Balance stats update from 219 to 220
Flayra
Game Director, Unknown Worlds EntertainmentSan Francisco Join Date: 2002-01-22 Member: 3Super Administrators, NS2 Developer, Subnautica Developer
Game Director, Unknown Worlds EntertainmentSan Francisco Join Date: 2002-01-22 Member: 3Super Administrators, NS2 Developer, Subnautica Developer
Here are the latest of our internal stats. The top image is Build 219 after around 6,500 games. The bottom is Build 220 after about 1,000 games. This is showing the aggregate balance across all maps.
This discussion has been closed.
Comments
That being said, at least we're past the days of ridiculous alien domination, hooray, hopefully you will address the above challenges now, then the game will in fact become amazing.
<a href="http://en.wikipedia.org/wiki/Intermediate_value_theorem" target="_blank">http://en.wikipedia.org/wiki/Intermediate_value_theorem</a>
That being said, at least we're past the days of ridiculous alien domination, hooray, hopefully you will address the above challenges now, then the game will in fact become amazing.<!--QuoteEnd--></div><!--QuoteEEnd-->
This. From the few games I've played, the sides feel balanced. However gameplay does not feel complete. Skulks small bite cone makes it a very poor class, sentries are pointless and strangely aliens feel more defensive than marines now. Suppose thats just as well, seeing how marine now seem to have the greatest offensive capability with exo and ARCs... but its not something I'm used to.
I am going to say this because I care...
The sad thing is, that if you knew enough about statistics to be using it in this way you would know that the information you just posted is, statistically, completely worthless. It is incredibly distressing to see people developing a game I care about showing such blunt ignorance of a topic they are attempting to intellectually employ to further the game.
<b>TL:DR</b> - The sample size is <b>far too small</b> and there are too many variables you <b>cannot control</b>. This shows nothing of any value.
This doesn't tell you anything about the game being balanced in terms of how likely it is for experienced players to win once they have mastered the mechanics, or how difficult it is for each player to kill each other (i.e. does skill scale equally between both teams, or is it easier for a certain team at higher levels?).
Yuuki has pointed out on numerous occasions that there are certain values of dmg etc. (or skulk bite distance) that would achieve 50/50 balance in statistics, but that 50/50 balance is essentially meaningless. Maybe if you give gorges a 1 hit kill melee then it will be 50/50 balanced, but is that a fun mechanic, is it enjoyable, is it skill based....etc?
There are so many flaws with using these stats to try and balance the game, making changes to the gameplay and then justifying them by showing that public games are closer to 50/50 is just mad.
The logic of perhaps breaking things to get to 50%, and then trying to fix them afterwards seems a bit tortuous.
ns2_rockdown huh? A little foreshadowing there? :P
Internal playtesting is a small sample of players and ability levels. Mass statistic gathering such as this is the only way to ensure balance for something as asymmetrical as NS2. Sure playtesters can have opinions, and usually ones slightly more informed than our own, but they are still of limited scope.
I agree again with Xarius, the game is balanced but at the price of fun. Giving one person some tweezers, another chop-sticks and telling them to dig a hole may be balanced - but its not fun. Currently needs some more work on the mechanics.
Yes exactly. I'm not saying that the balance is done - that will take years - nor am I saying that this is statically valid (we still see movement after 1k games), I'm just showing some data.
It's an older map from a while back. Mostly used because at the time it provided the best performance, but subpar gameplay. You can find a copy on this website somewhere.
It seems that your goal is actually to create a game that is statistically 50/50 balanced over all games rather than create something that is balanced from the perspective of gameplay. Balance should be about the skill it takes to perform a mechanic vs the skill it takes to counter it. The recent skulk nerf and then trying to justify it by posting these stats just highlights your incompetence. It's pretty clear to me why the game development for ns2 has been so rocky with you at the helm.
You have posted these types of stats before and celebrated achieving 50/50 on the twitter feed. Either you are lying in saying that you know these stats are meaningless or you are a fool for believing that they have any relevance. Sorry to be so blunt, but it just pissed me off to see a game with potential go down the drain because of incompetence and ignorance.
A thousand games seems enough, it gives you around 3% error (true winrate is between 47-53%), that's good enough. If it still change it's probably because people adapt to changes over time, that is the true winrate is changing in time.
The "I'm just showing some data" sounds suspiciously innocent ;) People don't just post "some data" on forums, I'm sure you have some thoughts about them.
Except the stats are not meaningless, they're a good rough measure of how things are shaping up, even with the skulk nerf.
Do you honestly believe the graph will jump to 80% marine wins by the time it reaches 5000 games?
Do you honestly believe the graph will jump to 80% marine wins by the time it reaches 5000 games?<!--QuoteEnd--></div><!--QuoteEEnd-->
Correlative data tells you nothing about cause. Statistically, its a nice graph, and it does give you information. But it does not tell you anything conclusive about anything at all.
It sounds to me like the Dev's know that, so that's good. I like seeing the data myself, but I also like the assurance that its not the only thing informing their balance decisions. ;-)
This graph doesn't say the game is balanced, and it doesn't say it isn't, but it does say that either team can win a pub, and that isn't something that should be underestimated.
You are interested in game balance (whether all the different gameplay mechanics come together in a balanced way). The statistics tell you that marines won 50% of their games and aliens won 50% of their games, but not why each time won those games. Not only this but they will never tell you anything about why.
The danger here in following the statistics as the main judge of balance is that you end up with vaguely equal chances of winning, with frustrating/boring/uninteresting mechanics (like the current Skulk). It would be far better to address gameplay mechanics in a logcal and psychological way first, then apply small changes to fine tune win/loss statistics after (no-one enjoys not feeling they have a fair chance of winning).
Not only that but there are so many variables you simply cannot control, that even with 50/50 ratios it might not be your game mechanics causing it, but the community instead. For example it has been true frequently that, newer players go marine, meaning the better players are more often on the alien team. By this logic aliens should have a higher win/loss rate and the current 50/50 rate would suggest you have unbalanced the game mechanics. This phenomenon will not last forever, developing a game around it is incredibly short sighted.
The objective should be to have the game be balanced while at the same time having it as fun as before to play. Try making it easier for the losing side of the win/loss ratio to play rather than making it harder for the winning side of the ratio to play. That way everyone is happy.
can aliens win without fades ?
im sure the aliens wouldn't even notice if the lerk was removed from the game
and dont get me started on bilebomb, its like the counter to EVERYTHING
doesn't seem very balanced to me. I could care less about 50:50 when the choices we are given are no brainer and most of the features seem like fluff
what the stats that are just now and listen I agree i don't like the change to skulks bite but , Beta is Beta !
and wilson ur a D*** learn to play nice or I'll sit ur ass on the naughty corner lol
Yes, shotgun rush.
<!--quoteo(post=1980961:date=Sep 20 2012, 11:17 PM:name=piratedave)--><div class='quotetop'>QUOTE (piratedave @ Sep 20 2012, 11:17 PM) <a href="index.php?act=findpost&pid=1980961"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->can aliens win without fades ?<!--QuoteEnd--></div><!--QuoteEEnd-->
Yes, skulk rush.
Yes, skulk rush.<!--QuoteEnd--></div><!--QuoteEEnd-->
ty for proving my point :)