148 replies to this topic

[Davers]

Targetloc, on 13 March 2013 - 08:33 PM, said:

This was pretty damn funny.

[Lyrik]

Mackman, on 13 March 2013 - 08:27 AM, said:

So you're saying that League of Legends did not become one of the most played games in the world by relying on Elo matchmaking? You're saying that they've been using an ineffective matchmaker for... what is it now, three or four years at least?

Every game will not be close: You're right about that much. But Elo will result in each player getting more equally-matched games on a consistent basis.

Also, am I to understand from your last comment that you don't want any matchmaking at all? Because if so, then I can rest easy knowing that your opinion on how to make competitive games fun for everybody is effectively worthless.

I want matchmaking. I'm just saying that no system is foolproof ;-)

[KuruptU4Fun]

You have to realize though that ELO matches you with a RANGE of others, anywhere from 100 to 300 as an example. If it can't find people close to your ELO then it widens that range up by a larger percentage maybe it opens up to 300 to 1000. We don't know the details, but that's a fairly good assumption. Mind you that's on BOTH sides of the scale. if you're at 1200 ELO then the MM system has to open up to find people who may have 900 to 1700 to fill the gaps. This is also over hundreds of games at any given moment.

[Mackman]

Lyrik, on 14 March 2013 - 01:42 AM, said:

I want matchmaking. I'm just saying that no system is foolproof ;-)

No argument from me there. Sorry for lumping you in with the people who think that matchmaking is somehow an insult to their "skills."

Is Elo foolproof, especially in an environment where you can't control your teammates and rely on them to win? Of course not. But League of Legends has demonstrated fairly conclusively that Elo can be extremely effective at grouping people of similar skill, for close, fun games most of the time. Not all of the time, because randomness still happens, but most of the time.

Given their demonstrated success, I think PGI made an excellent choice: My matches are much more fun and close than they were before Elo, and I anticipate even better things once they make their first promised adjustments.

[buttmonkey]

i agree about the competitions for sure, people will exploit this system to get easy mode during competitiions. there must be a better way than elo, why open the game up to exploits

[Stoicblitzer]

Livewyr, on 13 March 2013 - 08:31 AM, said:

Here's how to test if ELO is working.

Play a game where you're at.. note the competitiveness of the play (and the outcome, but that isn't as important)
Then make a new account and grab a trial mech.. note the competitiveness of play there (and certainly not the outcome)

There is a drastic difference.

Landslides are being caused by distribution of a few things: ECM, Missiles, Ravens, AC20 Cats, Splatcats, PPC Snipers, Jump Snipers, and things like that..

qft

Tickdoff Tank, on 13 March 2013 - 11:00 AM, said:

IMO, Elo alone will not be sufficient to balance the MM in MWO. We need to have some sort of a "BV" (Battle Value) system which works with the Elo system.

A pilot's effectiveness in a match is not determined purely by skill, His choice of mech and weapons also plays a part. Ignoring either of these aspects will cause problems.

We will not be able to use the base BV system from TT. The weapons and mechs work much differently in MWO for the TT BV system to be accurate. In order to develop an effective BV for MWO you will have to look at more than just the base numbers for a weapon or equipment. As a previous poster has stated, in TT all weapons fired at the same rate, all weapons did their damage instantly, and all weapons hit random locations. That is not how MWO works, so we must develop our own BV metric, one which would give a much higher value to a SSRM2 than an AC2. (In TT the SSRM2 has a BV of 30 and the AC2 has a BV of 37)

qft

[Stoicblitzer]

Adrian Steel, on 13 March 2013 - 08:12 PM, said:

I can corroborate what the OP is saying.

I often run a 2 man pre-made with my brother, Ryan Steel. We're of comparable skill. We coordinate on comms. We can be matched with bads on our team that do everything wrong, while the other team appears to be chock full of decent players.

When we each solo pug, there never seems to be the same disparity. Every bad on our team correlates to a bad on the other team.

We have a sneaking suspicion that the match making code is different for pre-mades than solo pugs, and that this is most noticeable in non-primetime hours where the population at the extreme ends of the Elo bell curve will be minimal. It's very possible that pre-mades are assumed to have a communication advantage in the match making code, and are purposefully matched with team-mates of a lower Elo in order to compensate.

If you're incredulous at this claim, ask yourself how many modifications to the game haven't been in the patch notes... There are entire threads devoted to discussing these easter eggs.

qft

[Vodrin Thales]

Glythe, on 13 March 2013 - 08:03 AM, said:

If you have 4 good players and 4 players who might as well not exist then the game is more like 4 on 8 and I really don't think the best 4 players of this entire game could take on those odds. I guarantee you that 8 average players with mid-low Elo will trump the top 4 players. Why? because they're smart enough to realize if they are only fighting 4 people those guys are going to be fantastic.

This is the OTHER valid Elo complaint.

It's really not. My groups of 4 that I play with are running a 75% win percentage or better when I am playing with vets from my group, and around 50% when I fill up with relatively green members. If you are having that much trouble carrying 4 pugs to wins they are not the problem. The problem is that ELO is frequently matching you against players like me.

[LT Kinslayer]

My guess is that your team isn't as great as you think it is.

Let's pretend the system works as you say:
Your team is gods gift to Mech Warrior Online and has an Elo of 2700 and you get four Elo 900 puggers to fill up your game.
Your opponents all have 1800 Elo.

Sure, you could argue that those 1800 guys would wipe the pugs easily, but that's where the argument ends. If your team is 2700 in a voicecom environment, you should still be able to wipe all the Elo 1800 guys without breaking a sweat. The skillgap between 2700 and 1800, if Elo is done correctly, should be huge. With the 2700 guy being a full plate Knight and the 1800 guy being the farmer that has just been drafted into the army two weeks ago.

(the Elo 900 guys would be the boys with a broken sling.)

[Karl Marlow]

Mackman, on 13 March 2013 - 08:27 AM, said:

So you're saying that League of Legends did not become one of the most played games in the world by relying on Elo matchmaking? You're saying that they've been using an ineffective matchmaker for... what is it now, three or four years at least?

Every game will not be close: You're right about that much. But Elo will result in each player getting more equally-matched games on a consistent basis.

Also, am I to understand from your last comment that you don't want any matchmaking at all? Because if so, then I can rest easy knowing that your opinion on how to make competitive games fun for everybody is effectively worthless.

This is not LoL. LOL goes to great lengths and pains to make every single one of the characters "balanced" Every character is supposed to have a 50/50 chance of being able to defeat any other characters. The difference resting only with skill. In MWO we do not have that. The Devs are making no effort in balancing one mech against another. All they are trying to do is balance by some nebulous value called "skill" while throwing us into wildly different mismatched scenarios.

I know I keep belaboring this point but it is my number 1 bugaboo about the ELO system. I don't care if you are trying to match me up with some kind of mystic ELO number you pull out of your crystal ball. What I do care about is a level playing field to start with. either by weight class, tonnage, or BV. That has to be the priority when putting together the match. I never got frustrated with the matchmaker before ELO came out. Yeah sometimes I would get stomped 8/0, Never did I feel cheated when I lost. even when people were allowed to run 8 mans versus pugs I never felt cheated when I got steamrolled by an 8 man. It was fun trying to take them down, and I think we sometimes were able to.

It is only after ELO has come out and we continually get rolled by a group that out tonnes us by 200 tons that I've frankly gotten sick of even dropping. I've not quit but I don't play anywhere near as much as I did. It feels cheep when I'm on the heftier side and It's annoying bordering on boring when you are on the smaller side. Now on those rare instances where the match is almost even then yes, the game is still fun. I just find it rare to drop into a match where the deck isn't stacked against you from the start. Maybe I am just unlucky but this is why sorting by weight or weight class is far more important than sorting by ELO. I understand the need to separate the pros from the little league players. All I want is for both sides to be playing the same game at the start.


EDIT: yes I know sometimes you can win while under tonned. I would point out that you should only be able to do that through skill which is supposed to be equal with ELO or by luck which isn't something we should be relying on. I also know that it is sometimes to your advantage to be lighter. Even so if you are at an advantage having the smaller force on a map then you are still stacking the deck against the other team.

Edited by ThomasMarik, 14 March 2013 - 10:49 AM.

[MaddMaxx]

Davers, on 13 March 2013 - 07:50 PM, said:

But how is the target value found? Is it based on the first guy to hit 'launch'? Does it make a whole team, THEN tries to find the equivalent so the first team might have wildly different Elo's?

Think of it like picking Teams on the school yard. The ELO looks at the available Pool of picks, then selects one player for team A, then a player for Team B.

Like the Captains in the school yard who pick, they also know the Pools over-all rating and each players level. So, if done, at least half reasonably, both Teams will end up with the same "over-all" average competence. Not all super stars and not all pigeon toed losers...

Well OK, except that Mouse chewing guy in the 900 block. lol

P.S. I played CoH and it had a Ranked system. Even there the system had to allow you to go up or down 3 brackets in order to not have to wait forever to get a game. There has to be some flex. To tight and no ones gets to play against anyone other than those of the exact ELO, the same gaming TimeZone and playing schedules.

Edited by MaddMaxx, 14 March 2013 - 10:56 AM.

[Glythe]

Lyrik, on 14 March 2013 - 01:42 AM, said:

I want matchmaking. I'm just saying that no system is foolproof ;-)

Last matchmaker was fair. This one is from what I see 90% unfair to me and several people I play with (grouped or independently). Random is a fairness all its own in that over a long enough timeline you actually do get pretty good matches. You might get 2 or 3 in a row that aren't fun but you can have a long string of fun games through independent assortment.

buttmonkey, on 14 March 2013 - 09:26 AM, said:

i agree about the competitions for sure, people will exploit this system to get easy mode during competitiions. there must be a better way than elo, why open the game up to exploits

The Elo MM is the worst we have had so far. I honestly would rather fight 8 man groups alone than this mode because at least the weights matched in the 8 man era. The only MM worse than that one was the initial system that just threw 8 people in a game with no regard at all.

Current MM is broken when you stack specific weight classes. The best team is going to be 4 lights and by that I mean Ravens (you may substitute with very fast mediums w/ SRMs) + 4 CTF (using poptart build) or 4 assaults. Problem is smart people stack weight classes because it wins more.

Throwing Elo out the window even a very good team in the wrong build can't beat a bad team with a stacked mech loadout. The fast movers will kill anything not in a group of 8 and the assaults have firepower and health to kill everything the others can't.

Recently you see teams with 6 mediums (trebs) against something stupid like 6 assaults. Good luck with that one.

[FerretGR]

Glythe, on 14 March 2013 - 12:35 PM, said:

Last matchmaker was fair.

You mean the random matchmaker? Please expound on how putting randomly grouped newbies in trial mechs against golds in tweaked 3Ls and Splatcats and DDCs in teams is fair.

Are you just being nostalgic for the days when the wins came easy and the c-bills flowed?

[Bilbo]

OneEyed Jack, on 13 March 2013 - 10:45 PM, said:

I could be wrong, but I was under the assumption that they hadn't started the class-based ranking, yet. Of course, it could explain a lot about some inconsistencies in my pairings....

You do indeed have a seperate Elo ranking for each weight class. The difference in competition when I jump into my light mech(you would think by now I'd know it would be better to just leave the ******** in the garage) versus any of the others is startling to say the least.

[Davers]

Glythe, on 14 March 2013 - 12:35 PM, said:

Last matchmaker was fair.

You mean to say "Having no matchmaker was fair." I disagree. When they further refine Elo (like they said all along they knew they would have to do) I think you will agree with me.

[Theevenger]

BillyM, on 13 March 2013 - 08:05 AM, said:

Your friend realizes that his ELO score cannot go down, right? All he has done is wasted a bunch of time for nothing but to negatively impact his team as a whole and ruin the game for everyone.

--billyM

Yes it can. Your ELO score will go up or down depending on what the actual result is compared to the predicted result. If you do better than expected, it goes up. If you do worse than expected, your score goes down.

[Zeh]

Theevenger, on 14 March 2013 - 12:58 PM, said:

Yes it can. Your ELO score will go up or down depending on what the actual result is compared to the predicted result. If you do better than expected, it goes up. If you do worse than expected, your score goes down.

I would clarify this a bit.

If you WIN it goes up. (If you weren't expected to, it goes up more)

If you LOSE it goes down (If you were expected to win, it goes down more)

[Lyrik]

Whatever our complaints are, the devs already said that current matchmaking isn't the final version. It will be adjusted and will take a greater emphasis on tonnage. So the feedback of the community didn't fall on deaf ears ;-)

Tournament team drops need also a restriction on maximum tonnage. An Atlas is of course better than an Awesome. It has 20 more for armor, weapons and equipment. And with nerfs/buff like maximal turn rate etc you can't balance them.

[Glythe]

FerretGR, on 14 March 2013 - 12:43 PM, said:

You mean the random matchmaker? Please expound on how putting randomly grouped newbies in trial mechs against golds in tweaked 3Ls and Splatcats and DDCs in teams is fair.

Random match maker was better than you think. Why? Weight is a huge factor in this game and if you believe otherwise you're either a liar or deluding yourself (with no consideration of base capture DDC should always defeat a Raven in combat). ECM is a huge factor as well and it has a much larger impact on the game than even weight right now because it really is that powerful.

I propose that matchmaker should be as follows:

1. Total team weights should be within 10% of each other (dont need exact match but close enough).
2. ECM should be balanced on both sides within 1 unit increments (games will not form with 1 ECM vs no ECM).
3. Elo should be a slight consideration..... perhaps 20% of the whole (compared to #1 & #2)


Overall I saw the majority of random matches being much closer than the Elo system. Most of the Elo ranked games are 7:1 and end with complete and utter steamroll. Even with 4 strong players vs 8 random players the games were very rarely 7:1 or worse. Most random matches ended with at least 3-4 kills for the losing side unless it was the occasional steam roll (which was about one in ten games... the exact opposite of current situation).

Glythe, on 14 March 2013 - 12:35 PM, said:

Last matchmaker was fair.

Quoted out of context for brevity..... I was saying last MM was MORE fair than the current because weight disparity throws off the balance for more than ANY other factor (besides ECM).

4 Assaults+4 Ravens will Always beat an equal ranked Elo team of mediums (or mostly medium team with 2 mismatched weights). Remember that the first team can have up to 8 ECM while the 2nd team will be doing well to have 1 ECM mech.

4 Ravens+ 4 Heavies ( Cheeseapult of any flavor/ CTF cheese [ballistic or poptart: 2x ERPPC+Gauss]) will always beat 8 mediums.

This is my point. Both matches occur often and it's a guaranteed loss when it happens. Random match maker based on weight was more fair. Note for both scenarios you may substitute a 2D trollmando for a Raven.

Conquest balance was completely destroyed when the number of lights and mediums on one team no longer equals the other team. I LOVED conquest because it forced people away from camping and sniping (far too easy with PPC). It forced them to move to the fight rather than forcing the enemy to come to their super fortified position. Conquest also forced build diversity because you can't sit in a group of 8 unless you want to lose.

Conquest is completely unplayable right now because you can bring four lights and the enemy team might have none. Lights are great at capturing bases but they also excel at swooping in on one enemy and killing him before help can arrive (and then you repeat this process). In the most extreme example of weight diversity 8 lights will always beat 8 assaults for conquest. Lights can split up and capture every base. Assaults cannot split up or they are at risk for getting swarmed.

Current match maker completely destroys the only game mode I liked to play.... as such I think it's horribly broken. Most people do not like or do not play conquest so they do not care.

Edited by Glythe, 14 March 2013 - 03:16 PM.

[FerretGR]

I do agree that tonnage/something else needs to be considered, and I've always said as much. But I really, really disagree that, even with its issues, the current matchmaker is a step down from the random matchmaking we had before. I don't recall the matches as being anywhere near as competitive as they are now, even with the occasional weight discrepancy. I don't know how often you PUGed in the former situation, but it was unpleasant at times, and it was really rough on beginners. This is a step up from that. In my humble opinion, of course. A combined Elo and accounting for tonnage or using some sort of BV-like system will be a further step up. Optimism!

Edited by FerretGR, 14 March 2013 - 05:37 PM.