35 replies to this topic

[Karl Streiger]

WO seems impossible. There are billions of combinations and every little change can have a dozen consequences.
Without a major change in game-mechanics, there is no way the progress of seasonal nerfs and buffs will ever stop.

However recalling the latest discussions about FW and the KDK-3 I realized that this community had already found the solution 3 ½ years ago.
During the reign of the PinPoint Poptarter, prior to the Ghost Heat development, the idea was born to create an incorruptible dynamical "BattleValue".

Independent other community members came up with this idea in the following years until it was successfully “field tested” within the MRBC and BWO Leagues.

In this thread I want to show you again the ageless beauty of this concept.
In a second part I want to show how this system influence the Game Modes and how the Game modes might need to change to support this system. I think terms like “Waves” and “StockMechs” is something that some people might want to see in this game and the new system can provide both without to corrupt the experience for casuals or competitive players.

---

Poll
Calculation (Q1)
Changes for the Game (Q2 & Q3)

---

In this thread I want to show you again the ageless beauty of this concept.
In a second part I want to show how this system influence the Game Modes and how the Game modes might need to change to support this system. I think terms like “Waves” and “StockMechs” is something that some people might want to see in this game and the new system can provide both without to corrupt the experience for casuals or competitive players.

Edited by Karl Streiger, 16 December 2016 - 12:37 AM.

[Karl Streiger]

The idea behind is to take the existing server data about Mech/Variant usage and take the simple assumption things that are good are taken more often. And what is taken often comes at a price. A simple truth of economics.

---

The name of the game? I don’t have a good name yet. Battle or Combat Value can be mistaken. I would like to hear your suggestions. I might run the name Combat Price at first.

---

The very principle of this balancing mechanism is not to balance the numbers and weapons and mechs but their usage.
The advantage is obviously that there are no tests needed, no additional development of Energy Draw, Quirks, Squirks, Heat Scale, Heat ramps or whatever you can imagine.
You can even bring a Rotary AC 20 or Hyper Assault Gauss to a game even when they are corrupt and totally imbalanced. The Combat Price will deal with them.

All PGI need to do this is a database script that calculate some values:
Here is a quick except how it might look like. (Numbers are set at random)

It's a pivot table of the Mech/Variants and the equipment used by those Mechs.

• The first value is the usage (I think It could be interesting to multiple the value with the number of mounts, did a Mech mount 1,2 or 12 of a kind has some influence)
  
• Second value is the W/L ratio of this combination
  
• Third the average MS (I would rather sort all MS and create a median, but maybe average is adequate enough)

When you sum up all the values in each row you get the total usage value of a Mech.



When you sum all the values in a column you get the usage of this equipment part.

A final value could look like (SUM(Usage) * AVG(MS) * AVG(W/L)) ??? -
Maybe equipment need to be divided by a constant.

When you create a Mech in a MechLab those values are individual calculated for your Mech. Resulting in a “Combat Price”.

Edited by Karl Streiger, 16 December 2016 - 12:32 AM.

[ice trey]

FP: Easy changes
1) Do not balance by tonnage, but points cost
2) Points cost is tonnage * Frequency modifier
3) Frequency modifier is based on how often the community use a given chassis in FP. The more overused, the more expensive it gets.
4) Frequency Modifer only applies to stats from two updates ago, in order to keep recent mechanics changes in mind. (No use gimping people with Highlanders and Victors when Poptarting has long since died)

Now stuff that too strong costs more. Stuff that gets under-used is cheaper.

Edited by ice trey, 16 December 2016 - 12:38 AM.

[Karl Streiger]

As you can see it is indeed very very simple - and it is always accurate and can be calculated every day or with every patch anew.

Changes for QP:
I think it is obvious that the changes need to put into the Match Maker, maybe instead of the PSR. If MWO ever gain enough numbers again, PSR could be used again - or at least when the calcul
So players can run almost any build and find enemys and team members that run equal builds.

Changes for FW.
Primary Invasion mode - first this mode need two changes.

• First : increase the max size of the Drop Deck for each player towards 12 Mechs. (yes 12 Mechs) - the sequece is locked (the left one is the first, then the second….and so on)
  
• Second: Change the modus of destruction into capture.

For each player there is a number of Combat Price Points available. It might be as high as the most expensive Mech in the Game.

You choose a Mech and the points are removed from your FP Battle account. You can take the “best” mech right from the start. Or you decide to take a worse mech, but are able to keep some CPPs on your account.

CPPs can be earned in combat by participating in the battle, similar to MS.
When your Mech is killed you “buy” the next drop. When you don’t have enough CPPs on your account the battle is over for you.
So the “drop sequence” becomes very important.

I think you can imagine how interesting battles could become. Run a Deck with 12 Vindicators or drop with the 4 UAC10 KDK-3 right from the start - and have 4 additional KDK3s in reserve.

(Oh yes 4 UAC10 KDK - Ghost Heat, Quirks and ED is not longer needed)

You see the Wave Idea?
You might push with ease through the first defenders but than you need to hold your ground. Fighting of waves of cheap attackers / defenders. But If they focus on a single target this might get down without earning enough CPPs to get another "life" decreasing the survival rate for the others.

[Duke Nedo]

You can balance organized matches like that, but you don't make all mechs viable. You'll just leave all bad mechs behind to rot, right? The point with balance is to make all mech variants in the game enjoyable to play.

It's one player behind every mech after all.

[Karl Streiger]

Duke Nedo, on 16 December 2016 - 01:03 AM, said:

You can balance organized matches like that, but you don't make all mechs viable. You'll just leave all bad mechs behind to rot, right? The point with balance is to make all mech variants in the game enjoyable to play.

It's one player behind every mech after all.

No, not to rot but to keep them in their own pool.

I don't dare not to say that a group with vindicators couldn't be successful but for the most players it won't touch of.
But instead matching the Vindicator vs Kodiaks - they are in their own pool
The Vindi fight panthers and Ice Ferrets and Gargoyles but no Kodiaks at all.

Of course we can hope that someday PGI add some 150% quirks on the Vindi and keep him for 30days on par with the KDK

Edited by Karl Streiger, 16 December 2016 - 01:12 AM.

[Kaeb Odellas]

Why leave bad robots bad and force them to fight other bad robots? Why not just make all the robots good?

[El Bandito]

Kaeb Odellas, on 16 December 2016 - 01:50 AM, said:

Why leave bad robots bad and force them to fight other bad robots? Why not just make all the robots good?

How would you bring the Spider-5V to Kodiak-3 level?

[Elessar]

The idea of the dynamic battle value definitely sounds interesting, especially since it also includes the W/L-Ratio.
(I definitely think it should be named a "value" btw. and not a "price")

An important thing, however, is to restrict the time frame for the calculation. You should only take the W/L-Ratio of the last 1-2 months into account for the calculation (value of weapons/mech combinations may undergo changes ... what worked before may not work today (for example due to changes in game mechanics, or due to changes in enemies the mech faces ... therefore you wouldn't want to have past successes of a combination overshadow the combinations failures in current game (or vice versa))

[Kaeb Odellas]

El Bandito, on 16 December 2016 - 01:57 AM, said:

How would you bring the Spider-5V to Kodiak-3 level?

Give it more hardpoints. Give it mid-air jumpjet maneuvering. Give it a wider torso pitch angle. Make it not losing speed on landing. Adjust its engine curve. All of these would probably be easier to implement than creating an entire self-correcting BV system.

[Karl Streiger]

Kaeb Odellas, on 16 December 2016 - 02:54 AM, said:

Give it more hardpoints. Give it mid-air jumpjet maneuvering. Give it a wider torso pitch angle. Make it not losing speed on landing. Adjust its engine curve. All of these would probably be easier to implement than creating an entire self-correcting BV system.

Easier?
Serious change ingame mechanics that fiddle with the engine, server hit registration, and dozen of other things is simpler but to run some data mining on server logs?

The data is already there - look into your stats - you have weapon logs, and mech logs...how difficult could it be to create a delta of Mech and Weapons?

Instead of reinventing the wheel development time can be used for different content - like moving on in time line and creating new equipment. The equipment don't need to be balanced by anything but feel. Because the usage and W/L will place it.

---

Gotten some more input - instead of pools the MM could also just match the same combat values for the teams.
So thunderbolts and Kodiaks drop together but are divided more equally.

as

Elessar, on 16 December 2016 - 02:25 AM, said:

The idea of the dynamic battle value definitely sounds interesting, especially since it also includes the W/L-Ratio.
(I definitely think it should be named a "value" btw. and not a "price")

An important thing, however, is to restrict the time frame for the calculation. You should only take the W/L-Ratio of the last 1-2 months into account for the calculation (value of weapons/mech combinations may undergo changes ... what worked before may not work today (for example due to changes in game mechanics, or due to changes in enemies the mech faces ... therefore you wouldn't want to have past successes of a combination overshadow the combinations failures in current game (or vice versa))

of course
it wouldn't make sense to collect today all the data for the last 3 years and put them into account.
You might start to collect data with the last patch.

The calculation for the Mech happens in the MechLab. You should see how the mounting of either LBX10 or AC 10 change this value.

However when compared to the TT BattleValue there will not a sweet spot - because people might think they have found a good Mech Variant that have a lower CV than normal but with superb combat performance this might change soon enough. When the number of usages of this combination is calculated

If those changes happens however - your mech is not nerfed. Your guns behave the same, your turn rate is not influenced, nor speed nor burn duration - everything stays the same - only that the CV marks your version a little bit more expensive.

Edited by Karl Streiger, 16 December 2016 - 03:20 AM.

[FupDup]

One of my concerns with BV systems based on usage rates is that lightweight filler weapons might end up having unfairly high BV simply because they're easy to mount. Statistically, you're going to find a lot more mechs equipped with Medium Lasers than Large Lasers. More Machine Guns than AC/20's. Etc.

This might also cause similar issues in the mech dimension if tonnage limits were retained, since TL reduce the frequency of the largest mechs (therefore reducing their BV) even though those big mechs are technically better than the middleweight mechs being used in strict TL conditions.


Well, that and the general concern of "why keep bad things bad" that other people have mentioned.

[Karl Streiger]

I must be the last person that don't want to bring the bad inline with the good - as it was promised.
But this would mean a 100% complete redesign of the whole game and every new Mech can destroy the fragile balance.

The fillers are indeed an issue. Cosidering that as part of the stats damage numbers are tracked either maybe the usage can be made more relative by comparing the total damage dealt in this battle, and how much damage each weapon has dealt.

For example the classic LRM 15 boat with MLAS support might have dealt 400dmg with LRMs and 140dmg with the MLAS
So the LRM usage is 75% and the MLAS 25%.

I wouldn't keep tonnage as a limiter in case of a combat value

[Lily from animove]

This sounds like it would mess up even more.

a current issue in the balance are the lower bottom skill players in FW and their distribution. these people hardly have an idea what they do. Now if you force them to regulary change mechs because your point system says so, whos the winner/loser of said system?

Loser:
newbies: with not enough own mechs fully stuffed and leveled.
Noobs: who suddenly have to pilot with bad skills in bad mechs.

the winners:
people who hoard chassis
people who know how the game works, they will suck less in bad mechs and therefore stil outperform the noobs and newbies

What you do is divide the community further between those who can adapt to the new rules by skill and their available emchs and grinded skills for the mechs from those not having them. That system would favour elitism to a scray degree where it would probably chase away a laod of people from CW once the new rules scrap over their owned mechs they have available.
in fact it would probably make thinsg even worse, think about 100 noobs badly piloting a kodiak, whole 10 people are piloting it good, the average score may eve start to lower the kodiaks costs, and soon you see premades roflstomp around because a lot noobs destroy the average of the mechs performance. That happens because your idea implies that a newb/noob in a good mech is automatically perfoming alos better, which hardly is the case.
Further, you give the community tools to exploit and manipulate your system, because irrelevant matches can be manipulated. Imagine a group like MS willingly doing extremely bad in some chassis in every unimportant game just to lower the costs of the chassis. And then when it matters BOOOM, roll with that chassis in tryhard mode because its actually performing well and cheap.

your idea would only work in a very idela setup where everyone has similar skills, and goes full tryhard mode all the time and not trying to abuse the system.

Sry Karl, but your idea has too many basic flaws to work out as well as you hope.

Edited by Lily from animove, 16 December 2016 - 04:48 AM.

[Karl Streiger]

OK this is also a possibility to see stuff.

The newb filter is already placed with the W/L although I think for all chassis it would be around 1. Because there are 2 less experienced mech jocks for every veteran in this chassis. But in the end its still used 3 times because its a good chassis.

I think you think that some premades can have any real impact on a mech. Oh look even when they are able to drive a W/L of 0.2 on a KDK they still have used it a hundred thousand times to do so, because all the other kDK player don't stop using it. And in the end its still a chassis that was used a hundred thousend times. So in the end, they might have the contrary result.

Also, such behavior has to activate trigger it might need some acknowledgment from a staff member. When he sees player names and performance far below average he only need to kick those values from the data and he is finished.

This could also be done automatically, when you have a non corrupted data set (not known by anybody that the collecting of data did has started) you can always compare current outcomes with the saved one. Is the margin to big the new data is rejected.

---

The goal is not to force anybody to run a Mech with a loadout he does not like. You can not game the system. Ok maybe you could with the above formula, but I don't have the data logs to mine them for a better one. Usage is the key indicator, match score and W/L are only "helpers" to add the right weight to the numbers.

Seriously - your favorite build is now 2% more expensive? And? Its not that it behave 2% worser in game it is exactly the same mech.
You don't need to chase any meta, you can and you will when you want to play competitive - because in this case the goal is the win.

Edited by Karl Streiger, 16 December 2016 - 05:08 AM.

[davoodoo]

I dont like idea of w/l and popularity dictating value of the mech.
Basically the better you do the worse your option becomes and it forces you to switch mechs up.

Not to mention that you can skew popularity with new releases like idk, mad 2c being in every match.

Have fixed values for each and tune them periodically if needed but dont do it automatically.

Edited by davoodoo, 16 December 2016 - 05:36 AM.

[jss78]

El Bandito, on 16 December 2016 - 01:57 AM, said:

How would you bring the Spider-5V to Kodiak-3 level?

Good point. IMO you SHOULD try to balance 'mechs, but it's a naïve expectation that it could ever be fully achieved.

The best I can tell, the worst offenders are generally the 'mechs which are relatively slow in their weight brackets. Take the Panther and Vindicator -- these are fire support 'mechs, basically Warhammers and Marauders for people who cannot afford Warhammers and Marauders.

They are great 'mechs in TT when cost is factored in. But in MWO they have literally zero reason to exist when we're all, in effect, infinitely rich.

You could shoehorn them into the light 'mech role (give them ludicrous engine caps) or give them massive offensive quirks, but let's me realistic, they'll never challenge top-tier striker lights and fire support heavies.

[Lily from animove]

Karl Streiger, on 16 December 2016 - 05:02 AM, said:

OK this is also a possibility to see stuff.

The newb filter is already placed with the W/L although I think for all chassis it would be around 1. Because there are 2 less experienced mech jocks for every veteran in this chassis. But in the end its still used 3 times because its a good chassis.

I think you think that some premades can have any real impact on a mech. Oh look even when they are able to drive a W/L of 0.2 on a KDK they still have used it a hundred thousand times to do so, because all the other kDK player don't stop using it. And in the end its still a chassis that was used a hundred thousend times. So in the end, they might have the contrary result.

Also, such behavior has to activate trigger it might need some acknowledgment from a staff member. When he sees player names and performance far below average he only need to kick those values from the data and he is finished.

This could also be done automatically, when you have a non corrupted data set (not known by anybody that the collecting of data did has started) you can always compare current outcomes with the saved one. Is the margin to big the new data is rejected.

---

The goal is not to force anybody to run a Mech with a loadout he does not like. You can not game the system. Ok maybe you could with the above formula, but I don't have the data logs to mine them for a better one. Usage is the key indicator, match score and W/L are only "helpers" to add the right weight to the numbers.

Seriously - your favorite build is now 2% more expensive? And? Its not that it behave 2% worser in game it is exactly the same mech.
You don't need to chase any meta, you can and you will when you want to play competitive - because in this case the goal is the win.

oh Karl, it's too much friday to explain you why many of the stated things here are wrong and not as you expect them to be untis shifted entire FW gameplay, of course they can and will shift emch stats too.

and just because you dont want to force people, doesn't meas your concept won't forse people, what forces people will be the price and those 2% can be exactly the difference.

and at the moment where you think someoen cna erase specific groups for sets of date because he expects "abuse" then we are already off an automated system and end not even better than with manual work based on a few people's perception.

thatw as the short version, the regular one would be like WALL of text that trump would get wet.

Edited by Lily from animove, 16 December 2016 - 06:51 AM.

[Karl Streiger]

Ok to set the margin we are talking about
A different Battlevalue might influence the pool you are dropping in or as an alternative the value of your mech for the MM.

In case of the wave invasion idea it's a cost for getting another drop.
In all those cases a small drop or rise will not affect anything

[Lily from animove]

Karl Streiger, on 16 December 2016 - 08:28 AM, said:

Ok to set the margin we are talking about
A different Battlevalue might influence the pool you are dropping in or as an alternative the value of your mech for the MM.

In case of the wave invasion idea it's a cost for getting another drop.
In all those cases a small drop or rise will not affect anything

if it won't affect anythign the entire idea is pointless because it won't have effect xD
Also, you would basically create a MM in FW which is giving people ways to abuse it.