Jump to content

Avoiding Endless Dev Timesink In Balancing


5 replies to this topic

#1 Slambert

    Member

  • PipPip
  • Legendary Founder
  • Legendary Founder
  • 21 posts

Posted 13 September 2015 - 02:18 AM

I'm frustrated by yet another balancepass consuming dev time and calendar time.

If we assume some level of imbalance will always exist the logical step is to introduce a mechanism to handle this (aside from annual manual balance pass).

I have read several suggestions on how to introduce such mechanisms over the years - some partly based on table top legacy.

My plea to PGI is to make sure the game has a basic level of Automatic imbalance handling before you Invest time in yt another manual balance pass.

The best suggestion I can come up with atm based on previous discussions is this:

1) Make an updated mech tiering system - either based on zero quirks (duplicate from last year) or the current quirk version.

2) change credit scoring to include a multiplier rewarding the individual player for using a low tiered mech.

3) Change match maker to only allow similar sums of "mech tier scores" from the 2 teams. Long waiting times could mitigated by allowing for instance 6 vs 12 matches in extreme cases.

4) Make an automated reevaluation (weekly or more often) of the mech tiers based on match results.

Edited by Slambert, 13 September 2015 - 02:36 AM.


#2 Nyuuu

    Member

  • PipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 93 posts

Posted 13 September 2015 - 03:34 AM

Sorry, but this just does not work.

Warthunder does something similar and it has horrible effects.

A plane is being called OP in the forum, this leads to all the bad players flocking to it since they want to do good. Since a lot of bad players are playing it and doing bad in it despite the plane in question already being OP the automated system sees "oh damn those poor guys are getting slaughtered, better buff it even more". This just continues on a spiral.

The exact opposite thing happens for underpowered planes. The bad players hear it is UP and don't even try it out, while the only ones still playing it are the very good players that want to play it since they love it. They manage to do well in it against the bad hostiles that flocked to the OP planes so the automated system once again gets to the conclusion "oh damn those things are doing pretty well, better nerf them even more".

Automated balance just does not work, its the same as developers having no idea about their own game, bringing in balance changes and statements simply going by numbers on a sheet. ("Metro" is our best Battlefield map ever, since it is by far the most played - BF3 design lead statement for those that remember the misery)

Edited by Nyuuu, 13 September 2015 - 03:36 AM.


#3 RedMercury

    Member

  • PipPipPipPipPipPip
  • Bad Company
  • 223 posts
  • LocationChina

Posted 13 September 2015 - 07:58 AM

Automatically nerf popular things and boost unpopular things. This solves the above dilemma, yes?

#4 Nyuuu

    Member

  • PipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 93 posts

Posted 13 September 2015 - 09:30 AM

View PostRedMercury, on 13 September 2015 - 07:58 AM, said:

Automatically nerf popular things and boost unpopular things. This solves the above dilemma, yes?

This would work in a fully rational world, without any emotion involved.
But with the situation we have, many people would play the Timberwolf, for it being THE 'mech people think about when they hear Battletech, until it would be so utterly nerfed below any reason to become entirely unplayable (And even then some hardcore lore fans would keep bringing it out).

On the other side of the spectrum we have 'mechs like the Kintaro (when was the actual last time you saw one in a match?) that would need to be buffed to insane levels, before people even think about taking it out due to the bias players have in their mind against it.

A good balance can only be created by human interference from people truly knowing how their game works. Or people having no idea how their game works actually listening to their top tier players...

#5 Slambert

    Member

  • PipPip
  • Legendary Founder
  • Legendary Founder
  • 21 posts

Posted 13 September 2015 - 10:18 AM

@Nyuuu
Great insights from a practical case, thanks :)

Would it work if point 4 was removed and any mech tier changes would only be updated manually?

Timberwolves and Kintaros would then be on opposite ends of the mech tiers for *always*.

The match maker would consider player tier, mech tier and queue sizes when setting up matches.

A bad player in a Timberwolf would then either meet opposing teams with similar mechs and player quality or risc being in a match with similar player quality but 6 vs 12: 6 Timberwolves vs 12 Kintaros. Note that mech types and player numbers are just examples here.

note: I strictly play IS but I really dont like that we are trying to bring IS and Clans on par. I would much rather have systems where the Timberwolf driver faced penalities outside his own mech making the choice on what to pilot a difficult one from a strategic view.

Edited by Slambert, 13 September 2015 - 10:19 AM.


#6 Appogee

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 10,966 posts
  • LocationOn planet Tukayyid, celebrating victory

Posted 13 September 2015 - 10:56 AM

View PostSlambert, on 13 September 2015 - 02:18 AM, said:

I'm frustrated by yet another balancepass consuming dev time and calendar time.


What time sink...? All they did was change values in a database.

.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users