Jump to content

State Of Match Making - Feedback/comments


1142 replies to this topic

#621 Kjudoon

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 7,636 posts
  • LocationWisconsin

Posted 05 September 2015 - 02:17 PM

View PostFire for Effect, on 05 September 2015 - 02:12 PM, said:


Oh I have data: it was already tried and it had failed. Can there be better data?

I sincerely doubt that PGI was able to interpret data. The past has shown that they have difficulty with interpretation...
Maybe they got better but I doubt it.

Apart from that its rather easy: compare damage and resilience of Mechs:
Oh cicada can do less that Stormcrow so it make no sense to put both into the same bucket...

basic logic also helps:
two identical pilots one in a locust one in a Mad Cat are simply not equal so you need BOTH metrics for matchmaking....

You have data? News to me. Everyone has circumstantial evidence, it's what we all base our opinions on.

The rest of your post deals with the need to use BV for matchmaking. I've agreed with the idea that your individual mech's battlevalue as based on the chassis and your loadouts as a PSR multiplier. Bad mechs should modify the psr of players the same as good mechs. So some hotshot player gets into his 9 flamer hunchie for a drunk lark, he should end up in a lower tier. On the other hand, some relatively new casual player makes his first DireGauss/PPC... up he goes!

In this I think everyone agrees.

The problem is going to be whether PGI implements such a system, and current impression is they are not.

#622 Wintersdark

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • 13,375 posts
  • Google+: Link
  • Twitter: Link
  • LocationCalgary, AB

Posted 05 September 2015 - 02:20 PM

View PostKjudoon, on 05 September 2015 - 02:00 PM, said:



PvE should not be just for new players and then force them out into 'the real game'. Most gamers I know would rather play Co-Op multiplayer than PvP just about every single time. You need a place that is an engaging PvE experience that is not lather-rinse-repeat and boring after the third time you play it. It needs to be a place also for those who have burnt out on PvP play but not MWO to go and cool off and that means good gameplay with depth that can't be had in PvP.

But you must confess that Tri is right: none of us have data and are posting are opinions... except for Russ. He's got data.
oh, man, give me co-op PvE and I'll never touch PvP again.

I loved the previous Mechwarrior games and played the hell out of them... And never touched PvP on them either. I hate PvP, and view it as a necessary evil in MWO.

#623 Wintersdark

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • 13,375 posts
  • Google+: Link
  • Twitter: Link
  • LocationCalgary, AB

Posted 05 September 2015 - 02:26 PM

View PostKjudoon, on 05 September 2015 - 02:17 PM, said:

You have data? News to me. Everyone has circumstantial evidence, it's what we all base our opinions on.

The rest of your post deals with the need to use BV for matchmaking. I've agreed with the idea that your individual mech's battlevalue as based on the chassis and your loadouts as a PSR multiplier. Bad mechs should modify the psr of players the same as good mechs. So some hotshot player gets into his 9 flamer hunchie for a drunk lark, he should end up in a lower tier. On the other hand, some relatively new casual player makes his first DireGauss/PPC... up he goes!

In this I think everyone agrees.

The problem is going to be whether PGI implements such a system, and current impression is they are not.


The problem with (traditional) BattleValue in MWO is its a balancing nightmare. When discussing it, people say idiotic things like "You just assign values to..." and crap like that, but that's like "you just solve world hunger, then everything is great!"

The TT values are utterly useless for MWO, not even as a starting point.

No insult to PGI here, but would anyone argue that quirks and weapons are perfectly balanced, and have made everything wonderful? No. Adding another system of values that need to be tuned against a moving backdrop of game systems would make it harder, not easier.

It's not actually impossible, and if implemented correctly would be awesome.

But the implementation cost would be enormous, and its FAR more likely that the system would be flawed. A flawed BV system lets players directly exploit matchmaking to end up in an actually OP build relative to his opponents. That's really bad.



(With that said, dynamically generated BV based on weapon/mech usage has possibilities, but that's still a substantial amount of complexity and design work).

#624 Mechteric

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 7,308 posts
  • LocationRTP, NC

Posted 05 September 2015 - 02:30 PM

I've pretty much given up on the group queue because of the large groups that are allowed.


Now that there's community warfare and private matches, I feel like groups of 12 (or between 4 and 12) have their outlet. It is time to make the public group queue more attractive for those of us who aren't easily able to get groups of 12 going.


So yes, 100% in support of limiting public group queue to 4 players.

Edited by CapperDeluxe, 06 September 2015 - 05:53 AM.


#625 Kjudoon

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 7,636 posts
  • LocationWisconsin

Posted 05 September 2015 - 02:36 PM

View PostWintersdark, on 05 September 2015 - 02:26 PM, said:

The problem with (traditional) BattleValue in MWO is its a balancing nightmare. When discussing it, people say idiotic things like "You just assign values to..." and crap like that, but that's like "you just solve world hunger, then everything is great!"

The TT values are utterly useless for MWO, not even as a starting point.

No insult to PGI here, but would anyone argue that quirks and weapons are perfectly balanced, and have made everything wonderful? No. Adding another system of values that need to be tuned against a moving backdrop of game systems would make it harder, not easier.

It's not actually impossible, and if implemented correctly would be awesome.

But the implementation cost would be enormous, and its FAR more likely that the system would be flawed. A flawed BV system lets players directly exploit matchmaking to end up in an actually OP build relative to his opponents. That's really bad.



(With that said, dynamically generated BV based on weapon/mech usage has possibilities, but that's still a substantial amount of complexity and design work).

yes I am not saying we should use TT numbers, I'm saying that the concept is very sound and we should use what PGI is doing for a BV on chassis as a basis for creating something similar to act as a multiplier for PSR. Their whole goal with what they've stated is to help standardize and make quirks better and representative of the 'physical' characteristics of the mech regardless of skill. that's what I got out of what I've read at least.

The next step would be to make it dynamic based on loadout and then implement it for MM with PSR.

#626 McMurl

    Member

  • PipPipPipPipPip
  • The Ominous
  • The Ominous
  • 186 posts
  • LocationEdmonton

Posted 05 September 2015 - 02:38 PM

i think that 4 man max for group sizes not only is a great idea, but is necessary. for people in higher PSR groupings, too often is it a MS or 228 comp team rolling against a group of people who aren't in meta mechs or competitive mechs. it makes the team part of the game very boring and frustrating.

I am dissatisfied with the current group queue mechanics. please switch it to 4 people max and leave 12 mans to CW.

#627 Osric Lancaster

    Member

  • PipPipPipPipPipPip
  • 447 posts

Posted 05 September 2015 - 02:51 PM

Reading the first page and . . .

View PostVlad Ward, on 03 September 2015 - 02:13 PM, said:

Every queue discussion in GD since the PSR change has ended with an Opt-in/Opt-out merged queue compromise. How is this not even on the radar?


View PostI Zeratul I, on 03 September 2015 - 01:56 PM, said:

Different perspective.

Can we have an option to pug in group queue?

It could decrease wait times and make those gaps easier to fill.


View PostGroovYChickeN, on 03 September 2015 - 02:04 PM, said:

Also as some have suggested. Would allowing solos to opt-in to the group queue help this situation at all?


Well it's good to see that someone has been paying attention. The problem is. . .

View PostRuss Bullock, on 03 September 2015 - 02:46 PM, said:

I really want to avoid this - it will only degrade the solo queue experience at least by some amount.


This is a non-argument. HOW would it degrade the solo queue?

View PostRuss Bullock, on 03 September 2015 - 03:11 PM, said:

I will think on it again - MAYBE if it pulls just a very few it might be okay. I do not want to effect the quality of the solo queue.


You're talking about an opt-in, remember. You can set it to default players to the solo queue, and only draw players out of that bucket that opts in when your group match maker needs some single player spackle.

View PostRuss Bullock, on 03 September 2015 - 03:28 PM, said:

Just a thought that is just a MAYBE atm.

What if we kept all group sizes, allowed a few solo players.

Then we make game mode random in group queue. If we could do that and limit group queue creation at least some what more, if not 1/1/1/1 than perhaps 2's as some have suggested.

Yes perhaps removing all odd sized groups would help - although the group size of 3 would be tough to get rid of.


Hurray, progress!

http://mwomercs.com/...odule-revision/

Edit: Remember guys, the 'number of players in group vs. solo queue is not a simple reflection of how many people want to be playing solo / in a group, but also of how easy it is to find a group to play with.

Edited by Osric Lancaster, 05 September 2015 - 03:49 PM.


#628 Fire for Effect

    Member

  • PipPipPipPipPipPipPip
  • Mercenary Rank 5
  • Mercenary Rank 5
  • 583 posts
  • LocationGermany

Posted 05 September 2015 - 03:16 PM

View PostKjudoon, on 05 September 2015 - 02:00 PM, said:


And we get back to the root problem: The player population is so darn low dividing it into new groups is a severe problem. Do you remember the 12man only queue? What a horror show that was. .... But to devote an entire queue for a group that could not even manage to get more than 36-48 players at any one time is abyssmal by every measure for a game.

PvE should not be just for new players and then force them out into 'the real game'. Most gamers I know would rather play Co-Op multiplayer than PvP just about every single time. You need a place that is an engaging PvE experience that is not lather-rinse-repeat and boring after the third time you play it.

But you must confess that Tri is right: none of us have data and are posting are opinions... except for Russ. He's got data.


well yes I was there and yea it was horrible and on top it did not work, so we have the best data that exits. it was tried and failed.

oh no one says it is only solo the campaign I mean. am very happy to play such things together, but the point is they have to finish the campaign so they get some experience.

new players into 4 vs 4 or even 3vs 3 seems like a very useful solution, but they should never see a veteran player the first 100 games.

see remark above concerning data... tried and failed is better than ANY data.

View PostKjudoon, on 05 September 2015 - 02:17 PM, said:

You have data? News to me.


trying and failing completely is pretty hard evidence....

#629 Fire for Effect

    Member

  • PipPipPipPipPipPipPip
  • Mercenary Rank 5
  • Mercenary Rank 5
  • 583 posts
  • LocationGermany

Posted 05 September 2015 - 03:35 PM

View PostWintersdark, on 05 September 2015 - 02:26 PM, said:

The problem with (traditional) BattleValue in MWO is its a balancing nightmare.


It is not since you completely decouple individual mechs. Mechs dont need to be "equal" anymore.
Balance groups NOT mechs. It is sufficient if both sides have the same potential to deal damage and to soak damage, you do not need to make all mechs equal, that way you can also get rid of those obscene quirks (hello dragon). Quirks are for distinguishing chassis not for producing Frankenstein abominations such as the Grid Iron.

View PostWintersdark, on 05 September 2015 - 02:26 PM, said:

The TT values are utterly useless for MWO, not even as a starting point.


and you believe that because of... ?


View PostWintersdark, on 05 September 2015 - 02:26 PM, said:

Adding another system of values that need to be tuned against a moving backdrop of game systems would make it harder, not easier.


It is not ANOTHER set of values it completely superseeds and gets rid of all other values except pilot skill. idecouples weight and combat power. you can even punish specific builts by simply giving these a penalty to combat potential rating.

Even up the sums on both side and you have a far better balance than ever before.

View PostWintersdark, on 05 September 2015 - 02:26 PM, said:

But the implementation cost would be enormous, and its FAR more likely that the system would be flawed. A flawed BV system lets players directly exploit matchmaking to end up in an actually OP build relative to his opponents. That's really bad.


do you really think it could be worse than throwing a locust and a Cheatah into the same bucket? (and yes I put the A into that mechs name on purpose.)

View PostWintersdark, on 05 September 2015 - 02:26 PM, said:

(With that said, dynamically generated BV based on weapon/mech usage has possibilities, but that's still a substantial amount of complexity and design work).


not really you just develop a basic formula, a value for each equipment. then add up for offensive and for defensive capabilities (only thingthat takes actual work is adding the good or bad hitboxes to the defensive part and presto you are finished. You can repeat that then for any design and any combination of equipments you can even put the calculation into the clients to make it easier for the server.

Easy pickings every engineer does things like that for a job every day. If they even have a working AI they can simply let all chassis let it slug it out against each other, this gives statistical strength to numbers. But the basics are insanely easy and no matter what they come up with it will be better than matching weight classes or weight since both are not in the least able to deliver an even remote approximation of the potential performance of a mech.

View PostCapperDeluxe, on 05 September 2015 - 02:30 PM, said:


So yes, 100% in support of limiting public group queue to 4 players.


we had that and you seem to have forgotten that time...

#630 Fire for Effect

    Member

  • PipPipPipPipPipPipPip
  • Mercenary Rank 5
  • Mercenary Rank 5
  • 583 posts
  • LocationGermany

Posted 05 September 2015 - 03:43 PM

View PostCapperDeluxe, on 05 September 2015 - 02:30 PM, said:

I've pretty much given up on the group queue because of the large groups that are allowed.


5.5% of all players are in large groups according to Russ so this very very small number is a problem for you?
The numbers PROOF that it can not be the large groups. Maybe better searce for another reason...

#631 Shell Game

    Member

  • PipPipPipPipPip
  • 112 posts

Posted 05 September 2015 - 04:00 PM

View PostTheStrider, on 05 September 2015 - 05:41 AM, said:

Last night was the first time in a long time that our 2 man team had a good night in Group queue. Wait times were fast, stomps weren't overly often, and we were on both sides of the coin enough not to start whining. Only time we stepped away for 5 minutes was when we faced a 8 man BMMU squad 2 games in a row, who were running nothing but high end meta.

By high end meta, you mean the BMMU players weren't running Lrms?

Edited by EpicWarlord, 05 September 2015 - 04:18 PM.


#632 Wintersdark

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • 13,375 posts
  • Google+: Link
  • Twitter: Link
  • LocationCalgary, AB

Posted 05 September 2015 - 04:10 PM

View PostFire for Effect, on 05 September 2015 - 03:35 PM, said:

It is not since you completely decouple individual mechs. Mechs dont need to be "equal" anymore.
Balance groups NOT mechs. It is sufficient if both sides have the same potential to deal damage and to soak damage, you do not need to make all mechs equal, that way you can also get rid of those obscene quirks (hello dragon). Quirks are for distinguishing chassis not for producing Frankenstein abominations such as the Grid Iron.
You still need balance. Not "This mech is equal to that mech"; but "this mech is worth <the assigned battlevalue>". The later is no easier than the former, because if you can say "Mech X is worth 900pts and Mech Y is worth 1000pts" with accuracy, you could have (not saying this is a better situation, but it was PGI's goal) buffed Mech X to be worth 1000pts, say with quirks. The trouble, however, is accurately assigning points to mech power is non-trivial, because mech power is an enormously complicated beast.

It's *REALLY FREAKING HARD* to assign accurate points values. This applies in a tabletop game (anyone who's played a point based tabletop wargame can attest to this) and it's trivially easy in a TT wargame vs. MWO.

Quote

and you believe that because of... ?
Because I have two brain cells to rub together.

Seriously, think about it for a minute.

An Awesome is a great chassis in tabletop. But tabletop has no concept of mech geometry, while mech geometry is critical in MWO. Right there, TT chassis point values are totally garbage.

Weapons? A PPC is a strong weapon in Tabletop. It's kinda junky in MWO right now, because it's extremely difficult to use accurately due to low projectile speed.... Another factor that's entirely absent in Tabletop.

In TT, you can build a reasonable formula for weapon balancing because ultimately (virtually) all TT weapons perform comparably - they have similar hit percentages, and any modifier can be directly factored in - in essence, (virtually) all tabletop weapons are the same thing. There's no real game mechanics difference between a Laser and an Autocannon, short of one needing ammo. They all fire once per game turn. They all hit random target locations. There is no convergene! No burn times! Mechwarrior Online is, functionally, an entirely different game from Tabletop, no matter what similarities remain.

Quote

It is not ANOTHER set of values it completely superseeds and gets rid of all other values except pilot skill. idecouples weight and combat power. you can even punish specific builts by simply giving these a penalty to combat potential rating.

Even up the sums on both side and you have a far better balance than ever before.
Did you read my post at all?

I said:

Quote

It's not actually impossible, and if implemented correctly would be awesome.
IF IT WORKED, it'd be great. Conceptually, battlevalue is awesome. But it's essentially impossible to implement in MWO quickly enough, at a reasonable enough dev-cost, and it's totally unreasonable to expect it to be near balanced enough to work.

Quote

do you really think it could be worse than throwing a locust and a Cheatah into the same bucket? (and yes I put the A into that mechs name on purpose.)
Objectively yes, it can be worse. Poor point costing leads to heavily, deliberately exploitable situations where you take under-costed mechs, and get matched against inferior targets because the points costs are balanced incorrectly. As trying to balance all those points costs correctly is virtually impossible, but exploiting them is simple, it becomes very easy to abuse the hell out of the system.

Quote

not really you just develop a basic formula, a value for each equipment. then add up for offensive and for defensive capabilities (only thingthat takes actual work is adding the good or bad hitboxes to the defensive part and presto you are finished. You can repeat that then for any design and any combination of equipments you can even put the calculation into the clients to make it easier for the server.

Easy pickings every engineer does things like that for a job every day. If they even have a working AI they can simply let all chassis let it slug it out against each other, this gives statistical strength to numbers. But the basics are insanely easy and no matter what they come up with it will be better than matching weight classes or weight since both are not in the least able to deliver an even remote approximation of the potential performance of a mech.
Ok, stop right there.

This is the "Oh, you just solve world hunger" foolishness I mentioned above.

There's no easy pickings, no "you just." What you want here is incredibly hard.

You can't just calculate DPS stats to determine weapon value. In TT you can (and they do!) but you can't here, because our weapons are nothing like tabletop weapons.

You've also go accuracy - how much does hitscan weigh, in absolute numbers of points (or even percentage of points) against varying projectile speed? What about tracking (SSRM, LRM)? What about bullet drop? How does the value of an AC10 change if it doesn't have bullet drop?

Now you've got physical geometry - something else that doesn't happen in TT. An Awesome is a great mech in TT, because it's no easier or harder to hit than anything else. Hitboxes, too. Everyone's seen how much hitboxes can impact how good a mech is.

How much more is a E and Ballistic hardpoints like the Firebrands arms worth vs. the AS7-DDC's RT Ballistics and GorrilaArms worth? How does 2B in the RT compare to, say, 1B in the RT and 1B in the LT? Same number of hardpoints, but different use.

Now remember, you need to build formulas for everything. Not just weapons.

Engines. Is an XL worth the same in any chassis, regardless of hitboxes? Are engines adequately represented in comparison to weapons? Or can you basically use any engine without significantly impacting your overall BV? Or, the other way, do minor engine changes change the overall BV too much, allowing you to downgrade your engine, take better weapons, and end up undercosted?

Twist rate. Turn rate. How does your degrees per second turn rate (and remember, it's a curve based on speed and engine size) impact your BV? Does it relate well to engines? What about weapons? What's 5% acceleration worth? How do Hoverjets compare with a PPC?

All these things matter, and they matter a lot. If you don't account for them, then the resultant BV must be incorrect.

It's not actually impossible, but it's so hard as to be utterly impractical and extremely expensive in terms of dev time.


Finally, say they do. Say the magically Make It Happen, have great BV scoring. Then they change an underlying game mechanic. Maybe they make ECM worse (thus making tracking weapons better), or they alter how mechs move. Suddenly, all those formulas are wrong.


In short: If the values for a BV system are wrong, the best result is poor matchups (say, Dragon's getting matched deliberately against Timberwolves); but the worst result is that people can deliberately take undercosted setups where they have a mech much more powerful than what it'll be matched against, because it's BV is under what it should be.


#633 kazlaton

    Member

  • PipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 170 posts
  • LocationNew Jersey

Posted 05 September 2015 - 04:14 PM

I would prefer to see the game mode choice removed before limiting group sizes again. And I would love to have the option as a solo player to opt into the group queue. I would do that every time.

#634 Kjudoon

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 7,636 posts
  • LocationWisconsin

Posted 05 September 2015 - 04:18 PM

Discounting others experiential evidence invalidates yours you know, Fire. Lets not be so quick to blow it off.

Also yes we had the 4man cap. The limit was not the problem. Syncdrops and elo averaging versus pugs unable to coordinate mixed against groups hiding ringers who never should have been in that match and on ts was.

Those are things that have been improved thanks to voip and now psr giving more accurate skill assessments. Averaging is still a problem but this is always going to be a problem unless you lock the tiers or who you can group with. Not saying those are good options but they are options.

But lets not blame 4mans limits for things it was not doing.

#635 MauttyKoray

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,831 posts

Posted 05 September 2015 - 05:07 PM

I think it should be an either/or choice here. Limit the group queue back to 4 (you're taking a whole lance, be happy with that), or limit group creation to a 1/1/1/1 makeup should be implemented.

People may ***** about this but we already had 4 man group queue limitations and it worked just fine, even to the point of creating BETTER matches than the average we have now. That and 1/1/1/1 is only an extension of our current 3/3/3/3 limitation, but without reducing it to 4 man queues it would allow groups to stay normalized. Thus if 2 6 man groups match together, you don't have the issue of having 5-6 heavies, at most you'll have 4.

#636 Surtosi

    Member

  • PipPip
  • Little Helper
  • Little Helper
  • 40 posts
  • LocationLouisiana

Posted 05 September 2015 - 05:17 PM

I like it, though my group would like to run together as a 12 man team for competition and major operations.

What about assigning a 13th player in a command role for these big groups so the match maker has a person to tie the 3 lances into a company? Would be cool to have that.

#637 Fire for Effect

    Member

  • PipPipPipPipPipPipPip
  • Mercenary Rank 5
  • Mercenary Rank 5
  • 583 posts
  • LocationGermany

Posted 05 September 2015 - 05:18 PM

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

You still need balance. Not "This mech is equal to that mech"; but "this mech is worth <the assigned battlevalue>".


the value is dynamically calculated according to loadout not per mech so an all flamer mad cat would have a vasty different BV than a laser gauss mad cat.


View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

The later is no easier than the former, because if you can say "Mech X is worth 900pts and Mech Y is worth 1000pts" with accuracy, you could have (not saying this is a better situation, but it was PGI's goal) buffed Mech X to be worth 1000pts, say with quirks.


it is not supposed to work that way mech DONT get flat values.

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

The trouble, however, is accurately assigning points to mech power is non-trivial, because mech power is an enormously complicated beast.


lets have a look a dire prime is worth 8901 points combat value, if you are off by a few hundred points is inconsequential then...


View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

Because I have two brain cells to rub together.


I am sure you have more...

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

An Awesome is a great chassis in tabletop. But tabletop has no concept of mech geometry, while mech geometry is critical in MWO. Right there, TT chassis point values are totally garbage.


Geometry and hitboxes are mainly a defensive function. BV would incorporate an offensive function (basically the damage you do in a given time frame) and a defensive function (basically the damage needed until you fold) calculating both is hideously easy you just have to add numbers. The geometry and hitboxes are just a modifier to that basic amount and mainly for the defensive part. you can do several approches to that modifyer.

The easiest is simply counting pixel from the front and from the back.

You can also easily add quirks to that, quirks that improve structure /Armor are just added like armor or structure points.
Quirks that increase cadence just increase the damage as if the weapon fires more often. Movement quirk just add to the engine rating making the engine having more points than it would have without quirks, and so on.

TT is not assigning BV to a chassis but to each equipment and then its simply added up. A chassis cost is simply the sum of its parts.

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

Weapons? A PPC is a strong weapon in Tabletop. It's kinda junky in MWO right now, because it's extremely difficult to use accurately due to low projectile speed.... Another factor that's entirely absent in Tabletop.


that is pilot skill.

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

They all fire once per game turn. They all hit random target locations. There is no convergene! No burn times! Mechwarrior Online is, functionally, an entirely different game from Tabletop, no matter what similarities remain.


again that is valid for all weapons so you can simply use the same approach for all it is combat POTENTIAL. How much dmg you can actually do is up to you... pilot skill

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

Did you read my post at all?


I even quoted it...

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

IF IT WORKED, it'd be great. Conceptually, battlevalue is awesome. But it's essentially impossible to implement in MWO quickly enough, at a reasonable enough dev-cost, and it's totally unreasonable to expect it to be near balanced enough to work.


I disagree, the workings are easy since, many things are already existing. (community has even counted the pixels of each mech from front and side!) Balance can be improved and will have several iterations but it will be invariably better than it is now.
Now is
locust = Cheatah;
Cicada = Stormcrow
Mad Cat = Quickdraw
Awesome = Daishi (Dire wolf)

It is hardly possible that they can make it worse in their first try.

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

You can't just calculate DPS stats to determine weapon value. In TT you can (and they do!) but you can't here, because our weapons are nothing like tabletop weapons.

You've also go accuracy - how much does hitscan weigh, in absolute numbers of points (or even percentage of points) against varying projectile speed? What about tracking (SSRM, LRM)? What about bullet drop? How does the value of an AC10 change if it doesn't have bullet drop?


All these things are essentially: "does the pilot hit the target". That is pilot skill !


View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

Now you've got physical geometry - something else that doesn't happen in TT. An Awesome is a great mech in TT, because it's no easier or harder to hit than anything else. Hitboxes, too. Everyone's seen how much hitboxes can impact how good a mech is.


That is the only thing that they cannot easily add up. But they can for example count the pixel of a mech front and side. Also geometry is quantifyable for example how much percentage of a mech is CT from sides and front. same for ST. and so on...


View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

How much more is a E and Ballistic hardpoints like the Firebrands arms worth vs. the AS7-DDC's RT Ballistics and GorrilaArms worth? How does 2B in the RT compare to, say, 1B in the RT and 1B in the LT? Same number of hardpoints, but different use.


Again count pixel how much of your mech surface area has to be shown to use the main weapons.

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

Engines. Is an XL worth the same in any chassis, regardless of hitboxes?


yes

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

can you basically use any engine without significantly impacting your overall BV?


no

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

Or, the other way, do minor engine changes change the overall BV too much, allowing you to downgrade your engine, take better weapons, and end up undercosted?


Any change in your mech changes its BV, it is NO flat BV for a chassis. So there cannot be any undercost.

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

Twist rate. Turn rate. How does your degrees per second turn rate (and remember, it's a curve based on speed and engine size) impact your BV? Does it relate well to engines?


Directly coupled to engine rating which already has BV cost.


View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

If you don't account for them, then the resultant BV must be incorrect.


Since the same formula is used for all mechs it evens out... also there is no correct or incorrect... only balance

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

Finally, say they do. Say the magically Make It Happen, have great BV scoring. Then they change an underlying game mechanic. Maybe they make ECM worse (thus making tracking weapons better), or they alter how mechs move. Suddenly, all those formulas are wrong.


why should they change it if there is balance? If they make the ML doing 1 point more damage per minute then BV for each ML goes up by one. Does not need magic just mathematics.


View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

In short: If the values for a BV system are wrong, the best result is poor matchups


cant be worse than now... see above

View PostWintersdark, on 05 September 2015 - 04:10 PM, said:

; but the worst result is that people can deliberately take undercosted setups where they have a mech much more
powerful than what it'll be matched against, because it's BV is under what it should be.


there cant be an undercost since even each point of armor is added and changes BV. It is NOT a chassis cost that gets a flat value no matter whats inside. Remember MM is not only BV it is BV times pilot skill...


View PostKjudoon, on 05 September 2015 - 04:18 PM, said:

Discounting others experiential evidence invalidates yours you know, Fire.


Its not my experiment it was PGIs experiment and it failed completely. Nothing that it was advertised to solve was solved.
Stomps are even now in solo queue as often as before you just dont have evil groups to blame it on.


I fear I have to go to bed its 3:00 here so good night everyone and read you tomorrow

#638 Wintersdark

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • 13,375 posts
  • Google+: Link
  • Twitter: Link
  • LocationCalgary, AB

Posted 05 September 2015 - 06:11 PM

View PostFire for Effect, on 05 September 2015 - 05:18 PM, said:

the value is dynamically calculated according to loadout not per mech so an all flamer mad cat would have a vasty different BV than a laser gauss mad cat.
Good lord. I get that, how can I be more clear with this? I UNDERSTAND HOW BATTLEVALUE WORKS.

But you're just not understanding the problem at all.

Try this way. Just try. Don't use words like "you just" and "it's easy", if it's so easy, try to do it.

How many points is one point of armor worth? Are points of armor worth the same regardless of where they are?

How many points is PPFLD worth? PPFLD is objectively better than spread damage or long duration burns - a long duration burn will result in damage spread around the mech. As well, high damage weapons (10+) can one-shot multiple internal components in a single hit, whereas a weapon doing lots of low-damage hits for the same overall DPS will destroy far, far fewer internal components.

How many points is 5 rating of engine worth?

How many points is a PPC worth, an AC10?

The important part here is that the answers to each of the questions above have to work in comparison to each other, as parts of the whole. THIS is extremely important - the comparison between categories more than individual components.




Even if you can do this, I'll show how it totally doesn't work at all with existing mechs. With zero quirks, a gauss rifle mounted on a Jagermech's arm is simply better than a gauss rifle mounted on a Thunderbolt's arm. They are NOT equal. If you cost them the same, and don't charge the Jagermech for it's high shoulder mount, it'd simply be a better mech when equipped with a Gauss rifle.

Otherwise, why do you think the JM6 totally eclipsed the K2 the moment it was released, doing everything the K2 did, but better.

But you're glossing over the hard part, and just calling it easy. This is a very difficult aspect of game design. It's hard.

And you know what? It damn well can be worse than it is right now. It *has* been worse. Much worse. Arguing that some totally vague idea "can't be worse" is utterly useless and quite frankly flatly wrong.

Edited by Wintersdark, 05 September 2015 - 06:11 PM.


#639 WANTED

    Member

  • PipPipPipPipPipPipPip
  • Knight Errant
  • Knight Errant
  • 611 posts
  • LocationFt. Worth, TX

Posted 05 September 2015 - 06:48 PM

As Asaru and Tsula of C4 said, I am not in agreement with reducing to 4 man groups again. I like jumping on and playing with my c4 friends and when a game says " sorry bub you can't join your friends " I'm out. I do agree with reducing play modes. I'm fine with taking down any play modes. Pick one. I know this is a hard choice and glad to see PGI listen to the community before making decisions. I think too many play modes are channeling players in different areas and reducing this will help. CW is another story but I did like you guys coming up with single events to channel players at one planet. That's a great idea.

#640 Killashnikov

    Member

  • PipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 187 posts
  • LocationSydney

Posted 05 September 2015 - 06:56 PM

All this ignores the point CW is underpopulated in oceanic peak times. MAke it <4, 8 or 12 man...





15 user(s) are reading this topic

0 members, 15 guests, 0 anonymous users