48 replies to this topic

[Greyhart]

OK bored. Simple question.

What things should Match Maker take into account when creating balanced teams. (nothing that isn't already tracked in game).

So off the top of my head (no particular order):

1. Tier
2. Tonnage
3. Number of Skills unlocked
4. Speed
5. alpha strike power
6. average match score in the mech

any other thoughts?

[Bud Crue]

Given the frequency that I see the same names over and over during an evening of matches, I'd say we should be grateful that the MM can get us matches with 24 people all having a pulse. We can hope for a MM to take all sorts of things into account such as players w/l rating or average score or mech tonnage, or whatever but without a large population of players playing at the same time all of that is irrelevant.

Apropos, I don't think alpha strike power is relevant unless your theoretical MM also takes into account other weapons stats such as potential range, dps, etc. I mean is a Catapult AI or a SRMcrow with a high alpha equal to a a gauss/vomit build of similar "power". Also what about GH? I can build a nova with a massive apparent alpha but I'll die if I fire it. Is that weighted equally to a mech that CAN successfully alpha without shut down? Similarly "speed" seems problematic. I mean is a full engine Quickdraw that is forced to run an XL in anyway equal to a Linebacker moving at the same rate? An 81KPH Timberwolf equally weighted against a similar speed Enforcer? Not seeing how relative speed could be weighed and have useful outcomes by any MM that PGI might come up with.

[Lord0fHats]

Quote

I'd say we should be grateful that the MM can get us matches with 24 people all having a pulse.

[Dr Hobo]

Or take into account the BV of the mechs on the field could do that too.

[Greyhart]

Dr Hobo, on 10 August 2017 - 05:15 AM, said:

Or take into account the BV of the mechs on the field could do that too.

The BV thing goes down like a lead balloon in some areas of this forum. But yes.

Also I would say that this is about balancing teams, if the population is low it means that the teams would be more balanced as the MM is not just using tonnage

[Dr Hobo]

Greyhart, on 10 August 2017 - 05:25 AM, said:

The BV thing goes down like a lead balloon in some areas of this forum. But yes.

Also I would say that this is about balancing teams, if the population is low it means that the teams would be more balanced as the MM is not just using tonnage

With skills tonnage and BV taken into account,maybe it would open up for some more diversity in builds,and you won't see as much of an alpha warrior arty strike online we have now.

[Ruar]

KMDD should be pretty high up the list for MM sorting as it shows individual skill better than most stats.

Average match score should also have a significant place in sorting.

After that though the rest is usually arbitrary. W/L is more about the team than the individual. Overall damage doesn't really mean much since you can get a kill with 30ish damage to the right spots.

I think to truly have an effective MM they have to go in and add some flags to each mech. Tonnage alone doesn't say much about the mech's combat potential. Just compare the Centurion, Trebuchet, Hunchback, and Shadowhawk to see that even when within a five ton difference there can be a significant difference in combat potential.

Each variant of the mechs needs a combat potential rating. I realize this is just a baseline because weapons make a huge difference, but there needs to be some room for surprises. So the combat potential is fixed based on chassis and not weapon loadout.

Each pilot will have a pilot rating based on KMDD and average match score for the past 14 days. This number provides the new tier based on percentage of people a certain rating. So the top 10% are the tier one players. 89-75 are tier two. 74-60 are tier three. 59-35 are tier four. 34-1 are tier five.

When a player enters the que their tier rating and mech combat potential are combined to provide a match rating. First the MM divides up teams based on tier and then match rating. Priority balancing for tiers one and two is a must since these people have more influence than others. Tiers three through five have more wiggle room to allow the match ratings to be put closer to equal.

Using this system should reflect both player skill and mech potential while still leaving enough room for each match to feel unique and allow for surprises.

[Kjudoon]

Throw tiers out completely or make a lot more of them. Every stat in the end of match page except W/L. Or just use match score if that is an actual aggregate representation of what you do in match... which I have no idea. The tool might already be there.

Then the rest is modified by your Mech's Battle Value. That means every piece of hardware, chassis, piece of armor, consumable and skill has a BV for it is a factor in match. This score modifies your base Average Match score. Possibly significantly. So you make a joke Atlas or a Meta Arctic Cheater it modifies who you are considered capable of facing. At least this means your matchmaking is handicapped like a golf or bowling average.

Regardless, equipment needs to factor into the equation, and W/L must be thrown out. If you're going full potato, you need to not benefit so much. But if you're doing the carrying, you deserve competent allies and opponents.

[Tarl Cabot]

OP, you are looking at it in the wrong way. MM portion itself works with what it is given. It would not matter what other factors are utilize until the PSR itself has been changed to where it does not act like it is an experience bar, moving people up into tiers they may not should be in simply because they drop a lot and do okay.

In another thread a player suggested that the Tier-PSR (still needs to be changed ) should be reset on a regular basis, reseeding everyone based on the previous seasons' performance since PGI has implemented Seasons = month. Basically shake things up. Those with excellent stats would still be Tier 1.

What would that do? I am currently Tier 1 maxed out and have been for a two or three seasons but my actual seasonal performance would likely put me somewhere in Tier 3 or low Tier 2. Those who are in perpetual Tier 4 hell due to some horrific starting stats may then be able to get out of that tier where they should be, leaving the the new players/new accounts/horrible players in Tier 4/5.

Thus I would suggest that Tiers be reset every quarter, using the previous quarter's stats. And for players who have been seeded at least once would never be seeded as a Tier 5 regardless of how much they attempted to tank their stats in the previous quarter.

Edited by Tarl Cabot, 10 August 2017 - 05:58 AM.

[Ruar]

Kjudoon, on 10 August 2017 - 05:43 AM, said:

Throw tiers out completely or make a lot more of them. Every stat in the end of match page except W/L. Or just use match score if that is an actual aggregate representation of what you do in match... which I have no idea. The tool might already be there.

Then the rest is modified by your Mech's Battle Value. That means every piece of hardware, chassis, piece of armor, consumable and skill has a BV for it is a factor in match. This score modifies your base Average Match score. Possibly significantly. So you make a joke Atlas or a Meta Arctic Cheater it modifies who you are considered capable of facing. At least this means your matchmaking is handicapped like a golf or bowling average.

Regardless, equipment needs to factor into the equation, and W/L must be thrown out. If you're going full potato, you need to not benefit so much. But if you're doing the carrying, you deserve competent allies and opponents.

Having every piece of equipment accounted for in a battle value would make for very boring matches. There needs to be some room for teams to do better or worse than expected. Otherwise every match will just feel the same.

[Dr Hobo]

Ruar, on 10 August 2017 - 05:48 AM, said:

Having every piece of equipment accounted for in a battle value would make for very boring matches. There needs to be some room for teams to do better or worse than expected. Otherwise every match will just feel the same.

Not really no.

Because then you can add in other things. The BV of a mech is just one stat. Going with some of the other ideas here(such as KMDD etc) then adding into skills on a mech(does it have 1/3,2/3 or all 91 pts of it's skills in play?) THEN on top of that,you can even go with what I suggested bout strike spamming(Where a mech has to be narc'd or tagged[to get even more accurate] or just under a uav[to get what we have now] then a mech with a command console can issue the strike)

Because doing all that,you won't see everyone gravitating to the biggest guns. It wont be LRM100 mechs. You will hopefully actually see some build diversity and that's honestly what we need.

[Ruar]

Dr Hobo, on 10 August 2017 - 06:28 AM, said:

Not really no.

Because then you can add in other things. The BV of a mech is just one stat. Going with some of the other ideas here(such as KMDD etc) then adding into skills on a mech(does it have 1/3,2/3 or all 91 pts of it's skills in play?) THEN on top of that,you can even go with what I suggested bout strike spamming(Where a mech has to be narc'd or tagged[to get even more accurate] or just under a uav[to get what we have now] then a mech with a command console can issue the strike)

Because doing all that,you won't see everyone gravitating to the biggest guns. It wont be LRM100 mechs. You will hopefully actually see some build diversity and that's honestly what we need.

Except no one builds mechs based on battle value. They build mechs based on what provides the most return for how they want to play. Trying to quantify every little nuance in the game so it can be reflected in the MM will just make for an incredibly boring match.

[Appogee]

Bud Crue, on 10 August 2017 - 04:59 AM, said:

we should be grateful that the MM can get us matches with 24 people all having a pulse.

Wait, you get all 24 people in your matches...?!

[Bud Crue]

Appogee, on 10 August 2017 - 06:38 AM, said:

Wait, you get all 24 people in your matches...?!

Well, not for the entire match. Of course not. I didn't mean to come off as sounding unreasonable.

[Kjudoon]

Ruar, on 10 August 2017 - 06:36 AM, said:

Except no one builds mechs based on battle value. They build mechs based on what provides the most return for how they want to play. Trying to quantify every little nuance in the game so it can be reflected in the MM will just make for an incredibly boring match.

Well that's an assumption on your part.

BV is not about 'calculating nuance' in the same way calculating damage doesn't have to figure out 'nuance'.

All BV would do is calculate the value of a mech on the field and force it to face comparable equipment. This assumes players of equal skill. But since the suggestion is to use BV as a modifier to better match PSR.

This would not change anything in how a match is played. But if you mean boring as in you would no longer be as able to club baby seals by abusing the PSR failings and were forced to face better players in crappy equipment, players of equal skill and equipment , or worse players in better equipment, then yes, it would suck the unfair advantage you were enjoying right out of the game.

Which is kinda the point.

Edited by Kjudoon, 10 August 2017 - 06:59 AM.

[Ruar]

Kjudoon, on 10 August 2017 - 06:56 AM, said:

Well that's an assumption on your part.

All BV would do is calculate the value of a mech on the field and force it to face comparable equipment. This assumes players of equal skill. But since the suggestion is to use BV as a modifier to better match PSR.

This would not change anything in how a match is played. But if you mean boring as in you would no longer be as able to club baby seals by abusing the PSR failings and were forced to face better players in crappy equipment, players of equal skill and equipment , or worse players in better equipment, then yes, it would suck the unfair advantage you were enjoying right out of the game.

Which is kinda the point.

Lol, classic. If I don't agree with your point then it's because I'm afraid of fair competition and only enjoy beating up on weaker players. There's just no way I could like having a balanced and fair match but with some room for surprises. I mean, if players were perfectly balanced and mechs were perfectly aligned there is just no way the matches would be stale because you'd be fighting the same people over and over with the same tactics and mechs being constantly used. Nah, that just couldn't happen at all.

[Dr Hobo]

Ruar, on 10 August 2017 - 06:36 AM, said:

Except no one builds mechs based on battle value. They build mechs based on what provides the most return for how they want to play. Trying to quantify every little nuance in the game so it can be reflected in the MM will just make for an incredibly boring match.

But in reality,you already are. XL engines,LFE,gauss PPC,etc all high BV equiptment,pushes you into higher BV's.

You'll have folks running super stocks,or stock builds(Newer players,casual etc),and you'll have people pushing the limits of what a mech can and can't do(Die hard meta chasers). And they won't often see each other unless they're really,really good with said build(or really,realllly bad with it)

This helps prevent stomps to teams that have mostly stock mechs v fully upgraded mechs.

I.e a 1500 BV mech won't fight a 15000 BV mech,but a 9000 BV mech can see both ends(on the high end of the 1500,but the low to mid end of the 15000)

Then you add in the KMDD to that stat,so someone who's really good in say,a superstock HBK-4G can sometimes see opponents who will be having slightly higher BV's than his superstock 4G. But someone in a bone stock Awesome won't be seeing a full meta tryhard Madcat Mk2 much,if at all.

Poor MM,mechanics,maps and other issues this game has is preventing it from retaining players(due to it's near vertical learning curve). Id love for us to get more players,not less. I think more folks want games that are closer and better,not a 3 minute match on a 5-12 bout.

I'm not saying you're wrong,but I'm not agreeing with you.

[Mystere]

Idealistic ideas, meet the real world. The overly complicated suggestions I deduce from these posts will either result in long wait times or release valves wide open given the realities of MWO today.

Just use wins and losses and nothing else as far as stats are concerned and add some weight class rules to keep mech variety in fights. It is much simpler and will be just as good.

Edited by Mystere, 10 August 2017 - 07:21 AM.

[Dr Hobo]

Mystere, on 10 August 2017 - 07:19 AM, said:

Idealistic ideas, meet the real world. The overly complicated suggestions I deduce from these posts will either result in long wait times or release valves wide open given the realities of MWO today.

Just use wins and losses and nothing else. It is much simpler and will be just as good.

You know,that's what War Thunder did,not openly however. I think behind the scenes.

Because back under it's old 25 tier system(where it was a legit progress and nowhere near as long as it is now) a bone stock BF-109 wasn't fighting maxed out Yaks and LAs.

We honestly need more players and better mechanics and a better PSR system. Like,the PSR is a decent idea(like ghost heat was,but we would have never needed ghost heat if we had a targeting system ala MW3/4 LL had)

[Coolant]

don't really need a matchmaker, all it has done is cause complaints (and unjustly). Figure out a fair way for new players to play against new players. Other than that, MW4:Mercs multiplayer didn't have a matchmaker and it had servers going for about a decade.