74 replies to this topic

[grayson marik]

Skarlock, on 07 October 2015 - 12:06 PM, said:

I don't like the idea of solving imbalance by leaving it in the hands of matchmaker. This is because leaving OP mechs as OP and under powered mechs as under powered means that if matchmaker hits a snag and can't balance the groups, it will simply create imbalanced matches anyways, or people will end up waiting forever to start a match. When you put that responsibility solely into matchmakers hands, it can't deal with a situation where everyone in single queue just takes the most OP mech they can take, except for a few people that want to play lower tier mechs that are an odd number. 3/3/3/3 has this same problem because it's attempting to enforce a constraint, and it just waits and waits forever for lights that don't show up and too many heavy mechs per side. If you make matchmaker strictly adhere to balancing rules, people wait too long to get an actual match, and the looser the rules are, the more the match quality degrades. The less matchmaker has to do in order to balance matches, the easier it becomes for it to make matches and the more time the players can spend actually playing the game as opposed to waiting forever staring at the search wheel.

This is also bad business for PGI. Every mech is essentially a product, and if people don't buy them because they suck, they lose money. If they are over powered, then other people feel forced to spend money to stay competitive, and people don't like games that essentially devolve into pay to win scenarios. If the mechs were completely balanced, then it wouldn't matter if the enemy team has 4 or 5 assaults and the other team is mostly mediums and lights. At that point matchmaker could only balance on player rating/skill and be much faster and accurate in creating balanced matches.

Well, the point is, if there are so many ppl driving the overpovered Mechs, they will wait longer and MM can open up the BV constraint by then creating 2 high BV teams to be matched, if time gets too long.

But in general it will work faster than the current system and create more even settings BV wise.

At the end, it will atz least be better than the nerv, buff, number maginc circle that is going on now for years.

[Grayson Sortek]

grayson marik, on 08 October 2015 - 12:46 AM, said:

Well, the point is, if there are so many ppl driving the overpovered Mechs, they will wait longer and MM can open up the BV constraint by then creating 2 high BV teams to be matched, if time gets too long.

But in general it will work faster than the current system and create more even settings BV wise.

At the end, it will atz least be better than the nerv, buff, number maginc circle that is going on now for years.

Keep in mind the source of that post is tier 1. I'm guessing the posters concern comes from either losing tier placement or not having enough people in tier 1 to get quick matches.

Maybe implementing the BV system without a ridged tier system would reduce this concern... of which I don't think I remember you suggesting that we adhere to the current ridged tier system.
BTW: I actually ran into a dev last night (which is incredibly rare for me). I believe it was Garry something. I mentioned your post and said that PGI should check it out, but I received no response. I'm guessing the devs don't play the game on flagged dev accounts because they don't want to hear the concerns/suggestions/anything from the playerbase. It's not like I was complaining or calling him names, I just said that they should check your post out and I got nothing in response.

Sorry, but I don't think they are going to respond to, consider, implement, or otherwise have anything to do with your idea, which is really a shame.

[Leopardo]

so what - BV+ balancing by tier (example - same quantity of people with same tier at both sides - with any tier alowed) faster matches ever!

[grayson marik]

Grayson Sortek, on 08 October 2015 - 04:54 AM, said:

Keep in mind the source of that post is tier 1. I'm guessing the posters concern comes from either losing tier placement or not having enough people in tier 1 to get quick matches.

Maybe implementing the BV system without a ridged tier system would reduce this concern... of which I don't think I remember you suggesting that we adhere to the current ridged tier system.
BTW: I actually ran into a dev last night (which is incredibly rare for me). I believe it was Garry something. I mentioned your post and said that PGI should check it out, but I received no response. I'm guessing the devs don't play the game on flagged dev accounts because they don't want to hear the concerns/suggestions/anything from the playerbase. It's not like I was complaining or calling him names, I just said that they should check your post out and I got nothing in response.

Sorry, but I don't think they are going to respond to, consider, implement, or otherwise have anything to do with your idea, which is really a shame.

Thanks for your effort.
Trith be told, with number mambo jumbo every few month's, you might anger some players now and then. But you always generate some new income by forcing people to buy other stuff for the new META each time.

So it might well be company politics to prevent getting into touch with any workable BV system at all costs!

Once, when i had the first draft of the former Inner Sphere Wars league online, Nico Snow ( if you remember him) wanted to take a look, because PGI was worried if the league would break their non disclosure agreement for closed Beta back then. After getting his fly through the league system, Nico contacted me back with info, that I could continue because they found no break of the agreement for closed Beta and maybe PGI would even consider supporting leagues like mine...

Short time later came the next big shitstorm IGP created and since then PGI has been absoluteley silent to any suggestion coming from guys like me: programmers, league administrators or such. Except a few that are tighly connected to folks from NGNG or general the NA environment.

Long story short: If you are not part of the "inner circle" chances for suggestions to even being reviewed are close to zero.

Leopardo, on 08 October 2015 - 05:34 AM, said:

so what - BV+ balancing by tier (example - same quantity of people with same tier at both sides - with any tier alowed) faster matches ever!

Yepp, you got it quite right, condensed into just one line of text :-)

[C E Dwyer]

I've read through and I have to say I have been very resistant to B.P.V's but

P.G.I for god sake use this or a system like it

[grayson marik]

Cathy, on 11 October 2015 - 08:09 AM, said:

I've read through and I have to say I have been very resistant to B.P.V's but

P.G.I for god sake use this or a system like it

Thank you!

[Jelan]

What an excellent idea and could be implemented in a staged way. Chassis BV first then Chassis and Player BV and finally Chassis, Player and Equipment BV

Could you clarify why it would need to be based on usage of the mech to identify the BV? This is the only part of my simple brain feels might be open to abuse and may flaw the statistics as every new mech would rise high initially and then potentially fall back down.

Wouldnt some form of mech effectiveness (W/L and KDR) with a multiplier of the player BV provide a better less volatile BV?

[grayson marik]

Jelan, on 13 October 2015 - 06:18 AM, said:

What an excellent idea and could be implemented in a staged way. Chassis BV first then Chassis and Player BV and finally Chassis, Player and Equipment BV

Could you clarify why it would need to be based on usage of the mech to identify the BV? This is the only part of my simple brain feels might be open to abuse and may flaw the statistics as every new mech would rise high initially and then potentially fall back down.

Wouldnt some form of mech effectiveness (W/L and KDR) with a multiplier of the player BV provide a better less volatile BV?

Well, of course the system could work with static values, given by PGI and changed every now and then for balancing reasons, like the nerv/buff cycle, that is going on right now.

Also, you are right. A new Mech will of course rise in BV at the beginning, maybe even a bit more than its true value in combat is and drop back then, until its values settle.

But this is not a bad thing to have. A new Mech would start with a relative low BV, because it is not levelled, no modules equipped etc.
As the pilot levels his new toys, he will customize them, add skills from skill tree, equip consumables and better weapons etc.
Also many players will try to find the sweet spot of the new Mech to use and eventually the new toy will also be the new flavour of the month.
So its BV will rise until a point, where the Mech is levelled and equipped to its best configuration. Then, either the next new Mech arrives or players tend to switch back to other Mechs, because they now used a lot of time for the new Mech and are a bit tired of the chassis. Then the numbers will fall back a little and the BV will settle around an average for this chassis.

I don't think, that this curve in BV is a bad thing at all.

Also keep in mind, that the same Timberwolf with the same configuration could have different BV values on player A and B because player A is Tier 1 and B Tier 5 for example. Or player A has it levelled to master and B is currently struggling with the basic skills.

As for the BV measurement by usage; As mentioned before, usage could only be one parameter for BV rise and fall. PGI already has a much bigger set of data to chose from to include it into BV calculation like hard point distribution, weapon accuracy, pilot Tiers etc etc etc. usage was just the most simple thing to use to explain the systems mechanics.

Edited by grayson marik, 14 October 2015 - 08:32 PM.

[grayson marik]

push

[Jelan]

The problem with a rating based on usage is its too volatile and affected by whims and flows of new mech releases. I really like the overall idea but every metric needs to be objective and based on statistics.

If every mech was allocated a base BV based on weight class to ensure that a light mech never has a rating equivalent to an assault and a range it can move between based on its effectiveness.

I think i've moved away from the idea of including the player BV into this calculation as it would force highly skilled teams to roll in sub par mechs which isnt fair on them.

This would enable the MM to field balanced teams in terms of hardware with a similar BV and then it can use the PSR rating to try and match the skillsets. No need for tonnage limits or 3/3/3/3 limits, a bit like building my 200 halfling levy army in Warhammer many years ago

[Foust]

I fully support this idea.

[Col Skaza]

I fully dont understand why PGI havent done that yet.

When they were talking about introducing BV system I really hoped that it will be something like this.

One more thingy that occured to me just now: mechs could be also balanced by map (impossible for CW at least for now), lets say we have brawling atlas on Alpine: he gonna get BV lowered, since... well, its alpine.
On the other hand same Atlas lands on Viridian Bog, and there his BV rises.

I also like the idea of using more data than just mech usage.

[Leopardo]

we have to create a poll and we must show it to PGI - they must read it! after this new balance fail and after tonnage fail - we have to push it to PGI !

Edited by Leopardo, 22 October 2015 - 10:27 PM.

[grayson marik]

Leopardo, on 22 October 2015 - 10:25 PM, said:

we have to create a poll and we must show it to PGI - they must read it! after this new balance fail and after tonnage fail - we have to push it to PGI !

What you want do do? Maybe 10 or 20 ppl have read it to the end and understood, that this would benefit players, not punish them.
The rest will flame it, because they fear they cannot farm PUGS or Noobs any more. Or simply because they do not understand.

And PGI? Bah, i have sent this stuff through any channel they provide multiple times. I already fear ban because of spamming...

[Leopardo]

guys vote plz we need it!)

[BearFlag]

I'm in favor of such a system with some disagreement on details.

I'd favor separating chassis value from weapons loadout value.

Chassis BV would be dynamic as enumerated. Weapons loadout BV would be calculated dispassionately on a per weapon basis and then summed. This result would be added to the chassis BV. The exact proportion/scaling is adjustable.

Finally, at match time, a pilot multiplier would be applied to BV (possibly .6 for a newbie, 1.5 for a topgun, for example). The mech itself would have a BV independent of the pilot for the sake of comparing it to other builds.

The advantage to separating these is that calculated weapons values respond immediately to mastering, module and quirk changes. The addition of a module would show an immediate change to BV in mechlab and would be constant - not varying by the week's flavor.

Some have said that weapons BV cannot be calculated. I disagree. The "systems" exist as pure numbers. As such they can be analyzed with pure numbers and an as-built value arrived at. Numbers are easy; humans are difficult.

The dynamic BV on chassis could help encourage the use of less popular mechs. One problem I see is that it looks like in open play the system relies on altruism. "If I take a low BV mech into pug play, my teammates can bring in higher BV mechs". If players are churning they're going to bring their money-making metas and the MM will be stuck trying to match meta-meta.

Chassis BV allows you to easily create an incentive system. BV by chassis lets you compare within family, within weight, within weight class. The Cicada 3C is fun to drive, but a poor performer and so not often seen. Its chassis would have a lower BV vs. its sisters like the 2B and 3M. So a chassis bonus/penalty somewhere <15% could be applied.

Overall, this is a great system method and I like the idea. But applying it to all mechs, weapons, modules is too much dynamism. I think chassis only is best.

IMO

Edited by BearFlag, 30 October 2015 - 10:01 PM.

[grayson marik]

BearFlag, on 30 October 2015 - 09:49 PM, said:

I'm in favor of such a system with some disagreement on details.

I'd favor separating chassis value from weapons loadout value.

Chassis BV would be dynamic as enumerated. Weapons loadout BV would be calculated dispassionately on a per weapon basis and then summed. This result would be added to the chassis BV. The exact proportion/scaling is adjustable.

Finally, at match time, a pilot multiplier would be applied to BV (possibly .6 for a newbie, 1.5 for a topgun, for example). The mech itself would have a BV independent of the pilot for the sake of comparing it to other builds.

The advantage to separating these is that calculated weapons values respond immediately to mastering, module and quirk changes. The addition of a module would show an immediate change to BV in mechlab and would be constant - not varying by the week's flavor.

Some have said that weapons BV cannot be calculated. I disagree. The "systems" exist as pure numbers. As such they can be analyzed with pure numbers and an as-built value arrived at. Numbers are easy; humans are difficult.

The dynamic BV on chassis could help encourage the use of less popular mechs. One problem I see is that it looks like in open play the system relies on altruism. "If I take a low BV mech into pug play, my teammates can bring in higher BV mechs". If players are churning they're going to bring their money-making metas and the MM will be stuck trying to match meta-meta.

Chassis BV allows you to easily create an incentive system. BV by chassis lets you compare within family, within weight, within weight class. The Cicada 3C is fun to drive, but a poor performer and so not often seen. Its chassis would have a lower BV vs. its sisters like the 2B and 3M. So a chassis bonus/penalty somewhere <15% could be applied.

Overall, this is a great system method and I like the idea. But applying it to all mechs, weapons, modules is too much dynamism. I think chassis only is best.

IMO

Good point to keep it simple. Static values for equipment, dynamic for chassis, like it.

But I disagree with the altruism point you brought up.
There are 2 different ways, the MM could solve a high BV queue:

1. The wait times would be longer, if too many hich BV Mechs are waiting for a match -> just as now, when you have to wait longer, when a certain weight class is overpopulated in the queue.

or

2. The BV limit could be raised on both sides of the match, so that we have an even match again, just with an overall higher BV.

[Osric Lancaster]

grayson marik, on 15 September 2015 - 10:24 PM, said:

We could kick Ghost Heat, DHS with 1.4 and 2.0 values and set all weapons back to TT values the dynamic BV will balance all this stuff automatically.

Balanced but not fun. You really do need to replace ghost heat with a good heat system and non-instant pinpoint convergence. Plus mixing CLAN vs IS in fights gives me the angry brain, even if the sides are balanced. Otherwise it looks good and I don't even care anymore.

Leopardo, on 30 October 2015 - 03:16 AM, said:

guys vote plz we need it!)

Why, so PGI can ignore it harder? Don't kid yourself, votes on forum polls have never made them pay attention before.

Edited by Osric Lancaster, 01 November 2015 - 02:45 PM.

[grayson marik]

Osric Lancaster, on 01 November 2015 - 02:31 PM, said:

Balanced but not fun. You really do need to replace ghost heat with a good heat system and non-instant pinpoint convergence. Plus mixing CLAN vs IS in fights gives me the angry brain, even if the sides are balanced. Otherwise it looks good and I don't even care anymore.

Yeah sure, you are absolutely right! The current heat scale implementation is a joke, compared to the BT heateffects. Also convergence needs love too.
Never the less, if the TT values would be taken that way, that the damage is dealt over a 10 sec time frame, we would be much closer to the fun, that we are now.... and it would only need XML file tweaking...

Dooh, I could cry when i remember, that it would only take XML editing for a first big step towards a better game. :-(

Osric Lancaster, on 01 November 2015 - 02:31 PM, said:

Why, so PGI can ignore it harder? Don't kid yourself, votes on forum polls have never made them pay attention before.

True, PGI can for sure ignore it harder, yet some things got through their ignorance wall during the years... who knows, maybe some day this too?

[Karl Streiger]

grayson marik, on 01 November 2015 - 10:41 PM, said:

Dooh, I could cry when i remember, that it would only take XML editing for a first big step towards a better game. :-(

Thanks for ruining my day - i really don't need another reminder