33 replies to this topic

[Hydrocarbon]

Ever get that feeling matchmaker's decisions are set to cull ELO than balance it? Ever think you could get a better W:L ratio if only you didn't get those 7+ losses in a row every other day?? Were you like me and got a record 10 losses in a row during the Free Victor event???

I'm curious if anyone else has some screenshots of what can only be described as a horrible failure in matchmaker algorithms. In the following pic the opfor has FOUR mechs with ECM. That can be delt with right? Not when 1/3 of your team is heavily vested in LRM's, and not a single ECM among them. BAP/CAP you say? Again, matchmaker made a bad match worse by providing the perfect storm - scores of ERML/ERLL/Gauss/ERPPC for the red team. Everyone knows 3/3/3/3 > balanced global loadout.




The previous match was also unusual - it pitted three DWF's (in a single lance, no less!) against our 3 strong-men - a meltmaster, a stinker, and warhobbit. Please note the 3 of them together valiantly score just 100 more points of damage than the DWF-B.



Stay classy, matchmaker.

[beerandasmoke]

Second match your assualt lance scored 6 kills between 3 of the mechs. Thats not bad and shows that they were competent pilots. It looks to me like your heavies and meds failed to support them. Look at the bravo lances numbers and your own lances for that matter. It seems to me by the pic your assualts engaged charlie lance and wiped them out while the rest of the team got melted by direwolfs. Enemy team also had two mechs who scored a grand total of 8 dmg between the two of them. Just seems like a failure to deathball instead of MM in that game.

Everyone gets stomped op. It happens and its just best to click battle again and jump in another game.

[Monkey Lover]

Here one from when we were having the server problems few months back hehe

Edited by Monkey Lover, 08 December 2014 - 08:22 PM.

[Sandpit]

Hydrocarbon, on 08 December 2014 - 06:59 PM, said:

Ever get that feeling matchmaker's decisions are set to cull ELO than balance it? Ever think you could get a better W:L ratio if only you didn't get those 7+ losses in a row every other day?? Were you like me and got a record 10 losses in a row during the Free Victor event???

I'm curious if anyone else has some screenshots of what can only be described as a horrible failure in matchmaker algorithms. In the following pic the opfor has FOUR mechs with ECM. That can be delt with right? Not when 1/3 of your team is heavily vested in LRM's, and not a single ECM among them. BAP/CAP you say? Again, matchmaker made a bad match worse by providing the perfect storm - scores of ERML/ERLL/Gauss/ERPPC for the red team. Everyone knows 3/3/3/3 > balanced global loadout.




The previous match was also unusual - it pitted three DWF's (in a single lance, no less!) against our 3 strong-men - a meltmaster, a stinker, and warhobbit. Please note the 3 of them together valiantly score just 100 more points of damage than the DWF-B.



Stay classy, matchmaker.

uhm
since when is your random teammates picked by your Elo bracket mech builds an Elo issue of the MM?

not to mention you really just tried to portray a conquest game where damage is secondary to capping points (proven by the fact that the game was won by cap points) an indication of "My team was "bad" because they didn't score enough damage"?

[MikeBend]

To be honest, it is not that uncommon to see 3 of assaults die with damage dealt under 200, or even 100. I play in PST +10 hours time zone, and every evening is the same. There are too few even games, mostly we get steamrolled, or 1 out of 5-6 matches we steamroll. Really pathetic to see assaults and heavies die with under 200 dmg. With my level of play i just cant pull off a win in Huginn or Spider, when it is 1 vs 8, or even vs 5. Not sure its MMs fault, probably just that black cat running around my house.

Edited by MikeBend, 08 December 2014 - 10:44 PM.

[Funkadelic Mayhem]

Ya I got sick of seeing that.


Since the 1st time since Jan 17th 2013 after a reformat MWO was not the 1st thing I dll. In fact. It wont be untill I see some REAL light love.

[TKSax]

Hydrocarbon, on 08 December 2014 - 06:59 PM, said:

I'm curious if anyone else has some screenshots of what can only be described as a horrible failure in matchmaker algorithms. In the following pic the opfor has FOUR mechs with ECM. That can be delt with right? Not when 1/3 of your team is heavily vested in LRM's, and not a single ECM among them. BAP/CAP you say? Again, matchmaker made a bad match worse by providing the perfect storm - scores of ERML/ERLL/Gauss/ERPPC for the red team. Everyone knows 3/3/3/3 > balanced global loadout.

MM does not take mech builds into consideration, all it takes is W/L ratio and tries to balance that by weight class. Your complaining about things that match maker does not do.

Hydrocarbon, on 08 December 2014 - 06:59 PM, said:

The previous match was also unusual - it pitted three DWF's (in a single lance, no less!) against our 3 strong-men - a meltmaster, a stinker, and warhobbit. Please note the 3 of them together valiantly score just 100 more points of damage than the DWF-B.
Stay classy, matchmaker.

Hmm All assaults have been in the same lance for a while now(As they were on your team), that was a change made to the match maker a while back to group mech by weight class as much as possible with 3/3/3/3.

Also score is not necessarily indicative of steam roll. I have been in some great matches that end up being 12-2, 12-1 and even 12-0 , because of an advantage gained by the other team just start to steam roll, As you lose mechs that just ramps up the TTK as more people are able to focus on each mech. I would be nice if the after match screen show the percentages of the other mechs, which would let you know how much of a "roll" it was.

[Karl Marlow]

Ok you have the matchmaker pics down but we need more steamroller.

[QuantumButler]

I really wish I'd saved a screenshot of this one game where my team had no assault mechs, heaviest thing was a Orion, and the enemy team had 3 direwolves and 3 mad cats.

[R Razor]

QuantumButler, on 09 December 2014 - 01:16 AM, said:

I really wish I'd saved a screenshot of this one game where my team had no assault mechs, heaviest thing was a Orion, and the enemy team had 3 direwolves and 3 mad cats.

Been seeing quite a bit of this lately, and in some cases the search takes less than 10 or 15 seconds, so either the release valves are stuck open, or, more likely, there just aren't enough people playing the game anymore.

[Lily from animove]

defeat is not always a steamrole, not even in a 12:0, its a bit about coordinated fire. if one team has actually a kind of natural fire coordination they can simply steamrole the other even if the skills are very close.

[YokiLin]

770tons vs 710tons
Time: 2014-12-08 - 2am

Edited by YokiLin, 09 December 2014 - 03:48 AM.

[Willard Phule]

Lily from animove, on 09 December 2014 - 03:22 AM, said:

defeat is not always a steamrole, not even in a 12:0, its a bit about coordinated fire. if one team has actually a kind of natural fire coordination they can simply steamrole the other even if the skills are very close.

Yeah, but, come on....you know as well as I do that the whole Matchmaker/Elo system has been broken since the moment they added it as a "temporary fix" until they came up with something better.....over a year and a half ago.

We could go through each and every little thing that's wrong with it again...it's been hashed over here about a million times...but, ultimately, the problem will never be fixed until they figure out how to actually separate players based on their skill, not having lower Elos be used to "even out" the high ones.

People can whine about tonnage, equipment and whatever....but, ultimately, it comes down to playing with people that are close to the same skill level you are. If you're playing with people that are significantly better than you, you're going to have a bad experience over and over again. If you're playing with people that are significantly worse than you, you're going to be frustrated repeatedly. Just fix the damn Elo system to something that's more stratified and less fluid...and you'll see a change.

[Joseph Mallan]

QuantumButler, on 09 December 2014 - 01:16 AM, said:

I really wish I'd saved a screenshot of this one game where my team had no assault mechs, heaviest thing was a Orion, and the enemy team had 3 direwolves and 3 mad cats.

I've also seen my team roll the assault heavy team quite a few times. True I almost always drop in an Assault, but that doesn't mean I haven't been on the lighter team and still win.

[Mercules]

Funkadelic Mayhem, on 08 December 2014 - 10:49 PM, said:

Since the 1st time since Jan 17th 2013 after a reformat MWO was not the 1st thing I dll. In fact. It wont be untill I see some REAL light love.

Give it a rest. I am a light/Medium pilot and Lights are okay right now. I run Commandos, Firestarters, Locusts, and Ravens and they all work.

[Joseph Mallan]

Mercules, on 09 December 2014 - 06:19 AM, said:

Give it a rest. I am a light/Medium pilot and Lights are okay right now. I run Commandos, Firestarters, Locusts, and Ravens and they all work.

Our Mileage definitely varies Merc!

Edited by Joseph Mallan, 09 December 2014 - 06:24 AM.

[TKSax]

Willard Phule, on 09 December 2014 - 04:05 AM, said:

Yeah, but, come on....you know as well as I do that the whole Matchmaker/Elo system has been broken since the moment they added it as a "temporary fix" until they came up with something better.....over a year and a half ago.

FYI you do know that they completely rewrote the matchmaker after private matches came out last year.

[Mister Blastman]

OP: You were in a blowva. How can you carry harder in one of the worst mediums in the game? The 'mech is junk.

Jokes aside (maybe I was being serious?), the 12-2 stomp happens far too frequently lately. More often than not I'm on a team where 1/2 (6 players) to up to 8 players do less than 100 damage (or at least less than 150). I'll be walking along, shooting people and minding my own business like a good little mechwarrior and all of a sudden look around and notice I have no teammates left. It is like they spontaneously combusted or something and eight of them have bought the farm.

However, in the majority of 12-2's I'm usually the only one with about 500 damage and the rest of the team is less than 240 with six to eight less than 125-150 (at least four to six are sub 100). It is pretty screwed up but that's just how it is.

What's your Win/Loss ratio, OP, in the PUG queue? Mine lately for PUG only games is around 61% (which is a good bit less than when I group queued a lot). I'm trying to get it higher. The only way is by carrying harder.

[Bront]

I don't PUG a lot, but the group queue can be murcurial. It happens.

In PUG matches, I find I win quite a bit more than I lose. I don't know if that means I'm a good PUG player, I'm getting great teammates, the enemy team is full of derps, or I've been lucky. I still lose, and still occasionally get rolled. It just happens.

In generall, a 12-4 win/loss is quite often a close game, and the difference is simply if your team manages to get that extra mech or two to drop or not. Once the sides start getting unbalanced, it gets harder to turn things around. This is mostly why it drives me nuts when teams hang back with a 4 mech advantage. When you have that advantage, closing in and finishing the job wins matches, while continuing to stand back and trade fire is a great way to let the other team come back as you aren't taking advantage of your numbers.

[-Natural Selection-]

10 guys drop against the same 10 guys 3 maps in a row. Results? 12/1, 2/12, 12/6.

2 of those matches would be deemed a steam roll. Now for the 2 random players who dropped in on each side and not knowing the results of all 3 matches may see this as a MM fail. I see it as they were not expecting to get rushed the 1st time, expected it the 2nd, and though we were going to do the same thing on the 3rd.

Thing is, sometimes people remember another groups strats and sometimes they misjudge what they think they are going to do. Either way,, MM can only predict outcome and not human behavior.