184 replies to this topic

[Xavori]

I think I can safely say, with little disagreement, that our current matchmaker sucks. There are far too many 12-0 matches, and not enough 12-11 matches. It doesn't have to be this way.

Even with a smaller player population, it's entirely possible to create matches that are more likely to produce balanced gameplay. I've been in a number of long-running small dart leagues that had teams with players who had skill from 6-7 dart outs to lucky to hit the board with all 3 darts. But because teams had balanced ratings, the matches still were competitive.

MWO could have that. There are many options available for creating player ratings, and then using those ratings to assemble roughly equivalent teams. This would dramatically improve the quality of matches, and should make the game more enjoyable as well.

[Exard3k]

how does your proposal influence wait times?

[Xavori]

Exard3k, on 29 November 2017 - 09:54 AM, said:

how does your proposal influence wait times?

Shouldn't change it at all. The goal isn't perfectly balanced, it as close as can be with the available players at any given time.

[MischiefSC]

Xavori, on 29 November 2017 - 09:57 AM, said:

Shouldn't change it at all. The goal isn't perfectly balanced, it as close as can be with the available players at any given time.

More to the point it doesn't extend them at all. It builds match with the best options available. The advantage of a better matchmaker is a better distribution of players. With tiers you end up with bigger swings one match to the next.

The ideal option would be an Elo score for each player averaged with an Elo score for each mech and a modifier for loadout. So player A may be good and in a laservomit MAD IIC he's worth value X, but in a 2 flamer Spider he's worth far less.

That would give the most accurate results based on what's available.

[Bud Crue]

Xavori, on 29 November 2017 - 09:48 AM, said:

Even with a smaller player population, it's entirely possible to create matches that are more likely to produce balanced gameplay. I've been in a number of long-running small dart leagues that had teams with players who had skill from 6-7 dart outs to lucky to hit the board with all 3 darts. But because teams had balanced ratings, the matches still were competitive.

MWO could have that.

Wait a sec!
So PGI balances by dartboard and we all know how that goes. Now you want to have the MM function by dartboard!?
Are you insane!!!?

[Xavori]

Bud Crue, on 29 November 2017 - 10:27 AM, said:

Wait a sec!
So PGI balances by dartboard and we all know how that goes. Now you want to have the MM function by dartboard!?
Are you insane!!!?

If the MM was using a dartboard, I think there'd be less 12-0 stomps

[Nightbird]

What ever happened to random numbers are random? If you believe that then no matchmaker can do what you want here.

[Escef]

Xavori, on 29 November 2017 - 11:23 AM, said:

If the MM was using a dartboard, I think there'd be less 12-0 stomps

Interestingly enough, you're probably right. Because a match where each player is equally valuable (which is what the MM strives for) will have a "snowballing" effect when a team loses a player. If players are of unequal value, losing your lowest value player is less of a handicap.

[RickySpanish]

"Stomps" don't necessarily imply a failure on the part of the match maker. There are a lot of circumstances where losing one or two 'Mechs early on can seal a team's doom in a fashion that snowballs out of control. A team with 0 deaths may have several heavily damaged 'Mechs that the flailing opponents failed to coordinate to finish off. If your team has a disconnect, the likelihood of you losing through a snowballing effect is even greater.

[Mole]

What the hell is it with you people and requesting something, PGI gives it to you, then you request the old system back thinking it's going to be good? People begged for map voting. Now they are begging for random maps like its the solution to everything. Which is what we had before map voting. I wasn't around for 8v8 but I can only assume it was changed to 12v12 by popular demand, now people are acting like 8v8 is the answer to all balancing problems ever. And now here you are requesting ELO which is exactly what we had before we pressured PGI into a tier system because the ELO was failing to make good matches too. Remember kids; history is important.

[Vlad Striker]

ELO rating must be calculated on every owned chasis separately, "avarage patients temperature on hospital" does not make any sense but PGI made exactly this mistake at near past when they try to apply ELO system.

[Khobai]

if you take 24 random people and divide them among 2 teams, most of the time the skill level should be pretty close, since skill level follows a bell curve and most people should be towards the center of the bell curve.

so its arguable whether you really even need a matchmaker to balance player skill


where matchmaker fails the most though is that it doesnt balance the tonnage or quality of mechs for both teams. So it might stick a boar's head with all flamers on one team and a meta madcatmk2 with laser/gauss on the other team.

thats the bigger disparity.

I mean how can you expect matchmaker to possibly work when it considers a bottom-of-the-barrel troll fatlas to be equal to a top tier meta madcat mk2?

they need some kindve rudimentary battlevalue system to rank how good mechs are

Edited by Khobai, 29 November 2017 - 12:40 PM.

[Escef]

Vlad Striker, on 29 November 2017 - 12:35 PM, said:

ELO rating must be calculated on every owned chasis separately, "avarage patients temperature on hospital" does not make any sense but PGI made exactly this mistake at near past when they try to apply ELO system.

The problem is that there are a LOT of ways to outfit any given chassis. There's a HUGE difference between Battlemaster 1S rigged up as an LRM boat, a 400XL splat striker, and a quad ERLL sniper.

[RickySpanish]

Mole, on 29 November 2017 - 12:29 PM, said:

What the hell is it with you people and requesting something, PGI gives it to you, then you request the old system back thinking it's going to be good? People begged for map voting. Now they are begging for random maps like its the solution to everything. Which is what we had before map voting. I wasn't around for 8v8 but I can only assume it was changed to 12v12 by popular demand, now people are acting like 8v8 is the answer to all balancing problems ever. And now here you are requesting ELO which is exactly what we had before we pressured PGI into a tier system because the ELO was failing to make good matches too. Remember kids; history is important.

It's as if using a singular numeric rating system to determine the potential of winning or losing a match that pits teams of opponents against one another in a complex, almost infinitely customizable game of skill, is actually a surprisingly difficult affair.

[Khobai]

you dont really have to balance skill though

when you take 24 random people and divide them on 2 teams the skill level should be distributed fairly evenly following a bell curve distribution. the current tier system probably suffices for that.

what you do have to balance is how good the mechs are on each team. because a bad mech, like an all flamer atlas, is practically the same exact thing as having one less mech on your team. and matchmaker currently doesnt account for that... it makes the stupidly false assumption that everyone is running an equally viable build. which is a really bad assumption given the level of customization MWO allows.

I think when solo queue games go 0-12 its more because of the composition of the mechs on one team is inferior to the composition on the other team. not because one team has way less skilled players.

group queue games are obviously different. because sometimes you have to play against top 1% player deathstacks of like 10+ players with a bunch of 2-man and 4-man groups on your team. and you just cant possibly win because of the difference in skill level, coordination, and teamwork. group queue is stupid because a bunch of 2-mans and 4-mans should never have to fight a 10-man or 12-man. there is no way in hell thats fair even with the tonnage limits.

Edited by Khobai, 29 November 2017 - 01:23 PM.

[MischiefSC]

Mole, on 29 November 2017 - 12:29 PM, said:

What the hell is it with you people and requesting something, PGI gives it to you, then you request the old system back thinking it's going to be good? People begged for map voting. Now they are begging for random maps like its the solution to everything. Which is what we had before map voting. I wasn't around for 8v8 but I can only assume it was changed to 12v12 by popular demand, now people are acting like 8v8 is the answer to all balancing problems ever. And now here you are requesting ELO which is exactly what we had before we pressured PGI into a tier system because the ELO was failing to make good matches too. Remember kids; history is important.

No, the problem with Elo is that it worked. What people wanted was a system they thought would involve them being in matches with people who would carry them. So we got tiers, which puts mediocres and bads in with good players. The result is the mediocres get destroyed by goods and struggle to get a 1.0. Good players have an easier time now getting a higher win/loss, everyone else is getting a harder time.

People didn't like Elo because it worked and they're not as good as they want to think they are. So we got a new system that's pushing everyone up to T1 and they're just food for people who actually play at that level.

[MischiefSC]

Khobai, on 29 November 2017 - 01:14 PM, said:

you dont really have to balance skill though

when you take 24 random people and divide them on 2 teams the skill level should be distributed fairly evenly following a bell curve distribution. the current tier system probably suffices for that.

what you do have to balance is how good the mechs are on each team. because a bad mech, like an all flamer atlas, is practically the same exact thing as having one less mech on your team. and matchmaker currently doesnt account for that... it makes the false assumption that everyone is running an equally viable build.

I think most solo queue games go 0-12 more because of the composition of the mechs on one team is inferior to the composition on the other team. not because one team has way less skilled players.

group queue games are obviously different. because sometimes you have to play against top 1% player deathstacks of like 10+ players with a bunch of 2-man and 4-man groups on your team. and you just cant possibly win because of the difference in skill level.

You'll get an even distribution in the average, but not aggregate. So individual matches will swing more wildly. Yes, averages are the same but the numbers it's averaging are bigger. I know my win/loss jumped a full 10%.

[Mole]

MischiefSC, on 29 November 2017 - 01:19 PM, said:

No, the problem with Elo is that it worked. What people wanted was a system they thought would involve them being in matches with people who would carry them. So we got tiers, which puts mediocres and bads in with good players. The result is the mediocres get destroyed by goods and struggle to get a 1.0. Good players have an easier time now getting a higher win/loss, everyone else is getting a harder time.

People didn't like Elo because it worked and they're not as good as they want to think they are. So we got a new system that's pushing everyone up to T1 and they're just food for people who actually play at that level.

Come on. Don't bull **** me. People hated ELO because they kept getting grouped with potatos that they felt were the cause of the 12-0 stomp they just experienced. I know, I was there. Then PGI gave is the tier system. A whole lot of nothing changed, and now people are touting ELO as the solution.

[Xavori]

Khobai, on 29 November 2017 - 01:14 PM, said:

you dont really have to balance skill though

when you take 24 random people and divide them on 2 teams the skill level should be distributed fairly evenly following a bell curve distribution. the current tier system probably suffices for that.

what you do have to balance is how good the mechs are on each team. because a bad mech, like an all flamer atlas, is practically the same exact thing as having one less mech on your team. and matchmaker currently doesnt account for that... it makes the stupidly false assumption that everyone is running an equally viable build. which is a really bad assumption given the level of customization MWO allows.

I think when solo queue games go 0-12 its more because of the composition of the mechs on one team is inferior to the composition on the other team. not because one team has way less skilled players.

group queue games are obviously different. because sometimes you have to play against top 1% player deathstacks of like 10+ players with a bunch of 2-man and 4-man groups on your team. and you just cant possibly win because of the difference in skill level, coordination, and teamwork. group queue is stupid because a bunch of 2-mans and 4-mans should never have to fight a 10-man or 12-man. there is no way in hell thats fair even with the tonnage limits.

That'd only work if we a normalized distribution of player skill. First, we don't know that that is the case. Second, I'm seriously doubtful it is. I'd bet we have a lot more potatoes than comps.

Elo/TrueSkill ranking would get a lot closer to making sure the two teams had balanced skills compared to each other. It's doubtful it would ever be an exact match, but it doesn't have to be. It just has to be close.

As for snowballing, it should be less likely, albeit not impossible. The most likely players to die first on a team will still be the least skilled, and since both teams will have an even total skill, that shouldn't be a total disaster. Now, that's not to say that will always be the case, but the goal here isn't to make perfect 12-11 matches every time. It's to increase the likelihood of getting quality matches.

[Asym]

Oh, good Holy Lord, here we go again.......

Nothing is going to change. PGI is NOT going to ever consider changes to MM nor anything they've done in the past because they will not disturb the cash flow of Status Quo: people are buying stuff consistently to keep the books in the black.

Otherwise, they'd be reading the forums to figure where we are coming off the rails and not buying....

Why discuss something we have no control over and PGI will not fix????