23 replies to this topic

[Thraka]

It was a blast loading up polar highlands with a nice giant yellow puddle in the middle only to discover that our entire team was too slow to get close enough to capture or even shoot the enemy to stop them.

[martian]

Thraka, on 11 September 2020 - 11:38 PM, said:

Fix Weights In Matchmaking

It was a blast loading up polar highlands with a nice giant yellow puddle in the middle only to discover that our entire team was too slow to get close enough to capture or even shoot the enemy to stop them.

Russ Bullock and Paul Inouye are okay with the tonnage imbalance and matchmaking.

Have fun and buy 'Mech Packs.

[crazytimes]

About my 4th drop today. My team had 3 fleas and 2 ravens- none in groups- and 1 assault. Their team had 5 assaults. It dragged out way longer than you would have expected, but the result was never a surprise. Too much fire-power and weight disparity.

I waited around long enough to see I got a PSR up still... then I hit alt f4 and played Divinity OS II instead.

Edited by crazytimes, 12 September 2020 - 01:53 AM.

[martian]

crazytimes, on 12 September 2020 - 01:50 AM, said:

About my 4th drop today. My team had 3 fleas and 2 ravens- none in groups- and 1 assault. Their team had 5 assaults. It dragged out way longer than you would have expected, but the result was never a surprise. Too much fire-power and weight disparity.

I waited around long enough to see I got a PSR up still... then I hit alt f4 and played Divinity OS II instead.

I guess that this is probably a relatively common reaction.

[Willard Phule]

With the addition of premade groups to QP, the MM is completely unable to match weights at all. Premade groups don't count toward the 3/3/3/3 the matchmaker is supposed to be putting together. If you have a 4 man premade group and they all drop light, your team will most likely be without assaults. Same thing for if they go heavy.

[VonBruinwald]

martian, on 12 September 2020 - 01:27 AM, said:

Russ Bullock and Paul Inouye are okay with the tonnage imbalance and matchmaking.

And so are the group players. Don't forget, they're the ones choosing what tonnage/classes they drop as.

[Beorning]

I am mostly ok with the tonnage thingy - been in a few wolf packs that ate up teams of enemy assaults. I even had one such match piloting the Anansi hero.

Also domination is a stupid mode.

[w0qj]

Yes, agree the MWO should consider total tonnage...

I'm having my fun in LRM boat...as an Assault mech (!).

Whilst this is fun for me, total group tonnage would be way out if more than a few folks do what I did!

[Capt Deadpool]

The majority of solo light pilots definitely got screwed (as well as whatever side they drop on) after introduction of soup que, as weight matching seems to be low on the match-making priority list. Being too heavy of a team is going to be a far less frequent problem, despite the original post, even given the presence of experienced wolfpackers dropping together sometimes.

For those posting in this thread who are not aware, MM prioritizes matching by group size and PSR before taking into account weight, so you would have to bump weight up the priority ladder instead of one of those criteria which would likely make things even worse than they are now.

Unfortunately, since PGI has decided to make group/solo que merger permanent, I think we are stuck with the tonnage imbalance issues.

[Tarl Cabot]

Willard Phule, on 12 September 2020 - 03:36 AM, said:

With the addition of premade groups to QP, the MM is completely unable to match weights at all. Premade groups don't count toward the 3/3/3/3 the matchmaker is supposed to be putting together. If you have a 4 man premade group and they all drop light, your team will most likely be without assaults. Same thing for if they go heavy.

The Soup queue is not using the previous Solo queue MM setup but the Group queue MM setup. Whereas, the Solo queue MM could potentially work if drop to max 3-man and limit them to 1 weight class/group, then only 3-man or 2-man, no pairing up more than one group to a side. Then lot would depend on HOW PGI has MM setup.

Seeded Group first, then find another group for the other side then fill up with matching weight classes for each side so weight classes are match up both sides.

Seed Solo Player, then only solo players for that drop, weight classes equal on both sides.

[Abu619]

I sure do miss having an assault lance, medium/heavy lance , and light lance. the drops are way off now.

[MrMadguy]

Thraka, on 11 September 2020 - 11:38 PM, said:

It was a blast loading up polar highlands with a nice giant yellow puddle in the middle only to discover that our entire team was too slow to get close enough to capture or even shoot the enemy to stop them.

Just don't vote for this map. It's broken and should be deleted from game. It was created only to sell Archer better. At that point LRMs were nerfed to the ground and nobody would buy this 'Mech, if they wouldn't be buffed somehow. And this map is obviously the best for Lights and LRMs.

Willard Phule, on 12 September 2020 - 03:36 AM, said:

With the addition of premade groups to QP, the MM is completely unable to match weights at all. Premade groups don't count toward the 3/3/3/3 the matchmaker is supposed to be putting together. If you have a 4 man premade group and they all drop light, your team will most likely be without assaults. Same thing for if they go heavy.

Current MM works perfectly. It's 3x4, that was wrong, because "skill" >> "weight" and MM should think about balancing SKILLS - not weight. For example I'm bad Assault pilot. I perform much better in Heavies, than in Assaults. And some pilots are extremely deadly in Mediums and Lights. That's, why 3x4 was working so badly. For example: game lacked Lights, so it was forcing Tier 1 Lights into my groups due to 3x4 rule, instead of putting Tier 3 Heavies or Assaults, as it was doing prior to implementation of 3x4. Guess, what result we had due to this mess? Stomps, stomps and even more stomps. Because I was supposed to carry matches, because I was in Heavy or Assault. And I couldn't.

Edited by MrMadguy, 13 September 2020 - 08:27 PM.

[Willard Phule]

MrMadguy, on 13 September 2020 - 08:19 PM, said:

Current MM works perfectly. It's 3x4, that was wrong, because "skill" >> "weight" and MM should think about balancing SKILLS - not weight. For example I'm bad Assault pilot. I perform much better in Heavies, than in Assaults. And some pilots are extremely deadly in Mediums and Lights. That's, why 3x4 was working so badly. For example: game lacked Lights, so it was forcing Tier 1 Lights into my groups due to 3x4 rule, instead of putting Tier 3 Heavies or Assaults, as it was doing prior to implementation of 3x4. Guess, what result we had due to this mess? Stomps, stomps and even more stomps. Because I was supposed to carry matches, because I was in Heavy or Assault. And I couldn't.

If you'll remember back to the day when they first implemented a "matchmaker," everyone had 4 "skill ratings." One for each weight class. But that adds too many variables for the MM to put together a match quickly and efficiently, which was one of the primary complaints about separation. So, they went to this whole PSR thing.

One of the better ideas that was put forth and promptly ignored was the idea of using a "battlevalue" type system that totals up the offensive and defensive capabilities of a chassis and adds a "pilot rating" to the BV based on something like we've got with PSR. I'm gonna guess it was ignored because it's way beyond the capability of PGI to code it.

As for it working "perfectly," I suppose it does from an organized, premade group player's perspective. But from a purely solo player's perspective, I can absolutely understand why they don't like being harvested every single match by people that are not only more skilled, but are organizing against 12 random people without a plan. Not that it matters. It's the same mechanic in FP, and FP doesn't even have a matchmaker. The concept is simple: MW:O is supposed to be a "team" game. A solo is a 1 man team, and each side of a match consists of 12 individual teams. If you want to win, join a premade group. Then, you'll be in a team of 4 against 12 individual teams.

[RickySpanish]

It was ignored because it was a flat out unworkable. A BV system could never, ever work in a game where even the physical geometry of a 'Mech can make or break its effectiveness. There are far, far too many variables to consider when implementing a singular numerical value to determine the worth of a 'Mech. Heck, that would even change on a map by map and game mode basis. Don't you find it strange that there has been so much difficulty reducing a player's combat effectiveness to one number (Elo/PSR), yet here's advocation for yet another even more complicated solution that also attempts the foley of condensing a player's potential to a single number?

Edited by RickySpanish, 14 September 2020 - 07:16 AM.

[Black Caiman]

I enjoy the weight disparities if Im honest. It adds an additional challenge to the game, and whether its real life combat, or combat from the Battletech universe their have always been disparities in strength to overcome. If youre team is severely out-tonned and you score a win thats a good feeling, and if you lose you can chalk it up to "oh well they just had too much beef for us." Its not any different then when you see a super-group on the other team. If you beat them thats a great feeling, but if you get cleaned up then oh well that was kind of expected.

I dont see why you couldnt assign some sort of sliding, constantly adjustable BV to each mech. All you would have to do was monitor the performance of each chassis/vartiant in each match it is involved in and move it up or down the scale based on how it performs. So if over 1000 matches played the WHM-6R averaged a 300 match score you can slot it to the appropriate BV, but if over the same amount of matches the CP-S averages a 350 match score then it would have a higher BV.

[Willard Phule]

Black Caiman, on 14 September 2020 - 07:47 AM, said:

I enjoy the weight disparities if Im honest. It adds an additional challenge to the game, and whether its real life combat, or combat from the Battletech universe their have always been disparities in strength to overcome. If youre team is severely out-tonned and you score a win thats a good feeling, and if you lose you can chalk it up to "oh well they just had too much beef for us." Its not any different then when you see a super-group on the other team. If you beat them thats a great feeling, but if you get cleaned up then oh well that was kind of expected.

I dont see why you couldnt assign some sort of sliding, constantly adjustable BV to each mech. All you would have to do was monitor the performance of each chassis/vartiant in each match it is involved in and move it up or down the scale based on how it performs. So if over 1000 matches played the WHM-6R averaged a 300 match score you can slot it to the appropriate BV, but if over the same amount of matches the CP-S averages a 350 match score then it would have a higher BV.

That's how they did it in BT. Chassis is worth X points based on tonnage. Each weapon system has a numerical value. Mobility has a rating. Add all that crap up and you get a BV. Give the pilot's skill rating a numerical value and add them together, and you get a number that's representative of everything combined.

[VonBruinwald]

Willard Phule, on 14 September 2020 - 08:01 AM, said:

That's how they did it in BT. Chassis is worth X points based on tonnage. Each weapon system has a numerical value. Mobility has a rating. Add all that crap up and you get a BV. Give the pilot's skill rating a numerical value and add them together, and you get a number that's representative of everything combined.

I'd rather see each chassis get it's own PSR rating with the PSR at purchase based on the average of your top 10 mechs in the same weight class.

Edited by VonBruinwald, 14 September 2020 - 09:22 AM.

[Nightbird]

Willard Phule, on 14 September 2020 - 08:01 AM, said:

That's how they did it in BT. Chassis is worth X points based on tonnage. Each weapon system has a numerical value. Mobility has a rating. Add all that crap up and you get a BV. Give the pilot's skill rating a numerical value and add them together, and you get a number that's representative of everything combined.

If you get a bunch of brown material together, you don't get anything more than a big pile o brown material.

[Willard Phule]

VonBruinwald, on 14 September 2020 - 09:20 AM, said:

I'd rather see each chassis get it's own PSR rating with the PSR at purchase based on the average of your top 10 mechs in the same weight class.

Nightbird, on 14 September 2020 - 10:17 AM, said:

If you get a bunch of brown material together, you don't get anything more than a big pile o brown material.

True enough, and both are great reasons to just let full 12 man premades into QP.

[GARION26]

Nightbird, on 14 September 2020 - 10:17 AM, said:

If you get a bunch of brown material together, you don't get anything more than a big pile o brown material.

Agreed or the old programming mantra GIGO "Garbage In, Garbage Out"
We can't agree as a player base on how Match Score should optimally be calculated, much less how PSR should be calculated, or how matchmaker should work. I don't think we're going to remotely be able to do a heavily nuanced more detailed PSR on the chassis by chassis level.