Jump to content

A Dynamic Self-Balancing Battlevalue System


115 replies to this topic

Poll: A Dynamic, usage driven system for Battle Value (196 member(s) have cast votes)

Do you support use of such a system in MWO?

  1. Yes (168 votes [85.71%])

    Percentage of vote: 85.71%

  2. No (28 votes [14.29%])

    Percentage of vote: 14.29%

Vote Guests cannot vote

#41 Lukoi Banacek

    Member

  • PipPipPipPipPipPipPipPipPip
  • WC 2018 Top 12 Qualifier
  • WC 2018 Top 12 Qualifier
  • 4,353 posts

Posted 06 March 2014 - 02:11 PM

I voted no.

Here's why.

I don't agree with Elo currently either, so I'm not defending that methodology and this idea has a LOT of merit.

I think there should be occasions that don't require the use of a premium private match (PPM) where a team might be bringing all of the "optimal" mechs.

If we're going to ask the Devs to add a dynamic system such as this (which on the surface I can agree with, but watching how it ended up in Proxis, a similar system in many ways, there are some limitations that would hurt the overall playerbase), then add a player driven economy as part of Community warfare, where Houses, Units, Clans etc all contribute to the galaxy-wide supply of mechs and materiel that would determine these BV's.

Edited by Lukoi, 06 March 2014 - 02:12 PM.


#42 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 06 March 2014 - 02:22 PM

Varent and I have been discussing this in PM's, but I figure that it's worth explaining this publicly since others may be under the same mistaken belief.

View PostVarent, on 06 March 2014 - 02:03 PM, said:

the jenner K is an unused variant that is not used simply because the founder and that variant of jenner is better with its hardpoints overall. You can do both builds the laser streak and the ppc on the founder, and everyone does, there is no real major reason to own the K overall. That would under this system allow you to abuse it. This is just one example because its the first to come to mind. However you can also do the same dual erlarge laser on ever raven just about as well. These are things simply off the top of my head.

Oh, without question, this system would result in the overall Jenner population being more evenly distributed across the different variants... But that's good. That's kind of the point of the system.

It still won't result in any of them being cheap, because those Jenners will still each be very good chassis.

Your fear here is that, since there are FOUR Jenner, that they are going to be able to somehow split their cost across all of them. But this is ok, because all the other variants of other mechs also exist in multitudes. The only way those jenners would become cheap would be if their usage actually dropped below the average drop rate for all variants of all mechs... Meaning that you'd be seeing the same number of them as you do mechs like the Raven 2X.

Again, I think part of the problem you're having is that you're hyper focusing on a small part of the overall market, and then forgetting that it's part of the overall market when BV values are calculated.

In the scenario you imagine, you're imagining that there will be one expensive Jenner, and the others will be real cheap... then everyone will just move to one of the other Jenners, and then another, then another..

But this can't actually work. In the real system, that's not what happens.

In the real system, what will happen is the usage will equally distribute across all of them at the same time.. because doing otherwise would involve intentionally spending more BV than necessary. Because as soon as you start using that K, it's price starts going up. That price movement is slow and gradual, but it happens and does so at the same time the first Jenner's price is going down. As soon as that starts happening, in the imagined scenario where all the other Jenners have a price of zero or whatever... then they are immediately cheaper than the K now (and much cheaper than the original Jenner everyone was using). So people aren't going to all migrate to the K. They're going to distribute across all the Jenners (if they can in fact all do what they want), because otherwise they'll be spending more BV than they need to.

As I mentioned previously, I specifically considered exactly the situation you are worried about here, because originally I had it implemented to just do the calculations on a monthly basis. But the problem you describe was pointed out.

That's why, as a result, I added in the notion of the sliding window calculation. This prevents the ability for the population to suddenly jump to a new cheap option and exploit it while it's cheap, until the next month where the cheap option moves.
The sliding window means that price changes will never jump in that way, and oscillate in a manner which enables the kind of exploitation you fear. It ends up being impossible to manipulate the system in that way.




Quote

You also under this system are basically encouraged to play the over under. You under bid with certain mechs and make specific meta builds out of unusual chasis (VERY DOABLE) so that you can free up more bv to do the super strong bv mechs (733c previously).

I'm not seeing how this is bad. You're right though, this is exactly how the system works.

The system doesn't let you just bring all of the most popular mechs (which is what happens today). Instead, you are forced to find other ways to fight efficiently.

Now, you claim that it's trivial to just make other mechs perform equally well to the top performers... But I have to say that's obviously untrue. Because if it was, then we'd see it in practice.. but we never do.


Quote

It just makes mech building the new meta..... Id rather not have to worry about this honestly myself since I honestly enjoy mech building and this would make it a grind....

I'm not sure what about it makes it into a grind.

I mean, you say you enjoy mech building... But if you were actually rewarded for making good, novel designs that other people hadn't made, somehow that makes it a grind?

I don't see how being rewarded for doing something makes doing that thing somehow less fun.

Saying "Mech design becomes the new Meta" is saying "Constantly changing meta will become the new meta".

That's the point here.. Creating a dynamic environment where the mechs seen are constantly evolving over time as players constantly adjust their configurations to maximize their efficiency for their allocated BV.

I'm thinking that there just must be some fundamental disconnect here, because I see this as exactly what we want to achieve.

View PostLukoi, on 06 March 2014 - 02:11 PM, said:

I think there should be occasions that don't require the use of a premium private match (PPM) where a team might be bringing all of the "optimal" mechs.

It's true that this system would prohibit that.

Indeed, it was specifically designed to prohibit it.

Potentially, you could have a separate Queue with unlimited BV.

However, the basis for this design really is to specifically prohibit what you describe there.. because the game will be more interesting if we see more mechs that just the 5 "good" ones.

And since your opponents have the same restrictions placed upon them, you're not putting yourself at a disadvantage... You can actually drive some of those other mechs, and still play competitively.

#43 Nik Van Rhijn

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,905 posts
  • LocationLost

Posted 06 March 2014 - 02:25 PM

While i can see that there are possible problems with Rolands (to my mind excellent) idea PGI could seed it with their existing data to give a starting condition. They could then put it into a patch and run it for 2 weeks. It is highly unllikely (to me) that it would be any worse than what we have now. If it does crash and burn then we have had 2 bad weeks. How often have we had much longer than that with bad balance changes?
There is no long term downside to trying it.
If used in conjunction with Elo buckets and giving new players a low Elo I think many people would be happy. Certainly I would like to see it.

#44 Cimarb

    Member

  • PipPipPipPipPipPipPipPipPip
  • Caladbolg
  • Caladbolg
  • 3,912 posts
  • Twitter: Link
  • Twitch: Link
  • LocationA hop, skip and jump from Terra

Posted 06 March 2014 - 04:24 PM

A system based upon popularity is extremely less efficient than one based upon statistics. I fully support a Battle Value system, but not the one you are proposing. I want a system where I can equip a mech with what works best for me and get a score immediately, on the fly, based upon what is equipped in it, not what hardpoints and chassis are popular at the time.

Anything that is based upon popularity is going to be behind the curve, as a Banshee has zero to begin with because obviously no one has played one up until it is in the game, and then a while after that to "build" one. A true Battle Value system, though, is based off of the statistics of the components, so the Banshee can be given a BV before it is even in game.

#45 Varent

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,393 posts
  • Facebook: Link
  • LocationWest Coast - United States

Posted 06 March 2014 - 05:16 PM

View PostCimarb, on 06 March 2014 - 04:24 PM, said:

A system based upon popularity is extremely less efficient than one based upon statistics. I fully support a Battle Value system, but not the one you are proposing. I want a system where I can equip a mech with what works best for me and get a score immediately, on the fly, based upon what is equipped in it, not what hardpoints and chassis are popular at the time.

Anything that is based upon popularity is going to be behind the curve, as a Banshee has zero to begin with because obviously no one has played one up until it is in the game, and then a while after that to "build" one. A true Battle Value system, though, is based off of the statistics of the components, so the Banshee can be given a BV before it is even in game.


the problem with the system is there are blatent loopholes in it that can be exposed and are encouraged for the sake of non meta users... wich I dont feel is correct either. Its just pushing an agenda.

#46 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 06 March 2014 - 05:17 PM

View PostCimarb, on 06 March 2014 - 04:24 PM, said:

A system based upon popularity is extremely less efficient than one based upon statistics. I fully support a Battle Value system, but not the one you are proposing. I want a system where I can equip a mech with what works best for me and get a score immediately, on the fly, based upon what is equipped in it, not what hardpoints and chassis are popular at the time.

The problem with your belief here is that you presuppose that you know what the "correct" value of a mech or weapon system is, and you don't.

What you are suggesting is that you fix the prices of everything, similar to the Soviets fixing the prices of consumer goods in their marketplace. The problem is, their prices did not truly reflect the actual value of things, and thus the economy didn't function.

The solution offered here is akin to a market economy, where prices automatically based on usage statistics, just like prices in a real-world free-market evolve based upon transactions. It's actually the most efficient means by which to establish prices.

Of course, if you magically knew all the correct prices for all mechs and weapons, you could just set them directly. However, you can't actually figure out what they are, because the interplay is too complex. So what will happen if you set a static table of BV, is that the metagame will shift to exploit the weapons and mechs which have prices set below their actual utility value. And what's worse, is that as the game evolves, those values would need to be reset constantly, not only in response to balancing changes, but also in response to new tactics and configurations developed by players.

That's why it's more efficient to derive BV's based on a statistical analysis of the playerbase's actions, which is what is being done here.

Quote

Anything that is based upon popularity is going to be behind the curve, as a Banshee has zero to begin with because obviously no one has played one up until it is in the game, and then a while after that to "build" one.

Sure, but it'll only be "behind the curve" for a short time.

A brand new mech could start out very cheap.. But by virtue of it being very cheap, it'll get used quite frequently (especially since it'll be brand new). This will have the result of its price increasing quite quickly, especially since changes at the edge of the price scale are accelerated due to squaring the sigma difference.

So even in the case you describe here, I would expect the new mech to quickly rise to its correct price point in a week or two, tops.

Quote

A true Battle Value system, though, is based off of the statistics of the components, so the Banshee can be given a BV before it is even in game.

But that Battlevalue that you assign to it will be wrong. It won't actually reflect its real utility in game.

With the market based system, the BV will automatically balance itself to the correct value, based upon the actual gameplay environment. With your system, it will likely NEVER achieve a correct value... because it'll just be a value that you picked out of thin air that isn't based upon how players would actually use it.

#47 GalaxyBluestar

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,748 posts
  • Location...

Posted 06 March 2014 - 05:43 PM

so how would this system handle clans?

could it make effective balance for 10vs12 and the diffirent weapon/equipment values?

if it could achieve this i'll give a yes vote.

#48 B0oN

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,870 posts

Posted 06 March 2014 - 05:48 PM

@ Varent ?

Have you ever met the JR7K with 2xErL´s+Adv-Zoom,UV,Arti and Airstrike ?
Runs around with 152.7kp/h, has a few JJ´s and nearly full armour (nearly an exact replica of the famous 6ML JR7) and is just a little longrange-scout that is able to do some serious things ...

Well, best part about the Jenner7-K is its 4 ( oh yeah 4 modules) module slots giving it enoug battlefield leverage to (nearly) counter the raw DPS of a 6ML Jenny,which is one of the main reasons I´m using it. It´s the mini RVN-3L (without ECM, but those moduleslots, man ^^) with JJ´s and full "lightmech" mach 1 (152.7 /w speedtweak) .

I have to confess that I really like it, and with the system Roland proposed even a change in the meta towards more "module-heavy" Mechs would equalize itself after some time, given usage .

You see, neither abusability nor robustness have been ditched in this system, one could go so far and more or less call it the "perpetuum mobile of balance" for MW:O.

Nifty system, take your time to think it through.

@Galaxy Bluestar :

I guess it doesn´t matter which side of the fence (IS or Clan) you are looking, heightened usage of Clanmateriel through community would equalize itself pretty quick in steeply rising "BV"-costs for these certain, often-fielded Clanassets, vice versa for IS.

Edited by Rad Hanzo, 06 March 2014 - 05:51 PM.


#49 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 06 March 2014 - 05:53 PM

View PostGalaxyBluestar, on 06 March 2014 - 05:43 PM, said:

so how would this system handle clans?

could it make effective balance for 10vs12 and the diffirent weapon/equipment values?

if it could achieve this i'll give a yes vote.

Honestly, since we no idea how the clans are actually going to be implemented or mixed with IS mechs, I honestly don't know.

I mean, technically, the system shouldn't care whether they were IS or Clan mechs... Also, if you segregate it such that clans are always allowed to drop fewer mechs, and that they were dropped on teams with only one type (clan or IS), then you could probably achieve the results you're looking for fairly easily, simply by making a separate market for clan mechs.

If you did that, then you'd have the clan mech prices determined independently from IS mech prices. Then, you could assign a separate PPBV, such that clan teams had less BV to work with, which would result in them being able to field less clan tech (i.e. the 10v12 thing). Alternatively, you could just use a different base value to calculate the PPBV for clan tech, which would have the effect of skewing their adjusted BV's upward.

So, yeah, I guess it could be adjusted to handle clan tech without too much trouble.

But I can't promise anything, without a more concrete explanation of how clan tech is going to be mixed into the actual game.

#50 Cimarb

    Member

  • PipPipPipPipPipPipPipPipPip
  • Caladbolg
  • Caladbolg
  • 3,912 posts
  • Twitter: Link
  • Twitch: Link
  • LocationA hop, skip and jump from Terra

Posted 06 March 2014 - 05:55 PM

View PostRoland, on 06 March 2014 - 05:17 PM, said:

The problem with your belief here is that you presuppose that you know what the "correct" value of a mech or weapon system is, and you don't.

"I" don't know the correct value, but "I" do know that the people that have developed the MUL/BV have spent decades of competitive time figuring it out. No, nothing is perfect, but tweaking some numbers in a database to fine tune things is much easier than trying to adjust an inflated popularity number.

View PostRoland, on 06 March 2014 - 05:17 PM, said:

What you are suggesting is that you fix the prices of everything, similar to the Soviets fixing the prices of consumer goods in their marketplace. The problem is, their prices did not truly reflect the actual value of things, and thus the economy didn't function.

The solution offered here is akin to a market economy, where prices automatically based on usage statistics, just like prices in a real-world free-market evolve based upon transactions. It's actually the most efficient means by which to establish prices.

Now you are the one presupposing that a free market is actually a good thing. There is a reason we have huge levels of poverty and amazingly rich 1%ers here due to that "free" market.

View PostRoland, on 06 March 2014 - 05:17 PM, said:

Of course, if you magically knew all the correct prices for all mechs and weapons, you could just set them directly. However, you can't actually figure out what they are, because the interplay is too complex. So what will happen if you set a static table of BV, is that the metagame will shift to exploit the weapons and mechs which have prices set below their actual utility value. And what's worse, is that as the game evolves, those values would need to be reset constantly, not only in response to balancing changes, but also in response to new tactics and configurations developed by players.

That's why it's more efficient to derive BV's based on a statistical analysis of the playerbase's actions, which is what is being done here.

Nothing would need to be reset when using component statistics, ever. All that needs adjusted would be individual components when their stats were changed. Say you have a ML that is worth 80 BV due to its damage, range, heat, etc. If you increase the heat slightly, you just recompute it's BV, using the already available tools for this very purpose, to get the new BV. It would be extremely easy to implement and manage.

The interplay between systems would only matter if it was something like an Artemis version, where it is not a component in itself, but instead an addon to an existing system. This, and non-tangibles like ECM, would take the most adjusting, but still drastically less than you would think.

View PostRoland, on 06 March 2014 - 05:17 PM, said:


Sure, but it'll only be "behind the curve" for a short time.

A brand new mech could start out very cheap.. But by virtue of it being very cheap, it'll get used quite frequently (especially since it'll be brand new). This will have the result of its price increasing quite quickly, especially since changes at the edge of the price scale are accelerated due to squaring the sigma difference.

So even in the case you describe here, I would expect the new mech to quickly rise to its correct price point in a week or two, tops.

So it works just like Elo. I am not a fan, sorry.

View PostRoland, on 06 March 2014 - 05:17 PM, said:

But that Battlevalue that you assign to it will be wrong. It won't actually reflect its real utility in game.

With the market based system, the BV will automatically balance itself to the correct value, based upon the actual gameplay environment. With your system, it will likely NEVER achieve a correct value... because it'll just be a value that you picked out of thin air that isn't based upon how players would actually use it.

It's not picked out of thin air, though. It is based off of the actual qualities that define the weapon: damage, RoF, heat, range, etc. If you want an idea of how this would work, check out http://www.heavymeta...om/bv_calc.htm. It doesn't take damage delivery (spread, duration or FLD) into account, so it would need some adjusting, but it gives you an idea of how the system works.

#51 Cimarb

    Member

  • PipPipPipPipPipPipPipPipPip
  • Caladbolg
  • Caladbolg
  • 3,912 posts
  • Twitter: Link
  • Twitch: Link
  • LocationA hop, skip and jump from Terra

Posted 06 March 2014 - 06:00 PM

View PostRoland, on 06 March 2014 - 05:53 PM, said:

Honestly, since we no idea how the clans are actually going to be implemented or mixed with IS mechs, I honestly don't know.

I mean, technically, the system shouldn't care whether they were IS or Clan mechs... Also, if you segregate it such that clans are always allowed to drop fewer mechs, and that they were dropped on teams with only one type (clan or IS), then you could probably achieve the results you're looking for fairly easily, simply by making a separate market for clan mechs.

If you did that, then you'd have the clan mech prices determined independently from IS mech prices. Then, you could assign a separate PPBV, such that clan teams had less BV to work with, which would result in them being able to field less clan tech (i.e. the 10v12 thing). Alternatively, you could just use a different base value to calculate the PPBV for clan tech, which would have the effect of skewing their adjusted BV's upward.

So, yeah, I guess it could be adjusted to handle clan tech without too much trouble.

But I can't promise anything, without a more concrete explanation of how clan tech is going to be mixed into the actual game.

Or, if you used actual BV, Clan mechs would be measured exactly like IS mechs and you wouldn't have to use any segregation or separate system.

#52 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 06 March 2014 - 06:22 PM

Quote

I don't know the correct value, but I do know that the people that have developed the MUL/BV have spent decades of competitive time figuring it out.

What does that tell you?
I mean, I'm sure you understand that the BV from Tabletop is totally unapplicable to Mechwarrior, given the dramatic differences in the game... So you can't actually use any of those numbers which were developed over decades.

The fact that trying to arrive at a static BV table for TT took decades of tweaks should indicate how futile an effort it is... and how badly it would fail here.

It's like trying to fix prices in a real world market. It doesn't work well.

Quote

No, nothing is perfect, but tweaking some numbers in a database to fine tune things is much easier than trying to adjust an inflated popularity number.

But.. no. It's not. That's the point.
You would have to adjust those BV numbers constantly... over and over again. Weapons change? ALL the BV numbers change.. not just the weapons that changed, but all the mechs... and really, all of the other weapons that didn't change, because their utility is now modified by virtue of the fact they are used in combination with the weapons that DID change.

And the player base is going to constantly chip away at it, and find the parts of the pricing table that are incorrect.

But the system presented here doesn't require you to adjust anything. It's MAXIMALLY easy. You have to do literally nothing at all. It will automatically adjust ITSELF. That's the whole point.

You seem to be really focused on the term "popularity" here.. What we're doing here is a statistical analysis of player data, in order to identify usage trends, and translate them into a price. This is infinitely more scientific that simply pulling BV's out of a hat.

Quote

Now you are the one presupposing that a free market is actually a good thing. There is a reason we have huge levels of poverty and amazingly rich 1%ers here due to that "free" market.

In this market system, you really don't have to worry about that. Given that players all have exactly the same amount of influence over the pricing mechanisms here, in that it is purely based upon what you choose to drop in, there is no mechanism by which any player could unfairly influence the system.

That is, unlike the real world market, where certain people have more money and thus more influence in the market, in the system presented here all players would have exactly the same influence, in that their influence stems solely from playing the game and dropping in mechs. I guess, technically, people who drop more often would influence the price more, but it would be by such a small degree as to be trivial.

Quote

Nothing would need to be reset when using component statistics, ever. All that needs adjusted would be individual components when their stats were changed. Say you have a ML that is worth 80 BV due to its damage, range, heat, etc. If you increase the heat slightly, you just recompute it's BV, using the already available tools for this very purpose, to get the new BV. It would be extremely easy to implement and manage.

No, as I pointed out above, your calculation would be wrong. You'd be missing huge components of the calculation, because its usage in the game is much more complex than you understand it to be.

This is why price fixing in economies is demonstrably inefficient, and why markets are proven to be the most efficient means by which to determine price.

Quote

So it works just like Elo. I am not a fan, sorry.

No, It doesn't really work like Elo at all.

I realize that the description is quite lengthy, but you may want to examine it in a bit more detail to get a better understanding of what is being proposed.

The only similarity to Elo is that it involves a number being derived mathematically.

Quote

It's not picked out of thin air, though. It is based off of the actual qualities that define the weapon: damage, RoF, heat, range, etc.

But you don't know how those things should properly translate into BV.

What you are suggesting is that intrinsic properies of a weapon or mech or whatever can be computed and translated into a singlular BV. While this may seem simple at first glance, the reality is that it runs into two major problems... and honestly, the problems are fairly well understood within the field of economics, as derivation of price is an important concept there.

FIrst, when you look at a calculator like the one you linked to (the link is broken by the way, as it accidentally included the period from the sentence), it appears very scientific. You put in numbers, it does calculations, and it spits out other numbers.

However, under the covers, at the heart of it, a few things have been hard-coded into the system... Specifically, scalar values representing certain core efficiencies, and relationships between different values. These two things have been arbitrarily set, likely though a process of trial and error. That's the only way to try and arrive at those values, because there's no real scientific principles to guide such things. And when those numbers are set incorrectly, then the resulting calculations are wrong. BV in Battletech was notorious for exactly this for years, which was why it required revisions.

The second critical problem is that it fails to account for the complex interplay of mechs when used by human players. It doesn't account for utility that emerges from combining all of the gameplay elements together.

Thus, when simple static values are applied to a complex game, a large playerbase will tend to identify the places where the BV doesn't reflect the REAL utility.. and then they will exploit it. Effectively, those weapons and mechs which have had their value set lower than it should be (and this ALWAYS happens, in every single instance of such a system, ever), and players will flock to them... Exactly what happens in an economy when a commodity's price is fixed below what the market derived price is, and everyone buys it, and you have commodity shortages.

In the dynamically derived price based on statistical analysis, this problem goes away. Because as soon as the player base identifies a mech which is priced below its actual utility value, and they start to use it more as a result... the price goes up. That's why markets are the most efficient means by which to derive price.

Quote

Or, if you used actual BV, Clan mechs would be measured exactly like IS mechs and you wouldn't have to use any segregation or separate system.

Except that you don't have actual BV values to use. You have no idea what the value of a madcat would be, or a cataphract, or an atlas.

You're basically saying, "If we had magically derived BV's which perfectly represented the utility of a mech at any point in time, then they would be perfect, and we should use them."

Sure.. But such a thing doesn't exist. It'll never exist.

Edited by Roland, 06 March 2014 - 06:30 PM.


#53 Amsro

    Member

  • PipPipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 3,437 posts
  • LocationCharging my Gauss Rifle

Posted 06 March 2014 - 07:29 PM

Quite a long read, well thought out and it seems to allow dynamic adjustment after balance tweaks.

I dig it.

#54 Macksheen

    Member

  • PipPipPipPipPipPipPipPip
  • Urban Commando
  • Urban Commando
  • 1,166 posts
  • LocationNorth Cackalacky

Posted 06 March 2014 - 07:38 PM

Perhaps one that didn't just rely on chassis use, but also a composite of weapon load-out.

#55 Cimarb

    Member

  • PipPipPipPipPipPipPipPipPip
  • Caladbolg
  • Caladbolg
  • 3,912 posts
  • Twitter: Link
  • Twitch: Link
  • LocationA hop, skip and jump from Terra

Posted 06 March 2014 - 08:54 PM

View PostMacksheen, on 06 March 2014 - 07:38 PM, said:

Perhaps one that didn't just rely on chassis use, but also a composite of weapon load-out.

I agree. At the very least, the popularity aspect needs to at least go down to the loadout level, not just the chassis.

#56 Lukoi Banacek

    Member

  • PipPipPipPipPipPipPipPipPip
  • WC 2018 Top 12 Qualifier
  • WC 2018 Top 12 Qualifier
  • 4,353 posts

Posted 06 March 2014 - 09:05 PM

View PostRoland, on 06 March 2014 - 02:22 PM, said:

It's true that this system would prohibit that.

Indeed, it was specifically designed to prohibit it.

Potentially, you could have a separate Queue with unlimited BV.

However, the basis for this design really is to specifically prohibit what you describe there.. because the game will be more interesting if we see more mechs that just the 5 "good" ones.

And since your opponents have the same restrictions placed upon them, you're not putting yourself at a disadvantage... You can actually drive some of those other mechs, and still play competitively.


Your comments here simply cement why I would not vote for this option. There are times/situations within a CW construct where the availability of optimal mechs should be allowed to occur...tying your system to a player driven economy (that includes production, salvage, raiding, commerce etc much like in NBT-league play) would encourage what I'm talking about. Your straight metrics-driven BV system, while admirable, would prevent it in ALL cases, so I would vote no.

I'm not a fan of the current system, but I honestly don't see enough of an improvement in spending the time and effort by PGI to jump to yours "as is" either.

Good thoughts though Roland. Maybe some form of it gets integrated into a broader system in the future, but hopefully with the understanding that sometimes asymetrical aspects within a PvP/CW construct are actually optimal (and more fun in the long run for many).

#57 Rebas Kradd

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,969 posts

Posted 06 March 2014 - 09:06 PM

Certainly a lot of work put into this.

I read on a dev tweet somewhere that there are voices within PGI that want to use BattleValue somehow, so perhaps this proposal will at least get its day in court.

#58 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 06 March 2014 - 09:17 PM

View PostLukoi, on 06 March 2014 - 09:05 PM, said:


Your comments here simply cement why I would not vote for this option. There are times/situations within a CW construct where the availability of optimal mechs should be allowed to occur...tying your system to a player driven economy (that includes production, salvage, raiding, commerce etc much like in NBT-league play) would encourage what I'm talking about. Your straight metrics-driven BV system, while admirable, would prevent it in ALL cases, so I would vote no.

Ah, don't get me wrong, there are certainly all kinds of extra situations where you could simply choose not to use this system. As you describe, there were planetary assaults back in the MW4 leagues where we certainly weren't bring garbage chassis to the field..

Although, at the same time, one of the most fun PA's we engaged in was when LA handed the defense of an important planet over to us with the simple orders of, "Just make them bleed as much as possible before they take it." LA had already suffered heavy losses, and had used all the good mechs in the planet's garrison.

But we ended up pushing all the way back through the entire PA against Wolf, using nothing but third string garbage tier IS mechs (things like Hellspawn and Chimeras) against top tier front line clan mechs. And frankly, it was pretty fun making the "trash mechs" work. ;)

Regardless, I understand your concern here. Within the context of a larger CW system, who knows what kind of thing could be used... Presumably, a much more complex framework would surround drops. In those cases, instead of having teams simply assigned a BV, you could potentially implement a bidding system. For the clans, especially, this could work well.

Ultimately, for what you describe, you could simply take the BV system, and then replace the team composition part of it with something else.

I focused on how it could be applied to our current game though, since CW isn't really defined in any concrete fashion yet.

Quote

I'm not a fan of the current system, but I honestly don't see enough of an improvement in spending the time and effort by PGI to jump to yours "as is" either.

Well, part of the reason I laid this out was to show that it really shouldn't take a ton of effort to implement. Certainly, the actual calculation part of it is trivial to implement. I laid it out in a spreadsheet in a few hours, including the time it took to seed the data and revamp it as I changed things around a few times.

The only work that would be involved would be simply tying it into the UI such that it summed up individual BV's prior to enabling the launch button for a team, and a display of the current BV price list.

#59 Lyoto Machida

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 5,081 posts

Posted 06 March 2014 - 09:30 PM

View PostRoland, on 06 March 2014 - 02:22 PM, said:

Varent and I have been discussing this in PM's, but I figure that it's worth explaining this publicly since others may be under the same mistaken belief.

Oh, without question, this system would result in the overall Jenner population being more evenly distributed across the different variants... But that's good. That's kind of the point of the system.

It still won't result in any of them being cheap, because those Jenners will still each be very good chassis.

Your fear here is that, since there are FOUR Jenner, that they are going to be able to somehow split their cost across all of them. But this is ok, because all the other variants of other mechs also exist in multitudes. The only way those jenners would become cheap would be if their usage actually dropped below the average drop rate for all variants of all mechs... Meaning that you'd be seeing the same number of them as you do mechs like the Raven 2X.

Again, I think part of the problem you're having is that you're hyper focusing on a small part of the overall market, and then forgetting that it's part of the overall market when BV values are calculated.

In the scenario you imagine, you're imagining that there will be one expensive Jenner, and the others will be real cheap... then everyone will just move to one of the other Jenners, and then another, then another..

But this can't actually work. In the real system, that's not what happens.

In the real system, what will happen is the usage will equally distribute across all of them at the same time.. because doing otherwise would involve intentionally spending more BV than necessary. Because as soon as you start using that K, it's price starts going up. That price movement is slow and gradual, but it happens and does so at the same time the first Jenner's price is going down. As soon as that starts happening, in the imagined scenario where all the other Jenners have a price of zero or whatever... then they are immediately cheaper than the K now (and much cheaper than the original Jenner everyone was using). So people aren't going to all migrate to the K. They're going to distribute across all the Jenners (if they can in fact all do what they want), because otherwise they'll be spending more BV than they need to.

As I mentioned previously, I specifically considered exactly the situation you are worried about here, because originally I had it implemented to just do the calculations on a monthly basis. But the problem you describe was pointed out.

That's why, as a result, I added in the notion of the sliding window calculation. This prevents the ability for the population to suddenly jump to a new cheap option and exploit it while it's cheap, until the next month where the cheap option moves.
The sliding window means that price changes will never jump in that way, and oscillate in a manner which enables the kind of exploitation you fear. It ends up being impossible to manipulate the system in that way.





I'm not seeing how this is bad. You're right though, this is exactly how the system works.

The system doesn't let you just bring all of the most popular mechs (which is what happens today). Instead, you are forced to find other ways to fight efficiently.

Now, you claim that it's trivial to just make other mechs perform equally well to the top performers... But I have to say that's obviously untrue. Because if it was, then we'd see it in practice.. but we never do.



I'm not sure what about it makes it into a grind.

I mean, you say you enjoy mech building... But if you were actually rewarded for making good, novel designs that other people hadn't made, somehow that makes it a grind?

I don't see how being rewarded for doing something makes doing that thing somehow less fun.

Saying "Mech design becomes the new Meta" is saying "Constantly changing meta will become the new meta".

That's the point here.. Creating a dynamic environment where the mechs seen are constantly evolving over time as players constantly adjust their configurations to maximize their efficiency for their allocated BV.

I'm thinking that there just must be some fundamental disconnect here, because I see this as exactly what we want to achieve.


It's true that this system would prohibit that.

Indeed, it was specifically designed to prohibit it.

Potentially, you could have a separate Queue with unlimited BV.

However, the basis for this design really is to specifically prohibit what you describe there.. because the game will be more interesting if we see more mechs that just the 5 "good" ones.

And since your opponents have the same restrictions placed upon them, you're not putting yourself at a disadvantage... You can actually drive some of those other mechs, and still play competitively.


This is what I envisioned when I read this post. Let's go with your example of Jenners...imagine each Jenner is a road or freeway or whatever. Sure, everyone is playing the Founder's Jenner or whatever and the usage is high for that particular variant (ie: the road is clogged with rush hour type traffic). The drop value for that Founder's Jenner is gonna be higher than the other two due to excessive usage. The other two variants (or roads) are gonna be lightly used and so their value is relatively lower. As drivers (pilots) find out about these other routes (the two lesser used variants with cheaper BV cost), some will utilize one of the lesser two variants (less clogged roads) and the total use of the variants (traffic flow) will alleviate and even itself out.

Basically, at some point, users in game will notice the lower BV on the underutilized variants and take advantage...the only way for this to happen without a total increase in Jenner drops is for it to come from drops that would have happened in the Founder's variant (thus evening things out without any sort of nerfs). If people continue to run the popular variants, it will cost their teams more to bring those specific mechs.

I've seen this idea (variant popularity) tied to how much a mech will cost to purchase or repair in CW but not for BV...great thread and idea so far.

#60 GalaxyBluestar

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,748 posts
  • Location...

Posted 06 March 2014 - 09:32 PM

voting yes. anything that helps the clan player immersion gets my vote.





3 user(s) are reading this topic

0 members, 3 guests, 0 anonymous users