HEY GUYS LET'S PRIORITIZE KILL TO DEATH RATIO DATA OVER DPS/HPS/MITIGATION+DPS BECAUSE TAPPING ENEMIES WITH YOUR BASIC ATTACK IS REALLY HARD AND IS A TRUE MEASUREMENT OF A CLASS'S WORTH

I don't even.

except that 9/10 times game metrics are not representative of what is actually going on.

 

metrics usually take into account the whole population. which results in comparing samples with different sizes, which you generally shouldnt do.

 

it also means that the metrics are heavily diluted the greater the population. bioware should be using the top 5% of each class and adjusting sample sizes to make sure all ACs are compared equally.

 

you cannot objectively compare 5,000 Maras to 500 Mercs, and balancing them game on the performance of bad players is a terrible idea

great examples of how unreliable player perceptions are and why player feedback has to be measured agaisnt real objective data.

where did you come up with your 9/10 times thing? just made it up because it sounds good? you also are assuming that they only look at broad general data and not specific data like how the top 5% performs. again, you are basing that on what?

great examples of how unreliable player perceptions are and why player feedback has to be measured agaisnt real objective data.

 

where did you come up with your 9/10 times thing? just made it up because it sounds good? you also are assuming that they only look at broad general data and not specific data like how the top 5% performs. again, you are basing that on what?

You keep asking for proof.... well here's the problem with the burden of proof. You have to prove it as well. What proof do you have that metrics ARE an intelligent way to balance a game.

Because I'll tell you what proof I have against that argument of metrics > experience and player feedback.

Answer: Star Wars: The Old Republic.

great examples of how unreliable player perceptions are and why player feedback has to be measured agaisnt real objective data.

 

where did you come up with your 9/10 times thing? just made it up because it sounds good? you also are assuming that they only look at broad general data and not specific data like how the top 5% performs. again, you are basing that on what?

its called critical thinking.

KDR is a terrible metric. the "operatives stunlocking players into un-subbing" metric is terrible. i could go on, but cba.

based on past comments, it is pretty clear that bioware looks at overall population metrics.

and while i would agree that math is a reliable tool for calculating damage output and the like, it is less reliable for calculating what goes on in PvP.

math will predict performance in PvE almost exactly. math in PvP is unreliable because you cannot reliably represent the number one variable in PvP: human ability.

which is why metrics are not the best way to balance actual class performance. looking at average DPS outputs only tell you part of the story in PvP

You keep asking for proof.... well here's the problem with the burden of proof. You have to prove it as well. What proof do you have that metrics ARE an intelligent way to balance a game.

 

Because I'll tell you what proof I have against that argument of metrics > experience and player feedback.

 

Answer: Star Wars: The Old Republic.

should i have to prove the value of scientific method as well? im not asking anyone for "proof" of anything. i am just telling you that player feedback is always subjective and has to be measured against real objective data.

by the way, your personal expreience and perceptions isnt actually "proof". just your opinion. i have my own opinions on game mechanics and on the dif classes as well, but my opinions are not "proof" either. just my own perceptions.

would be foolish for BW or any dev to only look at all those subjective perceptions and not measure them against real objective data.

My understanding from reading both articles is that Jonathan Crow is a Gameplay Analyst and not on the PvP Dev team. His job encompasses pulling data and stats from every part of the game. It is in no way centered around PvP.

So it doesn't surprise me that he compiles stats on K/D ratios, largely a peripheral metric, mainly because he works...well...on the periphery of PvP development.

That said, I don't think K/D ratio is that absurd a stat to monitor for imbalances. Keeping in mind that WZs are a team-based environment, consistent aberrations in K/D ratios would certainly alert a critical observer to something being out of whack.

You understand that just the fact that they used that as an example goes to show how clueless they can be about PvP and the metrics... right?

No it doesn't. Crow isn't a PvP Dev.

No it doesn't. Crow isn't a PvP Dev.

They've been quoted many times about using metrics as a reference for PvP as well... and in the original quote he was referring to PvP.

should i have to prove the value of scientific method as well? im not asking anyone for "proof" of anything. i am just telling you that player feedback is always subjective and has to be measured against real objective data.

 

by the way, your personal expreience and perceptions isnt actually "proof". just your opinion. i have my own opinions on game mechanics and on the dif classes as well, but my opinions are not "proof" either. just my own perceptions.

 

would be foolish for BW or any dev to only look at all those subjective perceptions and not measure them against real objective data.

Wow Alanis would be upset, two actual examples of irony in the last few posts. MR's post, and this one.

You realize the "objective data" they are looking at IS player feedback in the form of the results players achieve in WZs. Basing their decisions on player performance, is no more objective than basing them on player opinion, as one is typically dictated by the other. The only way to make the data viable is to establish a baseline of player competence. This can't be done by taking holistic averages, as class balance directly affects player population, as has already been pointed out. It is a reasonable assumption (but still an assumption) that anyone bothering to play a Merc DPS anymore, is very good at it. While it is a personal, yet also obvious, observation that there are hundreds of terrible Marauders running around burning up the WZs to the tune of 237 DPS. Without perhaps having a computer simulate them, the most objective data they can gather regarding player performance is the top 50, or 10%, or whatever of every class. But the sample sizes should be equal, and should NOT be gathered from the average of a skewed population.

I'm gonna make a commando and auto attack players and then run away / line of sight.

Incoming commando nerf.

should i have to prove the value of scientific method as well? im not asking anyone for "proof" of anything. i am just telling you that player feedback is always subjective and has to be measured against real objective data.

 

by the way, your personal expreience and perceptions isnt actually "proof". just your opinion. i have my own opinions on game mechanics and on the dif classes as well, but my opinions are not "proof" either. just my own perceptions.

 

would be foolish for BW or any dev to only look at all those subjective perceptions and not measure them against real objective data.

I'm not saying that metrics should be completely ignored. But in the quote, it appeared to be the focus and their way to balance the game and to anyone who is paying attention... it's not the first time they've pointed out that they rely heavily on metrics...

We seem to be reaching some kind of middle ground here... while in the beginning you seemed to be perfectly fine with them basing it off of metrics alone (or at least how it came off) and maybe I came across unclear, but I'm just saying they seem to be relying on metrics far too heavily.

It obviously is NOT working.

I'm not saying that metrics should be completely ignored. But in the quote, it appeared to be the focus and their way to balance the game and to anyone who is paying attention... it's not the first time they've pointed out that they rely heavily on metrics...

 

We seem to be reaching some kind of middle ground here... while in the beginning you seemed to be perfectly fine with them basing it off of metrics alone (or at least how it came off) and maybe I came across unclear, but I'm just saying they seem to be relying on metrics far too heavily.

 

It obviously is NOT working.

Correct.

For example, the metrics behind commando DPS are phenomenal. They do not account for, however, a warrior jumping in and interrupting grav round, hard interrupting grav round right after, a sorc walking by and interrupting the next unload and just completely crippling the players ability to be effective.

You need to account for class interaction.

[quote=DarthOvertone;5521853

That said, I don't think K/D ratio is that absurd a stat to monitor for imbalances. Keeping in mind that WZs are a team-based environment, consistent aberrations in K/D ratios would certainly alert a critical observer to something being out of whack.

KDR is probably the worst possible stat to use a evidence of performance in PvP.

a Healer gets credit for kills made by those that they have healed; ive healed wzs where i end up with 50 kills and 0 damage done.

Damage Per Kill (a metric i invented ) is a better representative of a players performance: http://www.swtor.com/community/showthread.php?p=3835275&highlight=DPK#post3835275

the point is that metrics, and statistics in general, can say whatever you want them to say. Mark Twain has a great quote "There are 3 kinds of lies. Lies, Damned Lies, and Statistics"

They've been quoted many times about using metrics as a reference for PvP as well... and in the original quote he was referring to PvP.

Ugly...the dude is a Gameplay Analyst. He also collects stats on World Events, Ops, FPs, Fleet activities, etc. He compiles metrics from all over the place. His Meet the Developer interview is actually pretty interesting if you haven't already read it: http://www.swtor.com/blog/meet-developers-jonathan-crow

Check out the gameplay chart he provided. Maybe I'm the only one, but I found it interesting that WZ participation was #2 only to Planetary activities up to 1.4 in the time period provided. Of course, Ops interest increased after 1.4 with the new NM mode, but WZ play stayed pretty consistent.

Every time I see new stats about this game, I'm always surprised by the strong performance of PvP in the metrics. It's not the sideshow that carebears would have you believe.

Edit (Didn't want to double post):

KDR is probably the worst possible stat to use a evidence of performance in PvP.

Nah. Among other variables, an anomaly in K/D ratio would be a fine red flag indicating something was out of whack.

Granted, it certainly shouldn't be the foremost and only stat for PvP, and it isn't. The guy was asked for a stat by the interviewer so he provided one. He's not even a PvP Dev. Dude just compiles statistical data and passes it along to the teams that need it. You guys are overreacting.

Ugly...the dude is a Gameplay Analyst. He also collects stats on World Events, Ops, FPs, Fleet activities, etc. He compiles metrics from all over the place. His Meet the Developer interview is actually pretty interesting if you haven't already read it: http://www.swtor.com/blog/meet-developers-jonathan-crow

 

Check out the gameplay chart he provided. Maybe I'm the only one, but I found it interesting that WZ participation was #2 only to Planetary activities up to 1.4 in the time period provided. Of course, Ops interest increased after 1.4 with the new NM mode, but WZ play stayed pretty consistent.

 

Every time I see new stats about this game, I'm always surprised by the strong performance of PvP in the metrics. It's not the sideshow that carebears would have you believe.

I did skim through it and was surprised by that too...

Personally I didn't expect PvP to be so strong in comparison... shall we start an uprising?

Nah. Among other variables, an anomaly in K/D ratio would be a fine red flag indicating something was out of whack.

 

Granted, it certainly shouldn't be the foremost and only stat for PvP, and it isn't. The guy was asked for a stat by the interviewer so he provided one. He's not even a PvP Dev. Dude just compiles statistical data and passes it along to the teams that need it. You guys are overreacting.

kills =/= kills in this game. kill totals in warzones are Kills + Assists. if you hover over it you will see "kills", "deathblows", "and solo kills". because this game awards a kill to anyone that basically looks at the enemy that dies, KDR is in no way a representation of how a player performed. if the game awarded kills to only the person who did the most damage to the target, and an assist to everyone else, then KDR would mean something

and im pretty sure that they dont use KDR to balance anything (for the reason i just laid out), i think youre right that the guy just spewed out the first metric that came to mind.

I did skim through it and was surprised by that too...

 

Personally I didn't expect PvP to be so strong in comparison... shall we start an uprising?

the devs have acknowledged PvP as a majority playstyle for a while. back in Feb or March a dev said that >50% of players participated in wzs, while ~35% participated in endgame Ops. why they have not made more effort to improve PvP, which clearly a large part of their playerbase enjoys, is beyond me

You shouldn't be using metrics on normal WZ with normal players- that's what led to the worst nerfs in the history of mmos in 1.2. Most of those nerfs were based off complaints back when the game had only one level 10-50 bracket, and almost nobody had full champion, much less BM.

They shouldn't have made such large balancing changes on such bad info.

Frankly- they came out with a preseason. They should have waited for stats from the pre-season RWZ and then made necessary changes. Do you think sorcs, mercs and ops would have gotten such large nerfs early on to dps when current stats show them as almost never being taken for dps slots in RWZ? I'd hope not.

RWZ is where skilled players are- pvp should be balanced around skilled players, not the whines of people who weren't capable of interrupting tracer missile spam or who honestly thought sorcs needed a nerf to their STEALTH ability.

Pre season doesn't need to be balanced before hand, it's preseason after all it's like a big long testing ground so that when you entire season 1 of RWZ you have balance. Balancing before hand was idiotic and it cost them big on the pvp population of the classes hit hardest by the nerfs.

It would be like planning out months of test firing a space shuttle- but sending people into space before you did that- it's backwards thinking that makes me think they don't actually understand what preseason is for, and put it in there because they saw other successful pvp games do so.

Furthermore- they refuse to show any of their statistics off- which only means that either metrics don't exist at all or they show something that doesn't match up with what they're saying. Of course- because of poor balancing there may not be enough RWZ teams anyway for there to be good metrics to use.

lol - you'd be surprised how many average guilds will bring in a commando - even a grav spammer. lol

 

speaking of which, is it just me, or do the grav spammers in reg wz's dwarf the number of assault specs? I mean...c'mmon.

Because assault isn't this god of mobility that people seem to think.

Gunnery full auto, grav static demo round and HIB mobile.

Assault full auto, charged rounds static HIB and AP mobile. If your unlucky with procs you can spend the whole Rail Shot, AP cooldown's static just trying to get it to frickin trigger with crap damage of those skills that offer no lowering of armour etc

They are both as mobile or static (choose your terminology) as each other and both put out very similar DPS on rare occasions if the Assault spec has had the proc gods with them sure they can do a bit better but Assault will never be remotely near what a VG can put out.

You shouldn't be using metrics on normal WZ with normal players- that's what led to the worst nerfs in the history of mmos in 1.2. Most of those nerfs were based off complaints back when the game had only one level 10-50 bracket, and almost nobody had full champion, much less BM.

 

They shouldn't have made such large balancing changes on such bad info.

 

Frankly- they came out with a preseason. They should have waited for stats from the pre-season RWZ and then made necessary changes. Do you think sorcs, mercs and ops would have gotten such large nerfs early on to dps when current stats show them as almost never being taken for dps slots in RWZ? I'd hope not.

 

RWZ is where skilled players are- pvp should be balanced around skilled players, not the whines of people who weren't capable of interrupting tracer missile spam or who honestly thought sorcs needed a nerf to their STEALTH ability.

 

Pre season doesn't need to be balanced before hand, it's preseason after all it's like a big long testing ground so that when you entire season 1 of RWZ you have balance. Balancing before hand was idiotic and it cost them big on the pvp population of the classes hit hardest by the nerfs.

 

 

It would be like planning out months of test firing a space shuttle- but sending people into space before you did that- it's backwards thinking that makes me think they don't actually understand what preseason is for, and put it in there because they saw other successful pvp games do so.

 

Furthermore- they refuse to show any of their statistics off- which only means that either metrics don't exist at all or they show something that doesn't match up with what they're saying. Of course- because of poor balancing there may not be enough RWZ teams anyway for there to be good metrics to use.

have a cookie!

Metric I'd most like to see: % of each AC on WINNING ranked warzone teams, further broken down by the talent tree with the most points in it.

and a cookie for you too!

You shouldn't be using metrics on normal WZ with normal players- that's what led to the worst nerfs in the history of mmos in 1.2. Most of those nerfs were based off complaints back when the game had only one level 10-50 bracket, and almost nobody had full champion, much less BM.

 

They shouldn't have made such large balancing changes on such bad info.

 

Frankly- they came out with a preseason. They should have waited for stats from the pre-season RWZ and then made necessary changes. Do you think sorcs, mercs and ops would have gotten such large nerfs early on to dps when current stats show them as almost never being taken for dps slots in RWZ? I'd hope not.

 

RWZ is where skilled players are- pvp should be balanced around skilled players, not the whines of people who weren't capable of interrupting tracer missile spam or who honestly thought sorcs needed a nerf to their STEALTH ability.

 

Pre season doesn't need to be balanced before hand, it's preseason after all it's like a big long testing ground so that when you entire season 1 of RWZ you have balance. Balancing before hand was idiotic and it cost them big on the pvp population of the classes hit hardest by the nerfs.

 

 

It would be like planning out months of test firing a space shuttle- but sending people into space before you did that- it's backwards thinking that makes me think they don't actually understand what preseason is for, and put it in there because they saw other successful pvp games do so.

 

Furthermore- they refuse to show any of their statistics off- which only means that either metrics don't exist at all or they show something that doesn't match up with what they're saying. Of course- because of poor balancing there may not be enough RWZ teams anyway for there to be good metrics to use.

I think by far the most surprising thing about this comment...

is it made sense to me and I agree with most of it.

Metric I'd most like to see: % of each AC on WINNING ranked warzone teams, further broken down by the talent tree with the most points in it.

DPS Operative: < 1%

DPS Mercenary: < 1%

DPS Sorcerer: 3%

Healing Operative: > 99%

Carnage Marauder: > 99%

Pyrotech Powertech: 92%

Shieldtech Powertech: 4%

You get the idea...

Metric I'd most like to see: % of each AC on WINNING ranked warzone teams, further broken down by the talent tree with the most points in it.

Personally I'd like to see them all.

Yes- I know, damage/healing/KDR, etc... you can look at any stat and say 'well there's this reason for it'. For example, it's easier to get more damage against teams with more healing and where you're alive the whole time, it's easier to get more 'kills' if you have aoe and dots even if you might not actually be killing as much.

Even solo kills only would prove the class is perhaps a better 1v1 class.

But- if we do see a trend of certain ACs being never wanted or always wanted- that would be a bit helpful.

I think overall all the data- including things like metrics on damage potential in 5 or 10 second bursts, amount of time spent CCed, DPS and time spent able to do DPS or healing, etc... all together though could tell us some things.

And seeing as this is data on how we are performing- it shouldn't be kept top secret from us.

This is me thinking here- if I had a game that was honestly balanced, and I honestly felt like my metrics supported my balance- I'd be happy to show it because it would quell a large chunk of the unrest. It's hard to support crying nerf or buff for classes when you have available stats that disprove the people asking for those things. However- when there's no stats you're stuck with everyone spouting opinions which are supported at best by skilled play, or at worst by 'last WZ I got pwnd by this class NERF!!!'.

It divides even the opinions of the best players whereas if those stats were available- players who know a bit about pvp could agree on more even at opposing sides of the coin. We don't really know what's going on- even when people put up their take on 100 games played at RWZ, these are the stats (at great expense to their own time to compile that data)- it may just be that server only has two RWZ teams so of course if neither has a merc on them you'll never see a merc, etc....

If we instead have data on tens of thousands of RWZ matches above 2000 rating (let's assume there has been that many...) and still see 'oh, there's 1% of players being mercs' that gives a much better idea of the balancing- true, being an overplayed or underplayed class doesn't necessarily mean the class sucks/owns, but in rated if you see a great disparity chances are the RWZ leaders aren't taking those classes for a specific reason.

Perfect balance is impossible- but I don't think we're even looking at something close to that- I think we're looking at some trees making up as much as 15% of all players in RWZ, while others are making up well under 1%. There's 8 ACs, and 3 trees per- that's 24 trees- we should see around 4% rep per tree- I don't need metrics to know it's not even in that ballpark.

But- ultimately, we can't even know because BW is hiding it from us.

You shouldn't be using metrics on normal WZ with normal players- that's what led to the worst nerfs in the history of mmos in 1.2. Most of those nerfs were based off complaints back when the game had only one level 10-50 bracket, and almost nobody had full champion, much less BM.

 

They shouldn't have made such large balancing changes on such bad info.

 

Frankly- they came out with a preseason. They should have waited for stats from the pre-season RWZ and then made necessary changes. Do you think sorcs, mercs and ops would have gotten such large nerfs early on to dps when current stats show them as almost never being taken for dps slots in RWZ? I'd hope not.

 

RWZ is where skilled players are- pvp should be balanced around skilled players, not the whines of people who weren't capable of interrupting tracer missile spam or who honestly thought sorcs needed a nerf to their STEALTH ability.

 

Pre season doesn't need to be balanced before hand, it's preseason after all it's like a big long testing ground so that when you entire season 1 of RWZ you have balance. Balancing before hand was idiotic and it cost them big on the pvp population of the classes hit hardest by the nerfs.

 

 

It would be like planning out months of test firing a space shuttle- but sending people into space before you did that- it's backwards thinking that makes me think they don't actually understand what preseason is for, and put it in there because they saw other successful pvp games do so.

 

Furthermore- they refuse to show any of their statistics off- which only means that either metrics don't exist at all or they show something that doesn't match up with what they're saying. Of course- because of poor balancing there may not be enough RWZ teams anyway for there to be good metrics to use.

One of the BEST posts I've ever read on this forum, or any other!