Teams coming out of lobby are pretty awful lately

(MarsRover) #41

He posted 2 screenshots from a pre-match lobby.

Have you ever seen a pub lobby? People leaving, other people joining and being assigned to level team sizes? The second one is Stopwatch where the situation is far worse because the teams are assigned even earlier - when going into lobby, before people have a chance to leave.

Those screenshots have zero value in this discussion, other than proving again that assigning teams should be done as late as possible.

Actual player performance in past X games is vastly more accurate than any statistical (and unproven) method. If statistics were driving reality I should be in top 1% players, getting Godlikes left and right. I don’t.

No, your argument that total XP corresponds with player skill is outright wrong. It may be correlated if you compare lvl5 with lvl50. Correlated, but not caused. But let’s wait a few months, those same players are now lvl60 and lvl100. Is the lvl100 better? By how much? Why? Lvl200 vs lvl300?

The metric you propose doesn’t make any sense. It is strictly worse. Every team I join would be doomed because at lvl39 I would be grouped with the lowest level players on the server, including all the levels <5 that didn’t stick to max lvl5 servers. On the other side there’s a mid-level team. There are games where entire teams have less levels than I have.

The game uses a metric that tracks actual performance. It works as good as it can given other factors. The problems you point out are caused by:

  1. pub game nature
  2. assigning teams too early
  3. bad team compositions

They are not caused by using a different metric for judging player skill. They will not be fixed by switching to a different one, and certainly not to a worse one.

There is not much to be done about 1) except autobalancing when one team is 2 players short. The difference between OBJ and SW shows how important is 2), and there is still much room to improve. 3) is much harder to mitigate, I think there is much more to gain from 2).

(Maverix) #42

[left][/left][quote=“Amerika;138835”]
You just explained my point. DB account levels indicate experience/time played but not a person’s actual skill level, their competence or their will to play correctly or even learn to play correctly. I know people who are in their 40’s who don’t know how to move around maps properly or they spam the jump button when they shoot due to panicing or try to use a grenade launcher at point blank range. And people who just run in and try to blow up people with mines on proxy despite basically suiciding every time. All high account level players but not high level competent players who care about “crosshair discipline”.

Just because you spent time trying to get better doesn’t mean everyone does. Some people just jam out to music and blow off steam and play for hours because it’s fun regardless of how good they do.[/quote]

You’re not understand what I just said. I’ll simplify it. You’re saying that experience doesn’t correlate to skill. I’m saying that it does.

And I provided my own example of how experience develops into skill when I bring up crosshair discipline. I was trying to imply that because of the situation I was in (having a very low end computer), I was able to adapt to the game. Regardless of the game, and regardless of whether or not you play this game casually or competitively, you will undoubtedly gain skill with the experience you gain.

I really don’t understand why you keep insisting that experience does not equal skill. Remember when I gave you the definition of the word “skill: the ability to do something well; expertise.” (Definition Source). What does expertise mean? “To have abundant experience in a field; expert skill” ([url=“https://www.google.com/search?espv=2&q=expertise&oq=ex&gs_l=serp.3.0.35i39j0i67l9.183252.183460.0.184948.2.2.0.0.0.0.168.322.0j2.2.0.cprnk%2Cegame%3Dlto…0…1.1.64.serp…0.2.322.2h0Uoqsk0NE”]Definition Source). Notice how both words are referred to each other in both definitions. They are both different words, but they correlate with each other.

(Mrarauzz) #43

Practice doesn’t make perfect, perfect practice makes perfect.

(torsoreaper) #44

He posted 2 screenshots from a pre-match lobby.

Have you ever seen a pub lobby? People leaving, other people joining and being assigned to level team sizes? The second one is Stopwatch where the situation is far worse because the teams are assigned even earlier - when going into lobby, before people have a chance to leave.

Those screenshots have zero value in this discussion, other than proving again that assigning teams should be done as late as possible.

Actual player performance in past X games is vastly more accurate than any statistical (and unproven) method. If statistics were driving reality I should be in top 1% players, getting Godlikes left and right. I don’t.

No, your argument that total XP corresponds with player skill is outright wrong. It may be correlated if you compare lvl5 with lvl50. Correlated, but not caused. But let’s wait a few months, those same players are now lvl60 and lvl100. Is the lvl100 better? By how much? Why? Lvl200 vs lvl300?

The metric you propose doesn’t make any sense. It is strictly worse. Every team I join would be doomed because at lvl39 I would be grouped with the lowest level players on the server, including all the levels <5 that didn’t stick to max lvl5 servers. On the other side there’s a mid-level team. There are games where entire teams have less levels than I have.

The game uses a metric that tracks actual performance. It works as good as it can given other factors. The problems you point out are caused by:

  1. pub game nature
  2. assigning teams too early
  3. bad team compositions

They are not caused by using a different metric for judging player skill. They will not be fixed by switching to a different one, and certainly not to a worse one.

There is not much to be done about 1) except autobalancing when one team is 2 players short. The difference between OBJ and SW shows how important is 2), and there is still much room to improve. 3) is much harder to mitigate, I think there is much more to gain from 2).[/quote]

Believe it or not, I actually I agree with a lot of what you say. Assigning teams should be done as late as possible. I also have tried to say a few times that skill and level are correlated which some people disagree with. I am glad you agree there is a correlation. Is there a convexity to the level at which you get diminishing returns on skill versus level? of course.

Ironically, you counter your own argument where you say my comparison to XP and skill is wrong, then say it is correlated so you agree, then counter your point by saying you don’t want to be with level 5 players which means you are implying that level DOES imply skill. So not sure your final decision on skill vs level so I will just give you the benefit of the doubt that you know level IS correlated with skill.

Anyone who disagrees that level is correlated with skill, maybe we can get the developers to make it so you are always put with random players of < lvl 5 for the rest of the time you play dirty bomb. Obviously you’d have no problem with this suggestion since skill and level have no correlation. Or maybe you can start a new thread where you petition to remove min 10 servers, max level servers, and the lvl requirement for playing ranked. Because obviously “level doesn’t mean anything”.

(MarsRover) #45

I shouldn’t have written “skill”. I meant “performance”, which is a sum of skill and experience.
The only thing this correlation does correctly is saying that statistically a lvl50 player will do better than a lvl5. But that is mostly due to experience, not skill. Level does not correlate with pure skill at playing FPS games at all. The are FPS pros playing DB for the first time. There are people playing DB for tens of hours with their pro gaming mouse set to 3200DPI that cannot hit the broad side of a barn. So the only thing left is correlation with experience, but it breaks very quickly at low and high levels, which makes it totally useless at predicting performance.

I reiterate again

  1. replacing an actual player performance metric with a statistical one would harm the balance
  2. the current metric is not a cause of imbalance
    I’m not going to repeat myself as to why. In my previous post I listed all the aspects of imbalance, perceived or real.

Also populistic methods like posting some screenshots of prematch lobbies with no history behind them does not help your case. Especially screenshots from SW, which basically has RNG-based balance.

I was talking about their experience, not skill. You can reach lvl5 without even playing on every side of every map. This is an experience gap that cannot be overcome with skill.

But at some point you get all the experience you’re going to get, and then it’s a matter of skill. There are people less than half by level who absolutely wreck me. We both know all the maps/mercs/abilities - the playing field is even. They are just better at clicking heads than me. So now take a very typical lobby - me at 39, two guys like I described at 15, and the rest somewhere lower all the way to a few twos and threes that may play their first match ever on this map.

Balancing by level would mean me and all the inexperienced lowest levels against two more skilled players and the rest more experienced players. Balanced? Nope. This is why balancing by level is such a terrible idea.

Balancing by actual performance, like it is now, would result in those two being in opposite teams. The teams are as balanced as they can be, but my team looks like stacked because I have number 39 under my nickname. This is why levels as they are now need to go.

Those two are not about skill. They’re there to set a certain minimum level of experience and as a shield from cheaters. They’re there for people wanting to play a match where everyone knows the basics. You expect that a guy playing Sparks there to not help you up manually, you expect people to know the map, what the objectives are, etc. It’s not about skill. It is about experience, which you get only by playing the game for a certain amount of time.

Also, DB if F2P. Nothing stops a cheater from creating another account. Wins and losses matter in competitive. It is gated by level as it is the easiest way to prevent cheaters.

(DMaster2) #46

[quote=“Kingsley;138931”]By the way, I don’t know, maybe try joining a team…?
Being that you have these opinions, make a team yourself.
Why shouldn’t players play with their friends?
Jesus.
Stop it.
Go Away.[/quote]
Because that may compromise the little balance we have in objective now. Balance >>>> Party matchmaking. I have a couple of friends playing DB, but that doesn’t mean people that don’t have friends that share this interest can’t get balanced matches because people want to stack teams.

(Lumi) #47

[quote=“DirtyMerc;138372”]http://i.imgur.com/h6jc0D7.jpg

this is what matchmaking does every time still it does balance but very rarely … join the server where you can see more ppl playing like 12/16 in servers [/quote]

In this pic the most frightening thing is not the level distribution, but the pings…

(Black) #48

[quote=“Amerika;138810”][quote=“Maverix_GT;138795”][quote=“Amerika;138757”][quote=“torsoreaper;138754”][quote=“Amerika;138751”]

That’s not a solution. That’s just something you want to achieve. How do you actually achieve balance?[/quote]

I know level != skill but I would say let’s try making teams based on level for a patch cycle and see if things are better or worse. While a guy who is a level 3 but has played a ton of CS might be able to kill a level 20 person every time, it doesn’t help when he does that while letting their mechanic repair the EV or if he doesn’t know how to flank because he hasn’t even learned the map yet.[/quote]

Why? Most matches become imbalanced when players swap sides, players leave (and it takes 30-60 seconds for a new player to join and help out) or because the balance system which assigns an actual skill rating to everyone based on actual performance couldn’t factor in that a guy is on the phone while playing, listening to music, playing a new merc they have never played or something else.

How does balancing by levels resolve any of this? That’s what I want people who keep throwing out this suggestion to do. Show, logically, how it works beyond being a placebo for a few people who obsess over numbers and think that a level in DB is the same thing as a level in an MMO.[/quote]

I agree with everything but the part about levels. You can say level isn’t everything, and that’s true SOMETIMES. I’ve seen a level 60 have crap aim. And I’ve seen a level 6 completely wreck the enemy team on his own. But for the most part, difference is, I would expect a level 30 to hesitate when hearing the sound of Proxy’s mine behind a corner and slowly peak it over a level 5. I would expect a level 30 Vasilli to throw his sensor in a vital spot that benefits his team the most. I would expect a level 30 nader not to trigger their martyrdom when a medic is about to revive them. I would expect a level 30 player to dodge an airstrike from skyhammer.

And I repeat, level doesn’t always represent how good a player will be. But it holds a lot more value that what some people think. I’m not debating the topic of balance in this post. Just my opinion that there is definitely some value to people’s level. and that it shouldn’t be discredited.[/quote]

Nobody is arguing that. Level is an indicator of experience but not an indicator of skill, competence or your current state of mind/play. It’s not used to balance teams and it shouldn’t be. You can be a crap level 60 and be balanced appropriately in Dirty Bomb. If you were a pretty bad player but played a lot and had a high level you’d get punished by any system that tries to balance based on time played. And your team would get punished as well.[/quote]

Amerika please, let’s not argue this.
You and I both know that experience STRONGLY correlates with skill. There really is no arguing about that. Level IS an indicator of skill, however I do I agree it’s not 100% accurate, it’s way better than determining a players skill over their recent 10 games.

The only argument I see is a player just coming from years of cs:go practice on dirty bomb in which case he will be a lot better, however the more he plays dirty bomb the better he will be.

The best way to measure a person’s skill in Dirty bomb is simply to get their average score per minute. A players worth in Dirty bomb is dependent on how many points they can earn, I mean you see it every time, the winning team has more points.

(MarsRover) #49

No it’s not. I’m almost 40 now, I’m not twice as good as at level 20. I pretty much stopped progressing around that time because I reached my physical limits. Actually, I’m probably worse now - at that time I was playing DB for ~3h a day, now I only play for 30-60 minutes. I have about 33% accuracy.

Now let’s say someone truly skilled, accuracy ~50%, is also at level 40. He has the same amount of experience, but he has much higher skill. Why should we be equal when balancing?

But that’s probably exactly what the current system does. And AFAIK it’s last 25 games now, it was 10 but they changed it.

(Amerika) #50

[quote=“BlackFro;139315”][quote=“Amerika;138810”][quote=“Maverix_GT;138795”][quote=“Amerika;138757”][quote=“torsoreaper;138754”][quote=“Amerika;138751”]

That’s not a solution. That’s just something you want to achieve. How do you actually achieve balance?[/quote]

I know level != skill but I would say let’s try making teams based on level for a patch cycle and see if things are better or worse. While a guy who is a level 3 but has played a ton of CS might be able to kill a level 20 person every time, it doesn’t help when he does that while letting their mechanic repair the EV or if he doesn’t know how to flank because he hasn’t even learned the map yet.[/quote]

Why? Most matches become imbalanced when players swap sides, players leave (and it takes 30-60 seconds for a new player to join and help out) or because the balance system which assigns an actual skill rating to everyone based on actual performance couldn’t factor in that a guy is on the phone while playing, listening to music, playing a new merc they have never played or something else.

How does balancing by levels resolve any of this? That’s what I want people who keep throwing out this suggestion to do. Show, logically, how it works beyond being a placebo for a few people who obsess over numbers and think that a level in DB is the same thing as a level in an MMO.[/quote]

I agree with everything but the part about levels. You can say level isn’t everything, and that’s true SOMETIMES. I’ve seen a level 60 have crap aim. And I’ve seen a level 6 completely wreck the enemy team on his own. But for the most part, difference is, I would expect a level 30 to hesitate when hearing the sound of Proxy’s mine behind a corner and slowly peak it over a level 5. I would expect a level 30 Vasilli to throw his sensor in a vital spot that benefits his team the most. I would expect a level 30 nader not to trigger their martyrdom when a medic is about to revive them. I would expect a level 30 player to dodge an airstrike from skyhammer.

And I repeat, level doesn’t always represent how good a player will be. But it holds a lot more value that what some people think. I’m not debating the topic of balance in this post. Just my opinion that there is definitely some value to people’s level. and that it shouldn’t be discredited.[/quote]

Nobody is arguing that. Level is an indicator of experience but not an indicator of skill, competence or your current state of mind/play. It’s not used to balance teams and it shouldn’t be. You can be a crap level 60 and be balanced appropriately in Dirty Bomb. If you were a pretty bad player but played a lot and had a high level you’d get punished by any system that tries to balance based on time played. And your team would get punished as well.[/quote]

Amerika please, let’s not argue this.
You and I both know that experience STRONGLY correlates with skill. There really is no arguing about that. Level IS an indicator of skill, however I do I agree it’s not 100% accurate, it’s way better than determining a players skill over their recent 10 games.

The only argument I see is a player just coming from years of cs:go practice on dirty bomb in which case he will be a lot better, however the more he plays dirty bomb the better he will be.

The best way to measure a person’s skill in Dirty bomb is simply to get their average score per minute. A players worth in Dirty bomb is dependent on how many points they can earn, I mean you see it every time, the winning team has more points.

[/quote]

Wait, first you say that determining balance by level is better than using a skill rating formula which the game currently uses (last 25 games I’ve been told as it was upped from 10 recently). Then you go on to say that the game should use score per minute, a value that would be used heavily in a skill rating formula, should be used. I am so incredibly confused by this.

People do not learn linearly over time. They have ceilings and at some point for people they stop learning. Either this is because they care to learn, they physically can’t get better, they stop caring, they play less, they have distractions or perhaps their good mouse broke and are now using a trackpad on a laptop but still want to play.

I am level 67 now. I personally think I hit my performance capped somewhere around level 10-12 where, by that amount of time, I had played all the maps, learned the movement system and learned all the mercs fairly well that were in the game at that point. I’m not any better now and I’m definitely not 6 times better. The only difference is I know more mercs and I know an extra map.

Other people will cap out much earlier and some much later based on how they learn or how quickly they choose to learn. Also, skill diminishes with less time played typically. If you don’t’ put in the practice and keep up with it, you get worse. Balancing by level doesn’t account for any of this at all. It in fact punishes people. But a skill formula adjusts for this and scales over time to a person’s approximate worth to a team. Which is something an account level can’t do.

I use levels as a cursory glance to see if there is anybody I should pay attention to on a map. After the first couple of minutes though I have a pretty good idea of their actual skill level and a lot of times it doesn’t match their account level.

I’ll simply never understand people wanting to use something that is only a small part of measuring actual skill as a system for balance. It just doesn’t work for all the reasons I’ve cited here and more.

@torsoreaper I’ve given multiple logical reasons why balance can’t be determined based solely on time played. You can disagree with my points and offer a counter-argument per point as to why I am wrong but I’ve yet to see one that isn’t just a number and equation that was made up on the spot to add false weight to a debate.

@Maverix_GT To reply minus the condescension, experience is a part of determining skill. It does not equal skill. I’ve explained this pretty well in multiple posts over many threads now. Experience/time played does not equal skill level. Or are you arguing this?

(Black) #51

[quote=“Amerika;139333”][quote=“BlackFro;139315”][quote=“Amerika;138810”][quote=“Maverix_GT;138795”][quote=“Amerika;138757”][quote=“torsoreaper;138754”][quote=“Amerika;138751”]

That’s not a solution. That’s just something you want to achieve. How do you actually achieve balance?[/quote]

I know level != skill but I would say let’s try making teams based on level for a patch cycle and see if things are better or worse. While a guy who is a level 3 but has played a ton of CS might be able to kill a level 20 person every time, it doesn’t help when he does that while letting their mechanic repair the EV or if he doesn’t know how to flank because he hasn’t even learned the map yet.[/quote]

Why? Most matches become imbalanced when players swap sides, players leave (and it takes 30-60 seconds for a new player to join and help out) or because the balance system which assigns an actual skill rating to everyone based on actual performance couldn’t factor in that a guy is on the phone while playing, listening to music, playing a new merc they have never played or something else.

How does balancing by levels resolve any of this? That’s what I want people who keep throwing out this suggestion to do. Show, logically, how it works beyond being a placebo for a few people who obsess over numbers and think that a level in DB is the same thing as a level in an MMO.[/quote]

I agree with everything but the part about levels. You can say level isn’t everything, and that’s true SOMETIMES. I’ve seen a level 60 have crap aim. And I’ve seen a level 6 completely wreck the enemy team on his own. But for the most part, difference is, I would expect a level 30 to hesitate when hearing the sound of Proxy’s mine behind a corner and slowly peak it over a level 5. I would expect a level 30 Vasilli to throw his sensor in a vital spot that benefits his team the most. I would expect a level 30 nader not to trigger their martyrdom when a medic is about to revive them. I would expect a level 30 player to dodge an airstrike from skyhammer.

And I repeat, level doesn’t always represent how good a player will be. But it holds a lot more value that what some people think. I’m not debating the topic of balance in this post. Just my opinion that there is definitely some value to people’s level. and that it shouldn’t be discredited.[/quote]

Nobody is arguing that. Level is an indicator of experience but not an indicator of skill, competence or your current state of mind/play. It’s not used to balance teams and it shouldn’t be. You can be a crap level 60 and be balanced appropriately in Dirty Bomb. If you were a pretty bad player but played a lot and had a high level you’d get punished by any system that tries to balance based on time played. And your team would get punished as well.[/quote]

Amerika please, let’s not argue this.
You and I both know that experience STRONGLY correlates with skill. There really is no arguing about that. Level IS an indicator of skill, however I do I agree it’s not 100% accurate, it’s way better than determining a players skill over their recent 10 games.

The only argument I see is a player just coming from years of cs:go practice on dirty bomb in which case he will be a lot better, however the more he plays dirty bomb the better he will be.

The best way to measure a person’s skill in Dirty bomb is simply to get their average score per minute. A players worth in Dirty bomb is dependent on how many points they can earn, I mean you see it every time, the winning team has more points.

[/quote]

Wait, first you say that determining balance by level is better than using a skill rating formula which the game currently uses (last 25 games I’ve been told as it was upped from 10 recently). Then you go on to say that the game should use score per minute, a value that would be used heavily in a skill rating formula, should be used. I am so incredibly confused by this.

People do not learn linearly over time. They have ceilings and at some point for people they stop learning. Either this is because they care to learn, they physically can’t get better, they stop caring, they play less, they have distractions or perhaps their good mouse broke and are now using a trackpad on a laptop but still want to play.

I am level 67 now. I personally think I hit my performance capped somewhere around level 10-12 where, by that amount of time, I had played all the maps, learned the movement system and learned all the mercs fairly well that were in the game at that point. I’m not any better now and I’m definitely not 6 times better. The only difference is I know more mercs and I know an extra map.

Other people will cap out much earlier and some much later based on how they learn or how quickly they choose to learn. Also, skill diminishes with less time played typically. If you don’t’ put in the practice and keep up with it, you get worse. Balancing by level doesn’t account for any of this at all. It in fact punishes people. But a skill formula adjusts for this and scales over time to a person’s approximate worth to a team. Which is something an account level can’t do.

I use levels as a cursory glance to see if there is anybody I should pay attention to on a map. After the first couple of minutes though I have a pretty good idea of their actual skill level and a lot of times it doesn’t match their account level.

I’ll simply never understand people wanting to use something that is only a small part of measuring actual skill as a system for balance. It just doesn’t work for all the reasons I’ve cited here and more.
[/quote]

I never disagreed with any of this. You are correct, there is differences when it comes to learning, however learning requires experience. Yes, some people will learn faster and slower than others but there is no such thing as not getting any better at the game. There is no such thing as a “performance cap”, you are always learning and getting better at the game. What you are probably referring to is a drop in the learning/improvement rate. There is no such thing as being a master or perfectionist at a game.

Also I’m not advocating for servers to balance out solely on level. I just want to clear up this inaccurate misconception people have that “level means nothing” when it means a whole lot. You and other users on here are trying to refute this claim by pointing out uncommon examples such as users coming on dirty bomb after playing cs:go competitive and destroying at a low level. However it still points to the same case.

Practice = Experience = Learning = Skill.

Level in dirty bomb references how much EXP(ironic how the game itself calls it experience) you’e gotten in the present moment.

I personally believe that measuring a players average score per minute out of ALL the games he or she has played is the most effective way of determining their “skill” or “value” to a team. Now you are fine to disagree with this because this is just my opinion from my own experience. I’ve seen to notice that the team that has players with high scores win more often.

(Black) #52

No it’s not. I’m almost 40 now, I’m not twice as good as at level 20. I pretty much stopped progressing around that time because I reached my physical limits. Actually, I’m probably worse now - at that time I was playing DB for ~3h a day, now I only play for 30-60 minutes. I have about 33% accuracy.

Now let’s say someone truly skilled, accuracy ~50%, is also at level 40. He has the same amount of experience, but he has much higher skill. Why should we be equal when balancing?

But that’s probably exactly what the current system does. And AFAIK it’s last 25 games now, it was 10 but they changed it.[/quote]
This isn’t debatable. Either you accept it or you don’t. Practice is what makes us skilled at anything we do in life. That’s how our brains work. Right now you are explaining how some players learn faster than others in practice which is completely understandable but that doesn’t negate that experience correlates with skill.

(Amerika) #53

[quote=“BlackFro;139384”]I never disagreed with any of this. You are correct, there is differences when it comes to learning, however learning requires experience. Yes, some people will learn faster and slower than others but there is no such thing as not getting any better at the game. There is no such thing as a “performance cap”, you are always learning and getting better at the game. What you are probably referring to is a drop in the learning/improvement rate. There is no such thing as being a master or perfectionist at a game.

Also I’m not advocating for servers to balance out solely on level. I just want to clear up this inaccurate misconception people have that “level means nothing” when it means a whole lot. You and other users on here are trying to refute this claim by pointing out uncommon examples such as users coming on dirty bomb after playing cs:go competitive and destroying at a low level. However it still points to the same case.

Practice = Experience = Learning = Skill.

Level in dirty bomb references how much EXP(ironic how the game itself calls it experience) you’e gotten in the present moment.

I personally believe that measuring a players average score per minute out of ALL the games he or she has played is the most effective way of determining their “skill” or “value” to a team. Now you are fine to disagree with this because this is just my opinion from my own experience. I’ve seen to notice that the team that has players with high scores win more often.[/quote]

We agree on a lot of things but I just don’t agree that time played is anything more than a small factor in determining overall skill. I’m not discounting it, I’m simply not making it the main point that should be focused on in regards to balance. It’s not the end all be all to making a more balanced game. A person’s amount of time spent playing something isn’t the same thing as the quality of that time spent playing something, their mindset, their ability, their hardware etc. etc. It’s not a good indicator of actual skill. It’s only a factor that goes into determining it. I could go Nader suicide over and over and be level 70 right now. Or I could be level 35 and have spent a ton of time reviewing how other players play and focusing on learning to be better and playing with purpose with and against other good players. Which player is more likely to be better at the game? That’s what I am trying to get across. This isn’t a black and white situation. It’s multi-layered and there are a ton of factors that go into determining how valuable time spent doing something actually is. I could go shoot a basketball mindlessly for 12 hours a day but that doesn’t make me better than an NBA player or even a high schooler.

My point is that using a system that intelligently tries to determine a player’s overall skill, which uses as much information as possible which includes time played but isn’t solely based on time played, is better than a system that does not. I would agree that using average lifetime score per minute would be better than a time played system because that is a calculation of skill and it probably already is used extremely heavily with the current private ELO system. I do agree with how SD is using it though in only taking the last 25 games as it’s a small but fairly long slice of how good a player currently is that isn’t weighed down by how bad they might have been or how good they might have been.

I do believe the system should factor in your merc choices though and use skill ratings based on mercs and also do sorting based on the likelihood of a player playing a particular merc. A quick explanation…the sorting should be smart enough to put players who did already pick mercs on opposite teams based on how good they are with those mercs (with a minor influence on overall skill rating of all mercs too) and then sort them based on the likelihood of a person playing a medic or engineer or something else when needed. So you don’t end up with two competent teams but one of them has zero medics or engineers while the other is flooded with them. The system most definitely doesn’t do anything like this. That would, IMO, make it quite a bit more consistent.

A system that allows player to pick mercs and even tells them that the team doesn’t have enough or has too much of something (hello Overwatch) would also go a long way in fixing this. Some people would be less likely to Rambo and go all gungho with their merc choices and instead pick based on their team needs assuming they don’t have specific missions. And specific missions doesn’t help balancing either.

Hopefully this not only explains my position a bit more but also shows one example of a suggestion that could be used to potentially help balance overall accompanied by my chain of logic in regards to how it would be a positive change and why it would work.

(MarsRover) #54

What isn’t debatable? That level is not does not correlate with skill? Or something else? I don’t think you are replying to my post. Or you didn’t understand it.

What do you mean by skill? You are clearly differentiating experience and skill. Like Amerika, I am confused.

If by skill you mean pure head clicking - hell no, they don’t correlate.

If by skill you meant performance as in player score etc - then yes, but in very narrow range of levels, from 1 to around 10-15. Over 15 you know almost everything about DB there is to know AND you reach your skill ceiling. Which brings me to:

OK to the general group of proponents of balancing by levels, some simplest examples I could think of.

Skill ceilings. I won’t run as fast as Usain Bolt no matter how hard I try. I won’t be as good as a sick6 player no matter how hard I try.
Skill degradation. Stop playing FPS games for a year - will you immediately be as good when you play again? Limit your playing time from 3h/day to 0,5h/day - will you maintain your abilities?
Attitude - some people just don’t care about getting better.
Hardware changes. Try going from silky smooth 300fps with a proper setup to a 60Hz laptop screen with a shitty mini mouse.
Levels in DB can only go up, they do not account for any of those.
Performance in last X games can go up or down, accounting for every single one of those.
Explain why balancing by level is better.

Situation number 1:
I am level 40. Let’s assume I have a lvl40 “skill”.
I am level 40, I have 33% accuracy, I spent ~400h to reach it.
Another player is also level 40, he has 50% accuracy, he spent ~200h to reach it.
Would balancing by level assign equal weights? Yes.
Which player will score higher?
Which player would you rather have on your team?
Should we have equal weight in balancing? No.
Explain why balancing by level is better.

Situation number 2:
I am level 40. Let’s assume I have a lvl40 “skill”.
I stop playing DB for a year and then come back.
Will I play as good as before? No.
Does my level drop to indicate that? No.
Will my performance in 25 last games indicate that? Yes.
Explain why balancing by level is better.

This debate is so stupid I just cannot continue until someone answers those.

Also, all of this is missing the point - no matter what metric is used for assigning teams in lobby, it is just a piece of the whole. Changing it to anything else won’t magically make pubs ideally balanced.

(Black) #55

[quote=“Amerika;139405”][quote=“BlackFro;139384”]I never disagreed with any of this. You are correct, there is differences when it comes to learning, however learning requires experience. Yes, some people will learn faster and slower than others but there is no such thing as not getting any better at the game. There is no such thing as a “performance cap”, you are always learning and getting better at the game. What you are probably referring to is a drop in the learning/improvement rate. There is no such thing as being a master or perfectionist at a game.

Also I’m not advocating for servers to balance out solely on level. I just want to clear up this inaccurate misconception people have that “level means nothing” when it means a whole lot. You and other users on here are trying to refute this claim by pointing out uncommon examples such as users coming on dirty bomb after playing cs:go competitive and destroying at a low level. However it still points to the same case.

Practice = Experience = Learning = Skill.

Level in dirty bomb references how much EXP(ironic how the game itself calls it experience) you’e gotten in the present moment.

I personally believe that measuring a players average score per minute out of ALL the games he or she has played is the most effective way of determining their “skill” or “value” to a team. Now you are fine to disagree with this because this is just my opinion from my own experience. I’ve seen to notice that the team that has players with high scores win more often.[/quote]

We agree on a lot of things but I just don’t agree that time played is anything more than a small factor in determining overall skill. I’m not discounting it, I’m simply not making it the main point that should be focused on in regards to balance. It’s not the end all be all to making a more balanced game. A person’s amount of time spent playing something isn’t the same thing as the quality of that time spent playing something, their mindset, their ability, their hardware etc. etc. It’s not a good indicator of actual skill. It’s only a factor that goes into determining it. I could go Nader suicide over and over and be level 70 right now. Or I could be level 35 and have spent a ton of time reviewing how other players play and focusing on learning to be better and playing with purpose with and against other good players. Which player is more likely to be better at the game? That’s what I am trying to get across. This isn’t a black and white situation. It’s multi-layered and there are a ton of factors that go into determining how valuable time spent doing something actually is. I could go shoot a basketball mindlessly for 12 hours a day but that doesn’t make me better than an NBA player or even a high schooler.

My point is that using a system that intelligently tries to determine a player’s overall skill, which uses as much information as possible which includes time played but isn’t solely based on time played, is better than a system that does not. I would agree that using average lifetime score per minute would be better than a time played system because that is a calculation of skill and it probably already is used extremely heavily with the current private ELO system. I do agree with how SD is using it though in only taking the last 25 games as it’s a small but fairly long slice of how good a player currently is that isn’t weighed down by how bad they might have been or how good they might have been.

I do believe the system should factor in your merc choices though and use skill ratings based on mercs and also do sorting based on the likelihood of a player playing a particular merc. A quick explanation…the sorting should be smart enough to put players who did already pick mercs on opposite teams based on how good they are with those mercs (with a minor influence on overall skill rating of all mercs too) and then sort them based on the likelihood of a person playing a medic or engineer or something else when needed. So you don’t end up with two competent teams but one of them has zero medics or engineers while the other is flooded with them. The system most definitely doesn’t do anything like this. That would, IMO, make it quite a bit more consistent.

A system that allows player to pick mercs and even tells them that the team doesn’t have enough or has too much of something (hello Overwatch) would also go a long way in fixing this. Some people would be less likely to Rambo and go all gungho with their merc choices and instead pick based on their team needs assuming they don’t have specific missions. And specific missions doesn’t help balancing either.

Hopefully this not only explains my position a bit more but also shows one example of a suggestion that could be used to potentially help balance overall accompanied by my chain of logic in regards to how it would be a positive change and why it would work.[/quote]

Yes adding more factors to determine a players skill is very effective however I believe experience/skill should be the top priority because it is the foundation of all the other factors. Overwatch did a great job with notifying teams when they lack certain classes which dirty bomb needs.

You are again explaining uncommon exceptions in which balancing from level/experience can be inaccurate with the nader being level 70 just from suicide. YES that’s possible, but it’s highly unlikely to be the case. Right now you are trying to propose a system to balance teams that is flawless. Not possible, no matter what kind of balance system we implement there will always be those exceptions that make the balance bad.

It’s just like hacking, no matter how strong you implement a anti cheat, someone is always going to be able to slip through.

I’m proposing that skill/experience should be the top priority when it comes to balance. It leaves less exceptions because like 80% of the time someone who is level 70 probably got there by playing the game the “right” way and not suiciding as Nader.
It’s never going to be black and white of course, the quality of the experience does matter, however I honestly don’t see why you would not make skill/experience a priority.

Why do you think there are max and min level servers? Because SD knows level has a significant correlation with skill.

(Black) #56

What isn’t debatable? That level is not does not correlate with skill? Or something else? I don’t think you are replying to my post. Or you didn’t understand it.

What do you mean by skill? You are clearly differentiating experience and skill. Like Amerika, I am confused.

If by skill you mean pure head clicking - hell no, they don’t correlate.

If by skill you meant performance as in player score etc - then yes, but in very narrow range of levels, from 1 to around 10-15. Over 15 you know almost everything about DB there is to know AND you reach your skill ceiling. Which brings me to:

OK to the general group of proponents of balancing by levels, some simplest examples I could think of.

Skill ceilings. I won’t run as fast as Usain Bolt no matter how hard I try. I won’t be as good as a sick6 player no matter how hard I try.
Skill degradation. Stop playing FPS games for a year - will you immediately be as good when you play again? Limit your playing time from 3h/day to 0,5h/day - will you maintain your abilities?
Attitude - some people just don’t care about getting better.
Hardware changes. Try going from silky smooth 300fps with a proper setup to a 60Hz laptop screen with a shitty mini mouse.
Levels in DB can only go up, they do not account for any of those.
Performance in last X games can go up or down, accounting for every single one of those.
Explain why balancing by level is better.

Situation number 1:
I am level 40. Let’s assume I have a lvl40 “skill”.
I am level 40, I have 33% accuracy, I spent ~400h to reach it.
Another player is also level 40, he has 50% accuracy, he spent ~200h to reach it.
Would balancing by level assign equal weights? Yes.
Which player will score higher?
Which player would you rather have on your team?
Should we have equal weight in balancing? No.
Explain why balancing by level is better.

Situation number 2:
I am level 40. Let’s assume I have a lvl40 “skill”.
I stop playing DB for a year and then come back.
Will I play as good as before? No.
Does my level drop to indicate that? No.
Will my performance in 25 last games indicate that? Yes.
Explain why balancing by level is better.

This debate is so stupid I just cannot continue until someone answers those.

Also, all of this is missing the point - no matter what metric is used for assigning teams in lobby, it is just a piece of the whole. Changing it to anything else won’t magically make pubs ideally balanced. [/quote]

What? Are you kidding me right now? It sounds like you really don’t understand. Your examples are ridiculous

Skill Ceiling? Umm, nothing is impossible and there is ALWAYS room for improvement. Saying you can never be better than someone is just pure pessimistic ideology. How do you think Bolt runs that fast? Spoiler alert: He practiced. Want to be as good as someone from sick6, practice…

Skill degradation? Reference my above post, exceptions are always going to exist but these are uncommon.

Attitude and hardware changes? Situational, again a rare exception.

I never said balancing by levels solely is good, I said making it a priority as opposed to other methods is. Balancing should go like this:

  1. Average Score Per Mintue
  2. Level/Experience
  3. Other factors

It should not balance from 10 games but either all or a good number of games.

Situation 1 and 2 again explains a exception. The system is not going to be perfect so stop pointing out the ways it’s not perfect.

(MarsRover) #57

…
No. Just no. This is not how reality works.

Current metric that tracks performance in last 25 games handles this exception. Why switch to an inferior one?

Current metric that tracks performance in last 25 games handles this exception. Why switch to an inferior one?

[quote=“BlackFro;139424”]I never said balancing by levels solely is good, I said making it a priority as opposed to other methods is. Balancing should go like this:

  1. Average Score Per Mintue
    [/quote]
    So pretty much what the current system does.

[quote=“BlackFro;139424”]It should not balance from 10 games but either all or a good number of games.
[/quote]
It takes 25 games, as I’ve written multiple times.
A total score/min metric is worse because it doesn’t handle many of the examples I’ve shown. Why switch to an inferior one?

Current metric that tracks performance in last 25 games handles those. Why switch to an inferior one?

See one of my previous posts about min level servers and ranked:

[quote=“MarsRover;139018”]Those two are not about skill. They’re there to set a certain minimum level of experience and as a shield from cheaters. They’re there for people wanting to play a match where everyone knows the basics. You expect that a guy playing Sparks there to not help you up manually, you expect people to know the map, what the objectives are, etc. It’s not about skill. It is about experience, which you get only by playing the game for a certain amount of time.

Also, DB if F2P. Nothing stops a cheater from creating another account. Wins and losses matter in competitive. It is gated by level as it is the easiest way to prevent cheaters.[/quote]

(Amerika) #58

Please don’t pull out made up numbers like “80% of the time” to use in a debate to add weight to an argument. You would get torn up in any moderated intelligent debate for doing something like that. Seriously people, stop making up numbers. Argue points with counter-points and give weight to them through logic but don’t try to make up your own numbers. It’s really hard to take people seriously when that is done.

I am not proposing a flawless system. I have torn up my own suggestions countless times because there is no such thing as flawless matchmaking as there are a ton of factors that simply can’t be calculated in. I just want a system that is weighted properly and accounts for relevant and current skills as opposed to being heavily skewed or only factoring in one element like the amount of time played and punishing not only that player but the teammates who get stuck with them.

The Nader example is an example of how the system, at it’s core, is flawed when trying to heavily weigh time played with a skill rating. Not everyone plays the same which is the point I was getting across. Ignoring that fundamental flaw simply because you chalk one example up to being rare isn’t how good design works.

There is max level 5 servers because early on time played is a huge factor. My issue, which I’ve explained so many times now and am going to refuse to explain further, is that it does not scale. Early on experience/time played is extremely important. Once you learn all the basics and the maps then your skill starts relying on other factors that have been pointed out. That’s what myself and others are trying to get across. SD gave Phantom a Katana so he’s a melee merc, right? See, I can skip over all the logistics and make an assumption too.

I think I’ve said what I need to say on this subject. There is no way that I would personally ever want any system that doesn’t try to intelligently create balance in favor of any system that either solely uses or heavily weighs in time played in regards to assigning a current skill value for balancing purposes.

(Xan) #59

the best part is when i have to play 3v8 with a lvl 20,me as lvl 46 and a lvl 1 against 8 people that range from 40-70

(Black) #60

[quote=“Amerika;139450”]Please don’t pull out made up numbers like “80% of the time” to use in a debate to add weight to an argument. You would get torn up in any moderated intelligent debate for doing something like that. Seriously people, stop making up numbers. Argue points with counter-points and give weight to them through logic but don’t try to make up your own numbers. It’s really hard to take people seriously when that is done.

I am not proposing a flawless system. I have torn up my own suggestions countless times because there is no such thing as flawless matchmaking as there are a ton of factors that simply can’t be calculated in. I just want a system that is weighted properly and accounts for relevant and current skills as opposed to being heavily skewed or only factoring in one element like the amount of time played and punishing not only that player but the teammates who get stuck with them.

The Nader example is an example of how the system, at it’s core, is flawed when trying to heavily weigh time played with a skill rating. Not everyone plays the same which is the point I was getting across. Ignoring that fundamental flaw simply because you chalk one example up to being rare isn’t how good design works.

There is max level 5 servers because early on time played is a huge factor. My issue, which I’ve explained so many times now and am going to refuse to explain further, is that it does not scale. Early on experience/time played is extremely important. Once you learn all the basics and the maps then your skill starts relying on other factors that have been pointed out. That’s what myself and others are trying to get across. SD gave Phantom a Katana so he’s a melee merc, right? See, I can skip over all the logistics and make an assumption too.

I think I’ve said what I need to say on this subject. There is no way that I would personally ever want any system that doesn’t try to intelligently create balance in favor of any system that either solely uses or heavily weighs in time played in regards to assigning a current skill value for balancing purposes.[/quote]

Rotfl.
Ok no random numbers. Would it be fair to say most the time the balance system I was proposing will balance teams properly?

We are both aware the amount of factors that can’t be calculated. Since you understand the system can’t be flawless you will also agree no matter what system we do there are going to be exceptions.

In reference to the sentence I bolded: Understanding the game, and improving your skill in the game are not the same thing. Yes they can relate with another, but not the same thing.
So the games you played early in dirty bomb mostly comprised of you learning and understanding the game. After you do this, it’s all about practice and improving your muscle memory, reflexes, and yada yada you get the point.

Also I think you there is a misunderstanding. I don’t want the TIME to be heavily weighed in, I want the EXPERIENCE to be heavily weighed in. They may sound like semantic games but someone can play 800 hours and still be level 10 while other can play that amount and be level 50.

I’m not trying to skip over logistics and make brash assumptions. DB’s current balancing system is effective however you and I both know it can be improved.