69 replies to this topic

[Roland]

Varent, on 02 March 2014 - 02:14 PM, said:

I like it but it also needs to take into account how those weapons interact with the mech weight, and lack or addition to jump jets. Also will need to play out with how they interact with the other weapons as well. For example a mech with jump jets and ppc/ac is far far superior to a mech without jump jets and the same load. In addition to that a mech with just one ppc and then all srm would have a wonky vaue as well since It wouldn't necessarily bring all those weapons to bear equally. That said in addition having them code this properly would be key.

Well, in practice, you could likely get away with a system that simply assigned a value to each chassis variant, and ignored the actual weapons and load out.

The effectiveness of individual weapons and equipment would be intrinsically included in the usage stats of the mech variants.

Honestly, it may even be more advantageous to ignore the weapons and only care about the variants, as this would encourage folks to explore other variants which can carry a lot of the same weapons as the optimal variants, but which can't carry quite the optimal load out. These would then also end up increasing in battle value cost, pushing players to explore other less similar variants.

Thus, the effective battle value of equipment would eventually show up in the variant costs, negating a need to actually bother tracking individual pieces of equipment. Just track the variants.

And we know that pgi already tracks the usage of each variant, so this data is already there.

So all that is necessary is to generate a value for each variant based on the usage numbers compared to total mechs dropped over a period of time.

Then, you implement a simple battle value limit for teams, similar to what pgi had planned for tonnage limits.

This system, however, addresses the potential flaw that was identified in simple tonnage limits, in that you could still make optimized lances by pairing the best assault mechs with the best light mechs.

In my proposed system, you wouldn't be able to do that, because both the assault and light mechs would end up having high battle values.

With my system, of you wanted to drop a high quality mech like a highlander, or Victor, or Jenner, you would need to pair it with a mech that is generally considered sub optimal. Not necessarily light, but merely not one of the "top tier" mechs.

What's more, you would be incentivized to make the most of those suboptimal mechs, coming up with new configurations which squeezed the most efficiency out of the least used mechs.

The result would be that the most competitive teams, in their desire to drop the best high value mechs, would also be dropping interesting and novel builds on mechs that the market had deemed low value. Thus, even in the most competitive environments, you would see a dynamic and varied selection of mechs. As new variants were discovered that exploited the current state of the market, and the current metagame, the market would automatically adjust prices to compensate, and the game begins anew.

The system is easy to implement, self adjusts, and results in a constantly shifting balance environment within which we can play. It solves so many problems for pgi, by automatically associating a cost with high efficiency in game.

Instead of trying to fight against the player base's natural tendency to optimize, you instead leverage that very tendency to balance the game for you.

[Willard Phule]

Varent, on 02 March 2014 - 02:14 PM, said:

I like it but it also needs to take into account how those weapons interact with the mech weight, and lack or addition to jump jets. Also will need to play out with how they interact with the other weapons as well. For example a mech with jump jets and ppc/ac is far far superior to a mech without jump jets and the same load. In addition to that a mech with just one ppc and then all srm would have a wonky vaue as well since It wouldn't necessarily bring all those weapons to bear equally. That said in addition having them code this properly would be key.

In regard to the highlighted text, I say "no it doesn't." Mobility, namely Jump Jets, is already taken into consideration with regard to standard BV calculations.

As for the current "flavor of the month" meta with the jump snipers...sure, they're nasty. I hate 'em as much as you do but....it's all take into consideration with BV. Point in fact, how is an Assault mech with the ability to just poptart in any way different than a mech that can jump over his head, while spinning during the jump to end up behind him and firing into his rear armor any different?

Let me throw this at you.....some clown with 2 ERPPCs and an AC5 with JJs has X battlevalue. How useful is he, really? He's REAL good against morons that run around the open plains without taking advantage of cover. How good is he against that idiot that just stands still, zooms in and target fixates until he dies? I see it every day.

How good is he against someone that knows the value of cover and concealment?

How good is he up close and personal? He's got that AC...but he's gonna touch off those ERPPCs to the point that he shuts down, eventually.

The jumpsniper meta works real well against noobs....and, unfortunately, that's a problem with the Matchmaker right now. You get such a HUGE range of talent in every match, a Jumpsniper is guaranteed to get a few kills. Using a BV system sort of negates that...since you're adding the Elo to the BV of the mech in question. An Atlas-whatever with a high Elo has a different BV than an Atlas-whatever with a low Elo.

Since this whole Medium vs The World thing...I can't tell you how many Highlanders I've taken out with a Streakhawk (4 SSRM2, 1 LL). And it doesn't take a whole lot of effort. You get in close and the guy can't spin with you...or keep up with you if you continuously jump over him, spin in midair and land facing his rear armor...use the LL on his legs, and just chainfire the SSRMs...he's done. But, I guess, that's a skill thing...it's what separates us from the noobs.

It all works out in the wash, man. BV is based on what you have, not how you use it. Let's take another look at that Highlander. If he had balanced out his load with some short-range stuff, I wouldn't be able to pull it off. Fair enough. Go for the meta...eventually, some idiot like me is going to figure out your weakness. And believe me, I'm an idiot. If I can do it, someone else can do it better.

Edited by Willard Phule, 02 March 2014 - 03:01 PM.

[PraetorGix]

BlackIronTarkus, on 02 March 2014 - 09:09 AM, said:

Seriously, it seems they try every other mechanics other than battlevalue. Why? The currently proposed mechanic of 3/3/3/3 will just end up as 3 jenners/3 shadowhawks/3 CTF-3D/2 highlander and 1 atlas DDC. Even tonnage limit is a better mechanic, but as they explained in their post it involve some problems.

I'm not trying to be offensive, but I'm honestly growing more and more tired of people trying to grab a game supposed to be for fun and making it "competitive". Maybe it's because I just got into WH40K and it's the same moronic trend there. This game is NOT balanced yet, and that doesn't take away from its fun factor.

Iacov, on 02 March 2014 - 09:26 AM, said:

some actually choose chassis by looks or feel
some perceive fun not only defined by efficiency
personally, i think it is very unlikely that there will only be "the lions of their weight class" in the new matches

otherwise i couldn't explain all those locust and quickdraws - that actually have their worth

Exactly this. As a Mechwarrior/BT fan, I cannot understand why another fan (and I'm assuming this because it seems most people in the forums are fans) would let the "competitive" issues get in the way of their fun.
Funny, I don't play any of the OP's suggested mechs, and guess what? I'm still having lots of fun and I manage to kill a good amount of these mega hyper ultra efficient uber competitive pro chassis.

Edited by Cmdr Hurrell, 02 March 2014 - 03:05 PM.

[Willard Phule]

Something to point out for CW:

BV would definitely change things. NO the matches don't have to be...and shouldn't be...even.

Imagine if you had a unit defending Hesperus. There's NO way that a defending unit should EVER have even forces with an attacking unit. That's just stupid. Even in BT canon, when Wolf's Dragoons (the closest anyone ever came to taking the facility) never had close to even odds.

So, in CW, if the MM is figuring out who should be defending Hesperus....one would think it could easily give a higher BV limitation to the defenders than the attackers.

It makes everything easier.

[Dirkdaring]

Mawai, on 02 March 2014 - 10:13 AM, said:

It doesn't need to be battletech to realize that trying to balance a match by assigning a Combat Value, Point Value, or whatever you want to call it to each mech would provide a potentially easier way to balance a match.

Ideally .. (and I mean ideally)

A mech could have a combat value derived from a combination of
- tonnage
- weapons
- engine
- armor
- modules
- jump jets
- other equipment

which would give the Base Mech Combat Value. This would then be modified by pilot factors
- mech efficiencies
- ELO
- total character earned experience

and then further modified by a group factor
- group size modifier

This process give ONE value for each mech in the queue which could then be used to form a match.

Advantages:
- the mix of mechs in the match should be more or less balanced based on all these factors combined
- the matches do not get stale with the same mixture of mechs ... personally, I think I will find matches where every team, all the time has 3 light, 3 medium, 3 heavy and 3 assault pretty boring.
- many of the weight classes only have one choice .. so if the matchmaker ends up exactly matching tonnage you will know that if you have an atlas ... so does the other team ... also boring.
- It acknowledges the fact that a Jenner piloted by an expert can be much more effective than an Atlas piloted by someone in their first match.
- it makes the matchmaker code much easier since there is only one quantity being used to form the match. It isn't combining ELO, mech class and trying to satisfy multiple constraints ... it only has to worry about the Effective Combat value

Finally, this system has the issue that you could mix great players with poor ones ... to avoid this the incoming queue could always be tiered as they plan in the new launch module so that each ELO group will be playing with folks in the same range.

Ok - you go right ahead and try to code that. Oh - and remember it has to calculate it all before a match begins will millions of players around the world playing. Don't forget the game isn't funded enough for more servers, so be sure to put in some text for players to read while they wait 15 minutes in queue before the match begins.

[Willard Phule]

Dirkdaring, on 02 March 2014 - 03:11 PM, said:

Ok - you go right ahead and try to code that. Oh - and remember it has to calculate it all before a match begins will millions of players around the world playing. Don't forget the game isn't funded enough for more servers, so be sure to put in some text for players to read while they wait 15 minutes in queue before the match begins.

No, it doesn't.

It calculates BV in the mechlab. Is it difficult? Maybe for you and me...we have to use a calculator, but we can get it done.

EVERY...EVERY...battlmech generator in both the free and communist worlds provide BV FOR FRE now. It can't be that complicated....then again, scratch that..it might be too complicated for PGI....regardless.

You have TWO...TWO values that matter. BV, which is generated when you splap that joke of a mech you put together. ONE that is simply the "machine." All the mechanical parts you thought were useful. Cool. Your BV is "XXXXXX."

The other is your Elo. I don't even pretend to understand how MW:O comes up with that number. Let's just say it exists.

When you hit "Launch" and click two dozen buttons, all with annoying beeps, to get into a game....the Matchmaker adds your BV and you Elo together to come up with...oh, let's call it the "Combat Value (CV)."

It finds people in the range of CV it's told to find.

Will you end up with noobs? Maybe. But they won't be noobs in completely useless mechs. They'll at least be able to help out a little before they end up as charred moron corpses in Terra Therma. Even complete ******* in the Stalker Trial will get a volley or two off before they die. At least they aren't the idiots in Trial Spiders, running around in 3PV, that drag the match out for 8 minutes because they don't have a clue that Skirmish is Skirmish.

[Khobai]

Quote

Well, in practice, you could likely get away with a system that simply assigned a value to each chassis variant, and ignored the actual weapons and load out.

You could. The problem is that would only work if weapons were perfectly balanced... and were a long ways from that. were almost 2 years out of beta and we still dont have balanced weapons :\

[Varent]

Roland, on 02 March 2014 - 02:43 PM, said:

Well, in practice, you could likely get away with a system that simply assigned a value to each chassis variant, and ignored the actual weapons and load out. The effectiveness of individual weapons and equipment would be intrinsically included in the usage stats of the mech variants. Honestly, it may even be more advantageous to ignore the weapons and only care about the variants, as this would encourage folks to explore other variants which can carry a lot of the same weapons as the optimal variants, but which can't carry quite the optimal load out. These would then also end up increasing in battle value cost, pushing players to explore other less similar variants. Thus, the effective battle value of equipment would eventually show up in the variant costs, negating a need to actually bother tracking individual pieces of equipment. Just track the variants. And we know that pgi already tracks the usage of each variant, so this data is already there. So all that is necessary is to generate a value for each variant based on the usage numbers compared to total mechs dropped over a period of time. Then, you implement a simple battle value limit for teams, similar to what pgi had planned for tonnage limits. This system, however, addresses the potential flaw that was identified in simple tonnage limits, in that you could still make optimized lances by pairing the best assault mechs with the best light mechs. In my proposed system, you wouldn't be able to do that, because both the assault and light mechs would end up having high battle values. With my system, of you wanted to drop a high quality mech like a highlander, or Victor, or Jenner, you would need to pair it with a mech that is generally considered sub optimal. Not necessarily light, but merely not one of the "top tier" mechs. What's more, you would be incentivized to make the most of those suboptimal mechs, coming up with new configurations which squeezed the most efficiency out of the least used mechs. The result would be that the most competitive teams, in their desire to drop the best high value mechs, would also be dropping interesting and novel builds on mechs that the market had deemed low value. Thus, even in the most competitive environments, you would see a dynamic and varied selection of mechs. As new variants were discovered that exploited the current state of the market, and the current metagame, the market would automatically adjust prices to compensate, and the game begins anew. The system is easy to implement, self adjusts, and results in a constantly shifting balance environment within which we can play. It solves so many problems for pgi, by automatically associating a cost with high efficiency in game. Instead of trying to fight against the player base's natural tendency to optimize, you instead leverage that very tendency to balance the game for you.

I dunno honestly I think the best way to do it would really be to take into account the factors of chasis, weapons, Jump Jets and other special equipment like ecm and bap since it would make a more fair system overall. The problem is 1) them implementing it properly. and 2) the increased wait time.

[Purlana]

Do you really think PGI can assign fair values to every mech / weapon / item ?

What is a JJ worth? How about a MG? LBX-10? Highlander vs Atlas?

Edited by Purlana, 02 March 2014 - 07:43 PM.

[Varent]

Purlana, on 02 March 2014 - 07:41 PM, said:

Do you really think PGI can assign fair values to every mech / weapon / item ?

What is a JJ worth? How about a MG? LBX-10? Highlander vs Atlas?

realistically it wouldn't add as much wait time now that I think about it since the overall mech would already havea simple single number attached to it taking into account the weapons, equipment, chasis. From there take that and also add in the pilots skill with that mech and you then have a basic idea of where that player lands. Then match up based on that.

[ReXspec]

The math behind Battle Value is readily available on various Battletech TT sources...

So the excuse that PGI can't "code it in" or has "no reference" is a weak excuse.

Battle Value is the value that effectively gives numerical value to a 'mechs, weapon, or piece equipment's overall combat effectiveness.

To quote Sarna, "Battle Value is a point-based system for the Classic BattleTech board game to measure the battlefield value of a given unit and to balance opposing forces.
Although it does factor the quality of the pilot, prevalence of a C3 network, and the size of the force relative to the opposition into the equation it is only a rough guideline; it does not take terrain features into account and naturally cannot account for the random element inherent to dice rolling (although the latter aspect is factored into the battle value assigned to individual weapons, to a degree)."

So yes. Battle Value does account somewhat for RNG, which, in a skill-based shooter/simulator like MWO, would be... difficult to translate into non-turn-based game UNLESS you gave the weapons in MWO some sort of random ballistic aspects to them such as accuracy/crosshair deviation based on movement, recoil, or some kind of firing cone like the machinegun has.

Edited by ReXspec, 03 March 2014 - 07:06 AM.

[Karl Streiger]

I had this idea a while back.
While should PGI at all find a calculation for BattleValue:

Look at your stats:
for each mech you have the W/L ratio, K/D ratio, damage and xp ratio.

Those values are highly dependable on the performance of this Mech - (the player too - but that should allready be covered by ELO)

If you take those 4 values - press it into a formula - and use in on all players - >
you get a "real" battlevalue for each Mech - including all possible builds

maybe you can add a modifcator how often this Mech is in usage - for example if the JagerMech is in average the Mech with the highest damage output - and so it is the most used heavy Mech - > you get a modification of 1.5 for the damage part

and so on.

Ok the drawback:
When you are a newb after 26 games - and you buy your self a very "bv" expensive JaegerMech - and run it in Stock Mode only -you ruin the MM for your team.

[Xeno Phalcon]

[IceSerpent]

Varent, on 02 March 2014 - 02:14 PM, said:

I like it but it also needs to take into account how those weapons interact with the mech weight, and lack or addition to jump jets. Also will need to play out with how they interact with the other weapons as well. For example a mech with jump jets and ppc/ac is far far superior to a mech without jump jets and the same load. In addition to that a mech with just one ppc and then all srm would have a wonky vaue as well since It wouldn't necessarily bring all those weapons to bear equally. That said in addition having them code this properly would be key.

Varent, you are overcomplicating things a bit except for JJs that definitely should be included in BV calc. Weapons themselves are fairly easy - just check what damage you do at various ranges (i.e. at 100m intervals all the way to max range in game), throw in a few modifiers for guided/unguided and spread/pinpoint, and add everything together to get BV of your weapon setup. That way you don't end up with wonky values.

[C E Dwyer]

its probably because people would cry and scream about things sucking even more than they do now.

An off the cuff example is the auto canon 2. In turn based it was a bit of a joke in this game its a nasty long ranged weapon because of its rate of fire, people still quad them though I don't now why, as 2xac2 and 2 x ac5 is cooler and better.

its dps is about 5 times higher than TT (i'm sure someone will correct me) so its value is going to be in theory 5 times more. if you quad them is that just the value mulitplied by 4 or is there an extra cost for each ac2 taken over two (much like heat scale)

Same with ammo turn based a ton was enough for paired ac2 but you need 3 or 4 tonnes is that cost a simple multiple or does it also have a gradient increase because you can sustain the hurt longer.

Are ac2 quad in a jaeger less or more effective than 3 in a victor or battle master.

A simple BV based on tonnage won't work because in turn based there are not hit boxes of different shapes and sizes.

Should a Zoom module cost more Bv when coupled with ac2/5 combo, than some klutz thinking its going to help them hitting with an AC20.


In theory its simple, in practice its hard, to balance.

Imagine the screaming and tears if Bv turns a persons cheese build guasspult into something costing twice as much than an atlas, with ML's ac10 and srms, and thats not outside a possible out come. neither is an ac2 streak shadowhawk costing more than a ppc ml awesome.

[Davers]

Willard Phule, on 02 March 2014 - 03:51 PM, said:

Will you end up with noobs? Maybe. But they won't be noobs in completely useless mechs. They'll at least be able to help out a little before they end up as charred moron corpses in Terra Therma. Even complete ******* in the Stalker Trial will get a volley or two off before they die. At least they aren't the idiots in Trial Spiders, running around in 3PV, that drag the match out for 8 minutes because they don't have a clue that Skirmish is Skirmish.

Won't the best players with the best builds be paired with the worst players and the worst builds?

How much weight is put on tonnage and equipment,? Is being a Cicada better or worse than being a Jenner?

In the end, doesn't all those formulas just come down to 'Did you win or lose?' So, basically, Elo?

[Tiamat of the Sea]

Because battlevalues, even BV2, are not a good representation of 'mech power. I've tried playing tabletop (which BVs are meant for) both by BV and by BV2, as well as by tonnage/'mech#, and of the three ways of limiting 'power', the third is the closest to an even keel- and most tabletoppers I've met of any skill level are in agreement on that.

[Deathlike]

To the title of the thread:

Short answer: Because PGI.

Long answer: It takes effort to implement things... which is "kinda" why Elo was added. The problem though is understanding how Elo functions per chassis or even per variant let alone per player and how they function per team.

So, it'll probably never happen... and if it did happen, it wouldn't work as intended.

[Willard Phule]

Then, the final solution is really very simple. Since all decision making is based on the "new player" experience....and "new players" will whine and cry if they get killed.....

Reduce the damage of ALL weapons to 1 point a piece. Make the maximum range for ALL weapons 10 feet (not 10 meters).

Double the armor value, triple the internal structure...and viola....you've got happy, fuzzy, newbville where noone dies, everyone gets to run around in their stompy robots. Teamwork doesn't matter, metabuilds don't matter....you get to stomp about for 10 minutes.

Oh, and everyone gets 10,000,000 cbills for simply participating.

[Crashingmail]

Within the german community we are using a system similar to BV 3.0. This was designed by a colleague to calculate everything that can be build into a mech except modules.

Right now S70 Tournament is running and for 3025 Stock this system is used. The people like it especially which a restriction how many points can be used, also some mechs are used on the field that normally nobody wants to drive.

For the other elements of this tournament 3025 Open class, 3050 Stock and 3050 Open Class they decided not to use this value.

Why PGI dont try anything similar is really bad and for myself i also cant understand it. Right now the MM, even with the changed, will be a mess. And this gets even worse when the Clans show up on the field.