefer that randomize readyroom is replaced with randomize all.
One of the issues with the randomize readyroom vote is that, after a map change, people will load the game at different rates. An early randomize readyroom vote (when people haven't already joined teams) will most likely fail, as not everybody has loaded the new map at this point and thus can't vote. Then when a random vote is done later, most people have already joined teams.
In any case, I think people need to get used to the fact that not every game is going to be even. Nobody has an accurate knowledge of every single player in the servers skill level. And there are far too many other variables that may throw a game off (perhaps some good players may have an off game, or some mediocre players do unpredictably well). Maybe some players end up playing gorge, or a lifeform they aren't proficient with, or a strategy they aren't well practiced with. Maybe a team is good, but has bad teamwork or a bad commander.
It's been a long long time since I've been on a server that has been skill-stacked so blatently that it must be deliberate. Sure, I've had my fair share of uneven games (ranging from only slightly uneven, to "there's no nope" levels of uneven), but I've long since come to terms with the fact that this is a symptom of most computer games, and that ns2 is no better or worse in this regard. You have a whole bunch of amazing players, good players, mediocre players, and bad players all thrown into one big melting pot, and it doesn't always seperate out like you'd hope (and even if it does, still doesn't guarantee a particular match will be even).
The one thing that seperates out ns2 from other games is that it has a higher learning curve, and thus brand new players will have a greater effect on balance as they can't just jump into the game like any traditional FPS. Even then, even if you make sure each side has the same amount of greens, there are still a lot of variables. Some new players are less willing or able to learn than others, and some might have a lot FPS experience whilst others might not, thus one green player may have 10x the impact that another does. Also, maybe there are people on one team trying to teach and coordinate the newbies and none on the other. New players also tend to flock to marines, and be more individually proficient as most people know how to aim and shoot. This typically results in newbies going marines most of the time (throwing out the balance if there's a sale and thus a lot on). However, on the other side of the coin, in the average pub game, marines require more teamwork. So said players may do better on aliens.
Anyway, it's a complicated issue. I'd like to see some sort of matchmaking/balancing system, but I really don't think that it will have the impact everybody would hope for - even if it's masterly designed and we have enough of a playerbase to make it work.
At the very least, lets change randomize readyroom to randomize all. Maybe spectators shouldn't be subject to this, but then it'd just be another way to bypass the random vote (i.e. jump in spectate after the vote starts, then join the team you want after the vote passes).
at least some versions of the vote random reset the game when they are used, so using them at the end of the round means the round ends prematurely (before concede).
And yeah, I usually want to play a particular side at a particular time for whatever reasons, so I'ff often make use of my binds to join the instant the round ends. Kind of hard to say someone is stacking merely by joining a side though, especially when there are 0-1 players on there by that point and you aren't even looking at their names.
Or just half way intelligent players who know about the bind function, since that's pretty much the most natural place to put it. Of course, you could always move it somewhere else.
I could, but being a RooBear of very little brain, there's no way I'd remember where I put it
1. Use something that ranks player skill such as gametime, K/D, Win/Loss, ELO or TrueSkill (from worst to best).
2. Allow cvars that prohibit too good players joining the game.
3. Possibly implement a cvar to enforce players to be of equal skill on both teams, either before joining or auto-team'ed
4. Experiment with server admins to see how this works out in practice
I think in NS1, at least sometimes, there was kind of an honor code for better players not to spawncamp or do something retarded that made the game horrible for everyone. I think too few players followed to suit, but yes there was this. And there were some who counter-stacked the best players to prevent a walkover. When you are part of some community (YO-clan hello?), you feel bit of a moral obligation to contribute to the gameplay experience, when its not some "random server". Competitive players had a bad habbit of ruining everyone's game just because they could.
You could also look into ways to make the game mechanics have more exponential skill curve. And limit the amount of skill variance a single player will have (a bad player shouldn't be able to stomp bad players 10-1 on a good day).
This.
Also, for team balance, for crying out loud, just TRY KDR. I know it's not perfect, but it's easy as hell to make, works to a degree GUARANTEED, gives you a skill level in one game, ...just try it already.
And even though KDR isn't the best balance factor, IT'S THE ONE PEOPLE CARE MOST ABOUT. They don't care about Harry Harvester who is the bane of all alien harvesters killing them in droves, they care about [xXxSNipEz] Superpwner with the 25/2 KDR. If they see Superpwner dropped into a team full of suck, they don't care about losing the game to Harry Harvester. ...that's a really good name.
I'm going to start calling people this. LOL
Best thing I've ever read.
I have over 1000 hours logged in the game majority as commanding, and when I'm not in the chair especially after build 250 was released I prefer battle gorge over skulk. I make it my soul mission to support my team by building structures, destroying res towers and stoping enemy advances. My KDR is normally 3/0 while everyone else is in double digit kills. Does this make my contribution any less important? When I can tie up 3-5 marines with a few clogs/hydras and only 1 skulk while busing forward structures I'd say that those contributions as a support role often mean a lot more than going Rambo and racking up a High KDR.
The game rewards score points for being a good team player, if you want better balanced games I would recommend balancing off of score-per-minute especially when you play on a server with regulars.
TacticalGamer servers have been using this technique for a long time now, for the most part it works pretty well. This puts the best team players on opposite teams. Personally I'd rather have 4 people who can follow orders and can communicate with eachother over 4 Rambos who get a 10-2 KDR but just run around doing their own thing.
The biggest change we can make (other than being able to have frequent updates to the 250 build because that's part of the balance issue) is to have commanders on all servers step up and be a leader for your team. I've played on a lot of servers where commanders don't speak. No orders, no strategy, advice to newbies especially on newbie friendly servers. If this changed the people coming to this game for the first time would see that it grisly is something that breaks the mould of the standard archetypes.
If you're still looking for an external fix, 250 needs to be rebalanced because for newbies playing against the experienced its BRUTAL. If you want to balance teams I say go by score per minute
-Montyp
*drops mic*
If one "Rambo" is killing harvesters alone constantly, he get all my attention and help if im com. This "rambo" helping the team more than 5 marines camping a hallway against 1 gorge and die against 2 skulks.
Teamplay isnt running around in groups of 3-4 all the time. Every "role" is important.
If one skulk is killing RTs all the time noone would say he is an rambo. They would say "Hey, thats a good skulk. Biting RTs is part of the skulks role"
If 1 marine is doing the same, he is an rambo and doing nothing usefull running around alone.
That makes no sense.
Btw, you can have an high score by sitting around doing nothing usefull all the time with 3 suicide bombing runs in the endgame.
Balancing by score isnt the holy grale like some mentioned here.