51 replies to this topic

[Willard Phule]

Ok...so before I start this HUGE post ( and it will be HUGE)...if you're only going to read the first paragraph or two before you start "White Knighting" PGI's decision to screw us all over...all I can say is "stand next to Sandpit and put on your tinfoil hat."

So...here it goes.

The concept of using a mathematical number, quantifiable by not only what we're using but how good with it is not only doable...it's simple. Well, simple to everyone else but PGI, apparently, but there you have it.

I'm not a statistician, mathemetician or anything like that...I'm just a disgruntled old guy with a grounding in BattleTech and the MW series of games. Take that for what it's worth..but I guarantee you on thing....you CANNOT "outstupid" me. I was a NonCommissionedOfficer in the US Army. We're professionally stupid. They pay us to do the most ignorant things you can imagine...and I did the "stupid" better than most. I've been Infantry, Artillery and Human Resources...I can shoot them, blow them up and file the paperwork afterwards. If that isn't the definition of stupid, I don't know what is.
-------------------------------------------------------------------------------
Let's talk about "Combat Value" real quick. It's a simple mathematical equation. It's the "BattleValue" of the machine you're driving added to the Elo of your skill. I won't pretend to know how your Elo is calculated. PGI has devised some complicated algorythm that compiles your wins, losses, kills, deaths, random dice rolls and a magic 8 ball to come up with that value. Just accept it. YOU have an Elo. Cool.

"Combat Value" is a simple addition thing. It adds the "Battle Value" of your 'mech with your personal Elo. A+B=C. Hell, even my dumb ass can grasp that. If you can't, then you're playing the wrong game....or, in the case of PGI devlopers, you're coding the wrong game.
---------------------------------------------------------------------------------
Ok, the first thing I need to talk about is "Battle Value." I'm SURE I'm about to get flamed by no less than a dozen people that don't understand how Battle Value (BV from here on out) is calculated about how the idea is flawed. Personally, I don't really care what you think but I'm going to go into GREAT detail to try to mollify you non-BT players out there. Bear with.

BV (Battle Value) is based SOLELY on the "machine." Every single BattleMech generator out there provides it for free, based on the concept provided by whoever...whenever. Does everything mentioned equate to how stuff perfoms in MW:O?

Of course not. But, you know, if I can figure it out...a 3rd grader can. I'm not smart, I'm stupid...were you not paying attention. The LB10-X in Battletech isn't the same as in MW:O. It doesn't use switchable ammo...it's got a crosshair...whatever. Am I the only one that can look at the BV of an LB10-X given in canon and say "gee, that ain't right...this would be closer?" I appears that way. Man. Everyone that plays this game must be so much more intelligent than me....I should probably just choke on my tongue while I sleep. It'd save the rest of the world later, right?

For what it's worth...it's pretty hard to come up with a BV of 3000 or higher. Can it be done? Sure...but you gotta work at it. And you're not going to do it with what's available with MW:O....don't believe me? Head on over to REMLAB and throw together any mech that you've put together in MW:O with the engine restrictions, hardpoint restrictions, etc. Try it. You can't do it.

But what about Mech Efficiencies and Modules? What about them? Assign them a value and drive on. Here's a good example with the efficiencies...Speed Tweak give you an addiontal 10% speed boost. It could take you from 5/8 to 6/9. How much would that movement increase increase your BV with TT rules? So, assign it and move on. It's not hard. But...Arty Strike! Ok..It has X potential of damage and is a 1-shot item....there are rules for it. Code it.

This stuff, although the longest part of this thing, isn't hard. Hell, my ignorant ass could come up with arbitrary values that could be adjusted over time in an afternoon of drinking beer. Maybe PGI should hire me.

That's it for BV....use canon rules. Add crap. Don't stress the small stuff.

Max BV for the "machine?" 3000.
----------------------------------------------------------------------------------
Here's the REALLY simple part of this...the "meat" part of the equation.

Whatever silly crap they're using to determine Elo, keep it. From what I understand, the max Elo is 2900. Increase it to 3000. Simple as that.
------------------------------------------------------------------------------------
Combat Value (CV)......

OMG...this is hard. Add BV (max 3000) plus Elo (max 3000). BV+Elo=CV. OMG...how hard is that?

Ok..so, let the whining begin, but let me put a couple of things out there for you.

Maximum end of the spectrum: You have a guy (Elo of 3000) with a completely maximized Atlas (BV of 3000) and he knows how to use it. Guess what? He has a CV of 6000. You can call this guy a "ringer." He's going to sway the battle no matter what you do, and he should....most of us don't fall into that range.

Example B: You have a guy who's badass with Lights (Elo of 3000) piloting a Locust (BV of 1000). Ok, so his CV is only 4000. But he's gonna run circles around the other lights. His CV is well earned.

If you limit a team to X CV, then the matchmaker has an even larger pool of people to pick from...and weight doesn't really matter. It WILL work out in the wash.

Sure, you'll have matches where people whine about weight indifferences...but...odds are, it'll be the team with the heaviest mechs that have pilots with lower Elos than what we're seeing right now.
---------------------------------------------------------------------------
And before you even say a thing, Sandpit, I realize this is MY opinion only.

Not every one of us lives in a happy, fuzzy world where every match is evenly balanced by skill or weight.

I can only imagine what I'd have to smoke to live in your universe.

Edited by Willard Phule, 02 March 2014 - 04:44 PM.

[Wintersdark]

Yes, yes, yes. I think everyone who understands what Battlevalue is understands that it would be a great tool.

But it's a great tool we can't make use of.

I've said it before, and will again:

You don't just "assign values and carry on" - the particular values assigned have to reflect the relative strength of each weapon/system/skill. If the balance between the item values - say, PPC and Large Laser - are incorrect, then the BV (and thus resulting CV) will be wrong. If the CV is wrong, it's at best no better than the current Elo only system, at worst it creates an exploitable system where players can use certain loadouts to smurf themselves into lower CV brackets, or others can accidentally end up fighting against better players with better builds.

The existing BattleValue data from tabletop is totally, utterly worthless. Things in MWO are nothing like Tabletop. Ghost heat, the heat system in general, convergence, armor, skill based aiming, things are just totally different here.

So, for BattleValue to work, PGI would need to create points values that accurately reflect the in-game value of everything relative to everything else. Then they would need to maintain these values as item stats/balance changes due to game mechanic/stat changes. For example, with the LPL changes last patch, they are worth more than they used to be worth. But how much more?

In tabletop, these values are much easier to come by.

For example, in Tabletop, you know the percentage chance of a given weapon hitting a given component. You can mathematically extrapolate and average Time To Kill for any given weapon then very easily. You can't do that here, because aiming is non-random.

Moreso, in TableTop, the game rules and item stats generally speaking just don't change, so you're JUST adjusting BV as your iterate over game results. Here, you'd be adjusting both stats/mechanics AND the point values. That's a tremendous game design challenge.


Finally, PGI. How good a job do you feel PGI does with prompt weapon balance changes, right now? How fast do you feel they correct imbalances and problems? Now, take some dev-time away from that to build an initial dataset, then some ongoing to review performance vs. point values on everything and adjust them. It's a non-trivial amount of time. Then, at last, how good a job do you feel PGI does with balance overall? Do you really think they'll be willing and able to build a good value set and maintain it over time as game systems change? I'd REALLY expect the values to be terribly wrong, and then not adjusted as balance changes making the whole system entirely useless.



TLDR: BV is awesome in theory, impossible for PGI to implement correctly and well.

[FupDup]

Or we can just balance out the weapons/chassis (i.e. buff underperformers, etc.) rather than using external point systems.

[CheeseThief]

The inclusion of battle value goes against PGI's current balance ideology of everything being equally viable. Hence why LRM5's fire faster and with a tighter spread than an LRM20.

If we rip the BV values out of the CBT rules as you suggest then the LRM20 (180) is worth 4 times the BV of an LRM5 (45), However 180 points BV of LRM5's is a much more focused group that is more able to penetrate AMS along with significantly higher DPS because of the faster refire rate.

To fit the TT BV system the LRM20 needs to be four times the potency of an LRM5 else the person taking an LRM 20 is at a matchmaking disadvantage, this means either the LRM5 gets nerfed and everyone takes the LRM20 because the hard point system makes small launchers worthless, the LRM20 gets buffed for the same effect, or every single item will need custom matchmaking weight balancing.

PGI is hard pressed balancing the weapons as it is, how do you expect them to deal with a second layer of matchmaker weight balancing over the top at the same time?


Weight Balancing is a better idea for closer matchmaker, it's half as complex, requires a fraction of the design work and doesn't need to be changed/updated every time they do a balance change, nor does it scew directly with individual weapon balance. Most people agree that a Hunchback is reasonably equal to a Centurion, that a Highlander is similar in power to a Stalker. CBT BV on the other hand doesn't do squat to help balance because an upgraded hunchback (1251) has the same BV as a Meta Victor (1260).

Edited by CheeseThief, 02 March 2014 - 05:38 PM.

[Willard Phule]

Wintersdark, on 02 March 2014 - 05:27 PM, said:

Yes, yes, yes. I think everyone who understands what Battlevalue is understands that it would be a great tool.

But it's a great tool we can't make use of.

I've said it before, and will again:

You don't just "assign values and carry on" - the particular values assigned have to reflect the relative strength of each weapon/system/skill. If the balance between the item values - say, PPC and Large Laser - are incorrect, then the BV (and thus resulting CV) will be wrong. If the CV is wrong, it's at best no better than the current Elo only system, at worst it creates an exploitable system where players can use certain loadouts to smurf themselves into lower CV brackets, or others can accidentally end up fighting against better players with better builds.

The existing BattleValue data from tabletop is totally, utterly worthless. Things in MWO are nothing like Tabletop. Ghost heat, the heat system in general, convergence, armor, skill based aiming, things are just totally different here.

But...it's a good place to start at. As opposed to the current system of "oh dear God, the noobs are crying...let's nerf something." that we're currently dealing with. At least, there's a numerical value that even the most ******** among us (myself) can understand.

Wintersdark, on 02 March 2014 - 05:27 PM, said:

So, for BattleValue to work, PGI would need to create points values that accurately reflect the in-game value of everything relative to everything else. Then they would need to maintain these values as item stats/balance changes due to game mechanic/stat changes. For example, with the LPL changes last patch, they are worth more than they used to be worth. But how much more?

Again...I could come up with a numerical value for ANYTHING that these morons have running in an afternoon of beer drinking and reviewing all the books that I have falling apart on my bookshelves. Would it be entirely accurate? Hell no, but it'd be close....close enough to tweak in patches. Considering the amount of tweaking we've received to try to make things work recently...it'd be better ( and easier if PGI has any sense) than what we've seen.

Wintersdark, on 02 March 2014 - 05:27 PM, said:

In tabletop, these values are much easier to come by.

For example, in Tabletop, you know the percentage chance of a given weapon hitting a given component. You can mathematically extrapolate and average Time To Kill for any given weapon then very easily. You can't do that here, because aiming is non-random.

Incorrect. The percentage chance of hitting anyone with any give weapon system...and this seems to be the #1 argument against any form of BV system...is the Pilots "Gunnery Skill." Hey, guess what? In MW:O, that is represented by your Elo. I won't even pretend to tell you how that is determined, but ultimately, it's determined by how well you do in any given weight class....both you your ability to shoot and to win. Whatever. Just consider your "Piloting Skill" and "Gunnery Skill" as your Elo. You have a problem with that? Take it up with PGI and how they calculate Elo. I'm not interested in it.

Wintersdark, on 02 March 2014 - 05:27 PM, said:

Moreso, in TableTop, the game rules and item stats generally speaking just don't change, so you're JUST adjusting BV as your iterate over game results. Here, you'd be adjusting both stats/mechanics AND the point values. That's a tremendous game design challenge.

Again. Not really. Your "Combat Value" is "Meat + Machine." Simple as that. All I've done is outline the rules for giving a numerical value to the "machine." PGI, in their infinite stupidity, has come up with a way to give a numerical value to your "meat." Deal with that. The Machine is easy.

Wintersdark, on 02 March 2014 - 05:27 PM, said:

Finally, PGI. How good a job do you feel PGI does with prompt weapon balance changes, right now? How fast do you feel they correct imbalances and problems? Now, take some dev-time away from that to build an initial dataset, then some ongoing to review performance vs. point values on everything and adjust them. It's a non-trivial amount of time. Then, at last, how good a job do you feel PGI does with balance overall? Do you really think they'll be willing and able to build a good value set and maintain it over time as game systems change? I'd REALLY expect the values to be terribly wrong, and then not adjusted as balance changes making the whole system entirely useless.

Are you kidding me? I wouldn't trust PGI to code it's way out of a one-way corridor with a single stop sign in it. That's beside the point. The dudes that programmed MW3 and MW4 had that part right.....at least, it felt that way at the time. These morons couldn't code a soup sandwich from what I've seen so far. But, then again, I'm an idiot. What can I tell you.

What I CAN tell you is that I've successfully managed any number of TT matches with different parameters based on BV that have ended up completely...COMPLETELY...reliant on battlefield tactics and dice rolls.

Maybe that's the point. Maybe PGI wants to eliminate tactics and random variables. Maybe they just want to sucker in as many ******* as possible so they can milk them in favor of producing a quality product based upon years of playability. The evidence, does in fact, lean that way.

Wintersdark, on 02 March 2014 - 05:27 PM, said:

TLDR: BV is awesome in theory, impossible for PGI to implement correctly and well.

Agreed. In fact, the only thing they seem to be able to implement at all is mulitple sales, mech designs and completely bugged maps. Seems to me they only want your money. So, stop giving it to them. Maybe when everyone on the design team is living on Ramen Noodles and wearing dirty underwear, they'll pay attention.

Too bad that'll never happen. Their marketing department seems to draw in new crops of ******* every three weeks.

[Wintersdark]

FupDup, on 02 March 2014 - 05:32 PM, said:

Or we can just balance out the weapons/chassis (i.e. buff underperformers, etc.) rather than using external point systems.

This is exactly the point.

If we can't even get balance between mechs/weapons/thingamagiggies right in the first place, what hope do we have of getting good BV numbers?

[Wintersdark]

Willard Phule, on 02 March 2014 - 05:55 PM, said:

But...it's a good place to start at. As opposed to the current system of "oh dear God, the noobs are crying...let's nerf something." that we're currently dealing with. At least, there's a numerical value that even the most ******** among us (myself) can understand.

We'd understand the number, see how bad it was, and cry about that. Knowing what the numbers were would allow players to directly modify their Combat Value as they see fit.

Quote

Again...I could come up with a numerical value for ANYTHING that these morons have running in an afternoon of beer drinking and reviewing all the books that I have falling apart on my bookshelves. Would it be entirely accurate? Hell no, but it'd be close....close enough to tweak in patches. Considering the amount of tweaking we've received to try to make things work recently...it'd be better ( and easier if PGI has any sense) than what we've seen.

Maybe you could, maybe not. I can definitely say as someone who's dabbled a lot in designing tabletop points based wargames as well as playing them, getting good points values that don't in themselves create balance problems is a whole lot harder than it looks. But, that said, perhaps you could. I doubt PGI could, because I don't think PGI really understands weapon balance (let alone mech balance) very well at all to start with.

Quote

Incorrect. The percentage chance of hitting anyone with any give weapon system...and this seems to be the #1 argument against any form of BV system...is the Pilots "Gunnery Skill." Hey, guess what? In MW:O, that is represented by your Elo. I won't even pretend to tell you how that is determined, but ultimately, it's determined by how well you do in any given weight class....both you your ability to shoot and to win. Whatever. Just consider your "Piloting Skill" and "Gunnery Skill" as your Elo. You have a problem with that? Take it up with PGI and how they calculate Elo. I'm not interested in it.

You totally missed my point.

I can assign a points value to a weapon in a points-based tabletop wargame based on exactly how quickly it's going to destroy an enemy mech on average. I can use exactly the same formula, applied to other weapons, to generate inherently balanced point values for other weapons. In short, in tabletop, you can take into account every normal variable in a fully mathematical formula (that doesn't take a math degree to build).

Elo, on the other hand, is being added to Battlevalue later, and cannot be used at data-creation time to calculate item point value. It's not a good indication of aim, either, when all is said and done.

Quote

Again. Not really. Your "Combat Value" is "Meat + Machine." Simple as that. All I've done is outline the rules for giving a numerical value to the "machine." PGI, in their infinite stupidity, has come up with a way to give a numerical value to your "meat." Deal with that. The Machine is easy.

I didn't disagree with you. But the machine part only seems easy. It's not. It's not even really easy for a tabletop wargame (see: GW's Warhammer, 40k or fantasy, point balance), but it's vastly harder with so many nebulous variables here.

I didn't disagree with any of your post, remember. It's a known system, it works well. It's just much, much more difficult to do here than it is in tabletop, and PGI is spectacularly poorly suited to do it.

Quote

Are you kidding me? I wouldn't trust PGI to code it's way out of a one-way corridor with a single stop sign in it. That's beside the point. The dudes that programmed MW3 and MW4 had that part right.....at least, it felt that way at the time. These morons couldn't code a soup sandwich from what I've seen so far. But, then again, I'm an idiot. What can I tell you.

No... It is the point. If PGI can't do it, then the whole thread is useless. Of course, that's so often the case with lots of great idea forum threads

Quote

What I CAN tell you is that I've successfully managed any number of TT matches with different parameters based on BV that have ended up completely...COMPLETELY...reliant on battlefield tactics and dice rolls.

Absolutely. No argument at all. As I've maintained from the get-go, Battle Value is a good system that works.

---

As an example: Lets take two Warhammer Fantasy style units, plucked from my hole filled memory from back in 5th edition. Well, No... lets go with two imaginary units because I can't remember real stats for anything.

In WFB, you have initiative, chance to hit, chance to wound, armor save; then the opposing force attacks back with whatever troops survive. All these dice rolls are completely quantifiable, you can give exact (to several decimal places) percentage odds of either unit winning combat and by how much; you can express numerically every aspect of how that combat will go. You can then compare point costs between individual troops, and modify them until you get resulting points that match the unit's combat ability.

You can't do that in MWO. You can't quantify how much worse an Awesome is than a Victor. instead, you just have to wing it, guess. You've got so many guesses to make, then you'd need to update them over time. Rank that Awesome too low (particularly combined with, say, a strong weapon ranked too low), and you'll have high Elo player's smurfing that bad boy all over the place, roflstomping noobs.

It's HARD to get good numbers for this sort of thing. Way harder than you'd think. Not impossible; maybe you could do it. Anyone could, given enough time. But I doubt that PGI would be able to do it good enough to not make the whole MM issue even worse than it was before they started... and, no matter what people like to think, the matchmaker absolutely could be a whole lot worse.

[Kaeb Odellas]

If PGI can't balance mechs and weapons, what makes you think they'll be able to balance Battle Value?

EDIT: Oh, wait, someone already said it. Whoops.

Edited by Kaeb Odellas, 03 March 2014 - 02:40 AM.

[Joseph Mallan]

Willard Phule, on 02 March 2014 - 04:42 PM, said:

Ok...so before I start this HUGE post ( and it will be HUGE)...if you're only going to read the first paragraph or two before you start "White Knighting" PGI's decision to screw us all over...all I can say is "stand next to Sandpit and put on your tinfoil hat."

So...here it goes.

The concept of using a mathematical number, quantifiable by not only what we're using but how good with it is not only doable...it's simple. Well, simple to everyone else but PGI, apparently, but there you have it.

I'm not a statistician, mathemetician or anything like that...I'm just a disgruntled old guy with a grounding in BattleTech and the MW series of games. Take that for what it's worth..but I guarantee you on thing....you CANNOT "outstupid" me. I was a NonCommissionedOfficer in the US Army. We're professionally stupid. They pay us to do the most ignorant things you can imagine...and I did the "stupid" better than most. I've been Infantry, Artillery and Human Resources...I can shoot them, blow them up and file the paperwork afterwards. If that isn't the definition of stupid, I don't know what is.
-------------------------------------------------------------------------------
Let's talk about "Combat Value" real quick. It's a simple mathematical equation. It's the "BattleValue" of the machine you're driving added to the Elo of your skill. I won't pretend to know how your Elo is calculated. PGI has devised some complicated algorythm that compiles your wins, losses, kills, deaths, random dice rolls and a magic 8 ball to come up with that value. Just accept it. YOU have an Elo. Cool.

"Combat Value" is a simple addition thing. It adds the "Battle Value" of your 'mech with your personal Elo. A+B=C. Hell, even my dumb ass can grasp that. If you can't, then you're playing the wrong game....or, in the case of PGI devlopers, you're coding the wrong game.
---------------------------------------------------------------------------------
Ok, the first thing I need to talk about is "Battle Value." I'm SURE I'm about to get flamed by no less than a dozen people that don't understand how Battle Value (BV from here on out) is calculated about how the idea is flawed. Personally, I don't really care what you think but I'm going to go into GREAT detail to try to mollify you non-BT players out there. Bear with.

BV (Battle Value) is based SOLELY on the "machine." Every single BattleMech generator out there provides it for free, based on the concept provided by whoever...whenever. Does everything mentioned equate to how stuff perfoms in MW:O?

Of course not. But, you know, if I can figure it out...a 3rd grader can. I'm not smart, I'm stupid...were you not paying attention. The LB10-X in Battletech isn't the same as in MW:O. It doesn't use switchable ammo...it's got a crosshair...whatever. Am I the only one that can look at the BV of an LB10-X given in canon and say "gee, that ain't right...this would be closer?" I appears that way. Man. Everyone that plays this game must be so much more intelligent than me....I should probably just choke on my tongue while I sleep. It'd save the rest of the world later, right?

For what it's worth...it's pretty hard to come up with a BV of 3000 or higher. Can it be done? Sure...but you gotta work at it. And you're not going to do it with what's available with MW:O....don't believe me? Head on over to REMLAB and throw together any mech that you've put together in MW:O with the engine restrictions, hardpoint restrictions, etc. Try it. You can't do it.

But what about Mech Efficiencies and Modules? What about them? Assign them a value and drive on. Here's a good example with the efficiencies...Speed Tweak give you an addiontal 10% speed boost. It could take you from 5/8 to 6/9. How much would that movement increase increase your BV with TT rules? So, assign it and move on. It's not hard. But...Arty Strike! Ok..It has X potential of damage and is a 1-shot item....there are rules for it. Code it.

This stuff, although the longest part of this thing, isn't hard. Hell, my ignorant ass could come up with arbitrary values that could be adjusted over time in an afternoon of drinking beer. Maybe PGI should hire me.

That's it for BV....use canon rules. Add crap. Don't stress the small stuff.

Max BV for the "machine?" 3000.
----------------------------------------------------------------------------------
Here's the REALLY simple part of this...the "meat" part of the equation.

Whatever silly crap they're using to determine Elo, keep it. From what I understand, the max Elo is 2900. Increase it to 3000. Simple as that.
------------------------------------------------------------------------------------
Combat Value (CV)......

OMG...this is hard. Add BV (max 3000) plus Elo (max 3000). BV+Elo=CV. OMG...how hard is that?

Ok..so, let the whining begin, but let me put a couple of things out there for you.

Maximum end of the spectrum: You have a guy (Elo of 3000) with a completely maximized Atlas (BV of 3000) and he knows how to use it. Guess what? He has a CV of 6000. You can call this guy a "ringer." He's going to sway the battle no matter what you do, and he should....most of us don't fall into that range.

Example B: You have a guy who's badass with Lights (Elo of 3000) piloting a Locust (BV of 1000). Ok, so his CV is only 4000. But he's gonna run circles around the other lights. His CV is well earned.

If you limit a team to X CV, then the matchmaker has an even larger pool of people to pick from...and weight doesn't really matter. It WILL work out in the wash.

Sure, you'll have matches where people whine about weight indifferences...but...odds are, it'll be the team with the heaviest mechs that have pilots with lower Elos than what we're seeing right now.
---------------------------------------------------------------------------
And before you even say a thing, Sandpit, I realize this is MY opinion only.

Not every one of us lives in a happy, fuzzy world where every match is evenly balanced by skill or weight.

I can only imagine what I'd have to smoke to live in your universe.

I worked with BV for a few months Phule, and though you are mostly right, a BV system is probably easier to game than a simple Min/Max. I played against a guy who used 5 Stone Rhinos each with a 1p/1g pilot. They had better pilots than My Sagittaire, Thunder Hawk & Atlas! If the players know how the numbers work A BV system is easier to game than raw statistics. Which is something I have a background in.

PGI saying 86% of players are playing solo is not completely true. The Mariks and Steiners have evenings where for hours they do sync drops. Everyone goes on TS3 load up A dropships someone counts down and they all hit launch. everyone who end up together moves to a new dropship. So those 4% numbers are only the players using the "group" Function in game properly.

So BV would help but only or those who don't game the BV system.

[IceSerpent]

Wintersdark, on 02 March 2014 - 05:27 PM, said:

But it's a great tool we can't make use of.

Yes, we can - see below.

Quote

You don't just "assign values and carry on" - the particular values assigned have to reflect the relative strength of each weapon/system/skill. If the balance between the item values - say, PPC and Large Laser - are incorrect, then the BV (and thus resulting CV) will be wrong. If the CV is wrong, it's at best no better than the current Elo only system, at worst it creates an exploitable system where players can use certain loadouts to smurf themselves into lower CV brackets, or others can accidentally end up fighting against better players with better builds.

Well, no kidding! Of course any system only works when you use correct values, it holds true for BV, Elo, cooking recepies, and virtually everything else in life.

Quote

So, for BattleValue to work, PGI would need to create points values that accurately reflect the in-game value of everything relative to everything else. Then they would need to maintain these values as item stats/balance changes due to game mechanic/stat changes. For example, with the LPL changes last patch, they are worth more than they used to be worth. But how much more?

No need to bother with "everything relative to everything else". Add a certain number to BV based on speed of the mech. Add another number based on total armor value. Add a 3rd number that reflects presence of JJs. This roughly covers the mech itself (you can always add smaller things liek trso twist speed later). Add a number based on heat efficiency, add an agregate number based on firepower at various ranges (modified according to damage spread and guided/unguided) - you got all weapons covered. Add a few weparate numbers for special equipment like ECM, BAP, TAG and modules. You're pretty much done, at least I can't think of anything else at the moment.

Quote

TLDR: BV is awesome in theory, impossible for PGI to implement correctly and well.

It seems to impossible for PGI to implement anything correctly and well. That doesn't really change how things should be done.

[Willard Phule]

@ Iceserpent

Add to all of that the fact that you're putting a cap on both BV and Elo which are equal. 3000 for BV, 3000 for Elo. Modules/Efficiencies should be a separate value.

So, you add the BV of the mech to the total of Modules/Efficiencies to get a number that could be easily shown in the mechlab.

You'll never know what your Elo is, but it gets added when you hit the launch button.

The MM, right now, is looking for a specific number range to put people into the match. That doesn't change. The spread may have to be a little different, but whatever.

So....let's say the MM is looking for guys in the 4500 CV range. There are any number of ways that number got put together. A guy with a 3000 Elo in a 1500 BV mech (that'd be a light or medium)....or a guy with a 1500 Elo in a 3000 BV mech.

Are they equal? Probably not but....that kind of spread is a lot closer to "even" than what we've got going on right now.

And, again, the MM could be tweaked any number of ways to make it work. With quantifiable numbers for both the player and the machine, it could very easily make better matches.

Could, that is, if it were being coded by people that gave a damn about it, I guess. Which means, that all this discussion is basically worthless..except as an intellectual exercise. If you want to play this game, you just bend over and take what you get.

[Khobai]

Quote

It seems to impossible for PGI to implement anything correctly and well. That doesn't really change how things should be done.

Agree. Battle value is ideal. We all know that. But PGI still hasnt balanced PPCs for an entire year and you honestly think they can balance battle value?

PGI is utterly incapable of looking at weapons objectively and weighing their pros and cons. Which is really easy to do... lets compare two weapons that SHOULD be equal, PPC vs LRM10 as an example:

weight/crit slots: PPC takes up 7tons and 3crits, LRM10 takes up 7 tons and 4 crits with 2 tons of ammo (bare minimum)

damage: PPC does 10 pinpoint damage, LRM10 does 11 spread damage, spread out in clusters of 1.1 damage. TAG and artemis despite their extra weight still dont make the LRM10 comparable in damage.

heat: PPC is 10 heat. LRM10 is 4 heat. But remember LRMs are one of the least accurate weapons in the game, so a lot of heat is being wasted on missiles that wont hit. LRMs are also doing spread damage instead of pinpoint damage. So I would say the PPC's damage done vs heat generated is better than the LRM10.

minrange: PPC min range is only 90. LRM10 min range is 180.

maxrange: PPCs still do pretty good damage at 700m. Max effective range for LRMs is around 500m-600m

delivery: PPCs can snapfire at 1500m/s and only need the target under reticle for a second. LRMs cant snapfire, only go 120m/s, LRMs give a warning, LRMs can be outright dodged. LRM has both soft and hard counters: ams and ecm respectively.

utility: PPCs can disrupt ecm. LRMs can indirectly fire, but only with a spotter, and even though it requires two mechs, its still inferior to one mech with a PPC.

So there you go... in less than 2 minutes ive weighed the pros and cons of two weapons. And from that we can gather that LRM10s need a massive buff to be equal to PPCs. Why is PGI so incapable of doing the same thing? I dunno...

Edited by Khobai, 03 March 2014 - 08:10 AM.

[Wintersdark]

IceSerpent, on 03 March 2014 - 07:51 AM, said:

No need to bother with "everything relative to everything else". Add a certain number to BV based on speed of the mech. Add another number based on total armor value. Add a 3rd number that reflects presence of JJs. This roughly covers the mech itself (you can always add smaller things liek trso twist speed later). Add a number based on heat efficiency, add an agregate number based on firepower at various ranges (modified according to damage spread and guided/unguided) - you got all weapons covered. Add a few weparate numbers for special equipment like ECM, BAP, TAG and modules. You're pretty much done, at least I can't think of anything else at the moment.

See? All those "certain numbers"? That's pretty much the heart and soul of game design. That's the part that everyone thinks is easy but is absolutely not.

You're all just "add a number for this, add a number for that" - no kidding. but WHAT number, specifically? Exactly?

And yes, the "relative to each other" is not only important, it's the most important part. The end result - the points values relative to each other - are what determines the balance of the system. My whole point in this thread - in every one of my posts, which I've gone over countless times - is that all these factors you want to just "assign a number to" are far harder to quantify in a game like MWO than they are in tabletop.

If your numbers are wrong, if you fail to account for something, or rate one apple inappropriately relative to an orange, you create an explicitly exploitable system where players can use things that are not actually weaker to face less skilled foes. Or, vice versa, a player could use equipment that does suck and be forced up against more skilled foes as a result - the system you've built, then, is basically just a random number added to your Elo, which is at best no better than the current Elo system.

BattleValue is great, but it requires good numbers. They have to be fairly accurate, too, "rough and good enough" is not in fact good enough, because it just makes the whole thing worse.

[IceSerpent]

Willard Phule, on 03 March 2014 - 08:04 AM, said:

@ Iceserpent

Add to all of that the fact that you're putting a cap on both BV and Elo which are equal. 3000 for BV, 3000 for Elo. Modules/Efficiencies should be a separate value.

So, you add the BV of the mech to the total of Modules/Efficiencies to get a number that could be easily shown in the mechlab.

You'll never know what your Elo is, but it gets added when you hit the launch button.

Strictly speaking, BV and Elo don't even have to be equal. You can certainly make it so, but system will work even if one has bigger impact on CV than another. Modules/efficiencies can be included in BV or added later - also doesn't make much of a difference.

[IceSerpent]

Wintersdark, on 03 March 2014 - 08:14 AM, said:

See? All those "certain numbers"? That's pretty much the heart and soul of game design. That's the part that everyone thinks is easy but is absolutely not.

Changing numbers is easier than coding a UI, it's easier than creating a new mech or new map, it's easier than designing and coding CW, it's easier than creating lobbies. If you look at the big picture of software development, it's practically the easiest task possible. Naturally, it's more difficult than scratching one's behind, but at some point developers have to actually develop the product.

Quote

You're all just "add a number for this, add a number for that" - no kidding. but WHAT number, specifically? Exactly?

It's up to PGI to decide. I would be willing to take a shot at it if I had a reasonable expectation of PGI being interested in me doing so. I know how to use a spreadsheet, so it's not an enormously complicated task for me. I am also fairly certain that quite a few other folks on these forums can do it too. That being said, why bother when we both know that it would be just an intellectual exercise?

[Trauglodyte]

You can't use Battle Value as a means to balance drops. It has too many parts and, as others have said, it is too easy to game. If people are willing to min/max weapons, speeds, etc, do you honestly think that they'll forgo the decision to do the same with that? BUT, and this is a big but (and I cannot lie - had to), they should use Battle Value as a means of balancing end game rewards. If team A wins the match and has, say 5000 BV points on team B, they shouldn't get the same level of cbills and XP. It is the same thing for when a Locust and Atlas do the same damage with the same kills and assists. Why does the 100 tonner get the same rewards when it is clearly more capable of doing things vs. that of the 20 tonner? That is, of course, debatable but is simply an analogy.

[IceSerpent]

Trauglodyte, on 03 March 2014 - 08:44 AM, said:

You can't use Battle Value as a means to balance drops. It has too many parts and, as others have said, it is too easy to game.

Why don't you give us an example of how exactly people will be able to game the proposed system (as opposed to verbatim copy of tabletop BV that's not even on the table for obvious reasons)?

[80sGlamRockSensation David Bowie]

IceSerpent, on 03 March 2014 - 07:51 AM, said:

It seems to impossible for PGI to implement anything correctly and well. That doesn't really change how things should be done.

This. Exactly this.

[Trauglodyte]

IceSerpent, on 03 March 2014 - 08:51 AM, said:

Why don't you give us an example of how exactly people will be able to game the proposed system (as opposed to verbatim copy of tabletop BV that's not even on the table for obvious reasons)?

So, I disagree with you and your rebuttle is that the burden of proof is on me? Brilliant.

Look, BV was a great thing in the TT game when everything fired at the same recycle rate. You can't simply import BV into this game when things aren't built on the same foundation. The Md Laser has a value of X and the PPC has a value of Y. Both have a recycle rate of 4s without efficiencies. But, put in efficiencies and their recycle rates differ which means that the BV foundation breaks. The SRM2 fires faster than the SRM6 so, again, the foundation breaks. The only way that you could do it would be to create some sort of algorythm that would alter the BV of a weapon based on a standarized value, which takes into account all different refire rates, that is also impacted by the state of efficiencies. What BV does is create a numerical value of a weapon in a vacuum but MWO isn't a vaccum and that is the problem.

Like I said, you could use it to alter end game rewards but that would be the extent at which you could do anything.



PS> Do you really want PGI taking battle value and doing some sort of equalization process? I mean, c'mon. They're starting to scare me as it is when they can't see past their hand on the easy decisions that they make.

Edited by Trauglodyte, 03 March 2014 - 10:29 AM.

[Almond Brown]

IceSerpent, on 03 March 2014 - 08:51 AM, said:

Why don't you give us an example of how exactly people will be able to game the proposed system (as opposed to verbatim copy of tabletop BV that's not even on the table for obvious reasons)?

Break out your spreadsheet, in its entirety, and we will gladly accommodate your request. Otherwise, you ask that we try and game a "non-existent system". wtf?