12 replies to this topic

[ClaymoreReIIik]

Hello folks,

there have been some fruitfull attempts in the german community to get tournaments in MW:O going and to keep them fun.

I am pretty sure the devs are not exactly monitoring all the foreign language boards here for ideas on what they could do and since the subject is a rather complex matter, the online translators reach their limits quickly.

The Matchmaker itself is often topic of discussion and a lot of people I play with and talk to have the opinion that the current implementation of the matchmaker does not do a good job of creating even matches.

My personal opinion is that ELO does a good job of judging a pilot. While the current implementation does not stop me from playing with people that have a different "pilot skill" then I do (it just makes me wait longer to play with them), it does a rather poor job at creating balanced games.

The reason for this lies, in my opinion, in the part of this game that makes it drastically different then other shooting games. The weapon you bring matters. A lot.

I often tell my team that success (wins) in MWO depend on 3 things: Teamwork, Individual Skill, Mech Lab choices.

The matchmaker metric ELO to some extend factors in Teamwork and Individual Skill in the fact that it tries to match Groups and assigns a "Skill Value" to the player. What it totally ignores right now is "Mechlab Choice" part of my equation.

A lot of people call for "weight class matching" or "tonnage matching" as an approach for this, but I think it is too simplistic and using that as a metric part would lead to even worse matches.

Strip a Jenner of all of its armor, weapons and equipment and fit the smallest engine and drop into a match. You will still be a "light", you will still be worth 35 tons. You will still have a high elo as a light player, but your contribution to the team will not be worth much.

The battletech board game had a solution for this called BattleValue. Yes the BV could be gamed. Yes it was never perfect. BUT: It was a lot more accurate then "Tonnage" or "Weight Class" and it was applicable to each and every mech build there was in the tabletop.

The german community made an attempt at creating a "MWO Battle Value" short MBV (This is the old forum thread where an old version was presented:http://mwomercs.com/...__fromsearch__1). The results were really good and tried and tested in tournaments that were balanced through it (using stock mechs to keep it easier to calculate).

The creators even pulled up a website where you can paste Smurfy links and it will calculate the MBV of the config. You can find the most recent values on this page as well. Its being tweaked with every tournament the german community completes. (http://mbv.36thdieron.de/)

If you play around with it, you will notice that the cheesier the config, the higher its MBV.

This gives you/us a nice piece for a matchmaker metric.

I would do it like this:

[MBV of the Mech] * [ELO of the player] = Probable Player Combat Contribution Value (PPCCV)

This accomplishes several things:
1) A crappy player in a great config will reach a low PPCCV and probably be less often stomped. As his ELO goes up, his PPCV goes up in leaps, bringing him to his limits (challenging games) pretty quick.
2) A great player in a miserable mech will not have to deal with matches where he is constantly matched against the cheesiest/"best" builds piloted by players of comparable skill.
3) You don't go nuts watching the players on your team after you died, constantly thinking "with that mech, I would do A LOT better". Or at least less often then this happens now.

In a first attempt you could then try to match players in the same way as you match them now, using the PPCCV instead of the ELO.

This should lead to much more balanced matches, since your not ignoring a very important fact in Battletech: "If 2 Mechs are different, they are not the same."

Edited by ClaymoreReIIik, 10 February 2014 - 07:12 AM.

[Sidekick]

No. I completely disagree.

If we would implement a battle value in any shape or form, medium mechs may become OP. Also, if we implement a BV limit to te teams, the players would need to know the Value of their mech AND their player stats. Clearly, this would blow up the complete matchmaking system. Just imagine. Using a bad or mediocre mech to buff your overall team.

That would blatantly DESTROY the current gameplay.


So, we cannot use this system because:
- It would motivate players to use medium mechs
- it would offer a undesired transparency
- it would destroy the current matchmaking system
- it would hinder all further balancing attempts

Some might say, such heresy may cause cancer, damnation, homosexuality, diabetes and erec. dysfunction.


Please. Stop posting obvious statements.

Or to say it in our language:

Das ist, wie Eulen nach Athen zu tragen.

[ClaymoreReIIik]

Sidekick, on 10 February 2014 - 01:21 PM, said:

No. I completely disagree.

If we would implement a battle value in any shape or form, medium mechs may become OP. Also, if we implement a BV limit to te teams, the players would need to know the Value of their mech AND their player stats. Clearly, this would blow up the complete matchmaking system. Just imagine. Using a bad or mediocre mech to buff your overall team.

That would blatantly DESTROY the current gameplay.


So, we cannot use this system because:
- It would motivate players to use medium mechs
- it would offer a undesired transparency
- it would destroy the current matchmaking system
- it would hinder all further balancing attempts

Some might say, such heresy may cause cancer, damnation, homosexuality, diabetes and erec. dysfunction.


Please. Stop posting obvious statements.

Or to say it in our language:

Das ist, wie Eulen nach Athen zu tragen.

While I do appreciate well formulated irony, this forum is for constructive feedback.

So in an attempt to stay constructive:

Sidekick, on 10 February 2014 - 01:21 PM, said:

So, we cannot use this system because:
- It would motivate players to use medium mechs

Oh well....motivating people into mech choices is a controverse thing. Some will like it, some won't.

Sidekick, on 10 February 2014 - 01:21 PM, said:

So, we cannot use this system because:
- it would offer a undesired transparency

They could implement it like they implemented ELO "We have some metric that will put a value to something regarding your mech. This will help the matchmaker." This would help against the pesky thing called transparency.

Sidekick, on 10 February 2014 - 01:21 PM, said:

So, we cannot use this system because:
- it would destroy the current matchmaking system

Lets call it "expand on the current matchmaking system" and the (good) work they did already was not for the recycle bin.

Sidekick, on 10 February 2014 - 01:21 PM, said:

So, we cannot use this system because:
- it would hinder all further balancing attempts

You kind of lost me here. I would guess that a system like this would give them another wheel to turn for much more complicated balancing attempts. You could make a weapon system worse and make its MBV value higher, so that the matchmaker considers it better, creating all sorts of new imbalances in the matchmaker. More balancing passes guaranteed. (Job Security is awesome!)


Or did I get something wrong?

Edited by ClaymoreReIIik, 11 February 2014 - 06:17 AM.

[Arnold J Rimmer]

As an habitual Medium pilot, I approve this message.

[Knockknock]

My point of view: This solution is a better one than the current system which is in use.
It is much more uncomplicated to understand, everbody can estimate the MBV he will drop with and it will be more less frustrating then getting a "great" 12/0 elo-game after another beacuse the possible balancing is much smoother.
As developer you'll get a ton of more options balance all items, mechs and waepons using MBV.

Sidekick, on 10 February 2014 - 01:21 PM, said:

Or to say it in our language:

Das ist, wie Eulen nach Athen zu tragen.

So we going to use a truck to take this owls to athen. Is much more uncomplicated than carrie them by hand

Edited by Knockknock, 18 February 2014 - 06:38 AM.

[nehebkau]

I like the idea of a Mech battle value but it would require a TON of work to get right and, now, there are far more dire issues to deal with. None the less, it should be something that we start thinking about and hashing out.

[ClaymoreReIIik]

nehebkau, on 18 February 2014 - 07:57 AM, said:

I like the idea of a Mech battle value but it would require a TON of work to get right and, now, there are far more dire issues to deal with. None the less, it should be something that we start thinking about and hashing out.

There is not really much work left to be done about the Mech battle value. Just check out the website of the 36th dieron and paste some smurfy links into their MBV calculator. There has been a lot of work invested by members of the german MWO community to get it right, already.

Like mentioned in the OP we played tournaments using the MBV as balancing factor for the 12 vs 12 mech loadouts and it was pretty much spot on. Hell of a lot better then balancing teams by tonnage or weight class.

[Khobai]

battlevalue could work, but youd need different battlevalues for different maps and gamemodes. for example a light mech is better on alpine conquest than it is on river city assault. So the battlevalue would have to change based on the map/gamemode.

personally I think battlevalue is overcomplicated and unnecessary. I think a better system would just be to divide mechs into 5 weight classes: tiny, small, medium, large, and huge. Then give each team 2 mechs from each weight class and 2 random mechs from weight classes that add upto 6 (i.e. 2 mediums, 1 tiny + 1 huge, or 1 large + 1 small, etc...)

[CDLord HHGD]

You mean this number?

[ClaymoreReIIik]

cdlord, on 18 February 2014 - 02:20 PM, said:

You mean this number?

More like a value that takes into account all the different parts of the mech:

I am not sure this is the current version of the values but it gives a pretty good idea of how much was taken into account:

Spoiler

IS = Points * 2 * EE Bonus (1,5)
Armor = Points * 4 * FF Bonus (1,1)
Engine = Rating + used additional heat sink place (15*x)
Single Heat sink = 30
Double Heat Sink = 42 (engine internal 60)
Jump jets = 20
Speed (16,2 per MP) = (Engine Rating / Tonnage) * 16,2 * 3
Module Slot = 50
Torso angle = <DEG> / 90
Arm angle = <DEG> / 100 / 2 +1
Torso pitch = <DEG> / 20
Arm pitch = <DEG> / 30

Weapons = Damage - Heat - Cool down + Range + Max Range - Slots - Tons + Speed + Health + Energy weapon bonus * (Weapon type Modificator)
Damage * 10 (LBX * 8) (UAC * 15) (LRM-SRM * 7)
Heat * 4
Cool down *4
Range / 20
Max Range / 80
Slots * 3
Tons * 3
Speed / 40 (Laser weapons 50)
Health * 2

Weapon type Modificator
Small Laser * 0,5
Medium Laser * 1
Large Laser *1,2
PPCs *1,4
Ballistic Weapons AC2/AC5/UAC5 * 0,8
Ballistic Weapons LBX10/AC10/AC20 * 0,9
LRMs * 0,7

Location Modificator
Head, Center/Right/Left Torso = Torso Angle + Torso Pitch -1
Right/Left Arm = Torso Angle + Torso Pitch + Arm angle + Arm Pitch -3


Heat Efficiency Modifier = 2*(1-EXP(Heat Efficiency)/0,35))
Heat Threshold Modifier = (Threshold-40)*0,015+1

Ammo
Ballistic Weapons = Shots * Damage / 3
LRMs = Shots * Damage / 4
SRMs = Shots * Damage / 3
AMS = 15
NARC = 10

Equipment
Artemis = Weapon Value * 1,2
AMS = 150
NARC = 50
TAG = 120
ECM = 300
BAP = 200
C.A.S.E = 50

Total = IS + Armor + Engine + Heat sinks + Speed + (Weapons Total * Heat Efficiency Modifier * Threshold Modifier) + Ammo

[BladeXXL]

Doing this I would miss my favorite endings like "12:1" or "2:12"
... oh wait... not really! ^^

[Foster Bondroff]

Second this request. All for it.

To be honest, the effort from the german community and the exellent games that MBV creates is the only reason, why PGI and IGN have any hope, that i will ever spend a single dime again on this game.

If those where not present, MWO would have been uninstalled long long ago.

[Sh4nk0h0l1c]

Well whritten Clay,

hope it falls on the Dev's ears, instead on deaf ears

I support this!