182 replies to this topic

[Duke Nedo]

Attrition? Just urban myth, Russ said so.

J/k, hard to resist.

Attrition is one thing, at least on my behalf I have no problems with matches ending in big numbers to one side. That is like you say, almost to be expected. But, there is a clear difference when you lose 2-12 and your team actually fights well, and 0-12 when they all run around like chickens away from the fight instead of joining the fight, leaving the assaults to die and ignoring team chat.... until after they are dead, then they blame you for not having learnt to play indy500 in your Stalker...

The latter is not attrition, it's just imbalance...

[Quaamik]

Honest question here to each of you:

Do you really care about your (random grouped) teams wins / losses?

To put it another way: Would you prefer mostly winning games but being matched against clearly inferior opponents or somewhere close to 50/50 win/loss matched against opponents of equal skill?

[Joseph Mallan]

Apnu, on 07 December 2014 - 07:59 AM, said:

During the Victors challenge I tracked my stats accross 60 games. Took 56 matches to unlock the Victor. I had a KDR of 2.09, 550 average damage per match -- all in the same mech, the HBK-4J in the solo queue. I did that to test the MM vs my skill.

MM needs work. Or we need lobbies based off Elo rating.

So lets see you get 2 kills per every one death. That does not mean you don't die after killing 2 mechs in a game. Does it? The rule was 1 Kill, 1 Assist, win, Stay alive. That's the challenge from the last Challenge You have to get ALL 4 requirements to earn a point. Its why I stayed out. My KDr is 1.0 in may Atlas-S, So I die just as often as I kill.

Duke Nedo, on 07 December 2014 - 08:02 AM, said:

Attrition? Just urban myth, Russ said so.

J/k, hard to resist.

Attrition is one thing, at least on my behalf I have no problems with matches ending in big numbers to one side. That is like you say, almost to be expected. But, there is a clear difference when you lose 2-12 and your team actually fights well, and 0-12 when they all run around like chickens away from the fight instead of joining the fight, leaving the assaults to die and ignoring team chat.... until after they are dead, then they blame you for not having learnt to play indy500 in your Stalker...

The latter is not attrition, it's just imbalance...

That difference is Teamwork, and it doesn't happen often enough in PUG Que.

[Charronn]

Quaamik, on 07 December 2014 - 08:03 AM, said:

Honest question here to each of you:

Do you really care about your (random grouped) teams wins / losses?

To put it another way: Would you prefer mostly winning games but being matched against clearly inferior opponents or somewhere close to 50/50 win/loss matched against opponents of equal skill?

Not really as my k/d ratio is well over half so i'm not much fussed,I just don't like the idea of some mechanism working in the background making you get the crap team time after time to balance your stats out.

[Quaamik]

Charronn, I'm guessing from that response it's the second: that your rather play quality matches against equally skilled opponents (win, lose or draw) than win constantly against weak opponents.

I think that's the case for most of us. I also think that's the frustration people have with the matchmaker. It's not measuring / balancing based on the right criteria.

I'd be happy if they scrapped the ELO and went to a forced ranking. Consider number of matches played overall, number in that weight class, number in that mech chassis, overall hit percentages, overall kill/death ratio & kill/death ratio for that weight class. Use the overall numbers for a gross ranking, with the weight and finally chassis numbers used to refine position. Match players so they normally don't play outside of a +/- 100 ranking in PuGs (might have to average rankings for teams). If it can't find you a match within that +/- 100, either drop you in a less than 12v12 or give you the option to go up (not down) the rankings.

The +/- 100 might need to be refined wider, depending on how many players are normally online, but you can get the idea.

Edited by Quaamik, 07 December 2014 - 10:08 AM.

[generalazure]

After skimming 5 pages of confusion, I have the distinct feeling most players do not know how an Elo rating system actually works. Yes, it produces streaks. No, it does not actively try to fix matches. It is meant to keep you mostly winning/losing until your Elo score matches your actual skill. You do not get a binomial p=0.5 distribution unless all involved players have the skill matching their score (that's pretty much never), so it cannot be modeled like a damn toin coss.

Let me give you an example of where streaks come from: Say you're playing a light mech on a weekend where they are unpopular. Since the matchmaker currently has trouble finding a better candidate than you, you're placed into a group where your Elo is significantly below average. Now assume, because the other guys in your team actually know their stuff (and your single light mech just didn't matter all that much anyways), your team wins.
Since you've just beat an enemy with a significantly higher Elo score, your own score goes up a lot. It's what Elo does, if your opponent is rated stronger than you (and it was, since you were below team average), you get more score from a win.
Guess what happens next? You're now overrated and the matchmaker places you against people of similar score... which will crush you more often than not, because you didn't really earn yours. But since the difference in scores is not that great now, for each match you will not lose as much Elo as you gained during that one match where you were the outlying score. So it takes quite a few losses to get your score back to where your skill says it should be. And that's pretty much the definition of a streak.


TL;DR: Matchmaker working as advertised, even with streaks.

[Duke Nedo]

generalazure, on 07 December 2014 - 10:15 AM, said:

After skimming 5 pages of confusion, I have the distinct feeling most players do not know how an Elo rating system actually works. Yes, it produces streaks. No, it does not actively try to fix matches. It is meant to keep you mostly winning/losing until your Elo score matches your actual skill. You do not get a binomial p=0.5 distribution unless all involved players have the skill matching their score (that's pretty much never), so it cannot be modeled like a damn toin coss.

Let me give you an example of where streaks come from: Say you're playing a light mech on a weekend where they are unpopular. Since the matchmaker currently has trouble finding a better candidate than you, you're placed into a group where your Elo is significantly below average. Now assume, because the other guys in your team actually know their stuff (and your single light mech just didn't matter all that much anyways), your team wins.
Since you've just beat an enemy with a significantly higher Elo score, your own score goes up a lot. It's what Elo does, if your opponent is rated stronger than you (and it was, since you were below team average), you get more score from a win.
Guess what happens next? You're now overrated and the matchmaker places you against people of similar score... which will crush you more often than not, because you didn't really earn yours. But since the difference in scores is not that great now, for each match you will not lose as much Elo as you gained during that one match where you were the outlying score. So it takes quite a few losses to get your score back to where your skill says it should be. And that's pretty much the definition of a streak.


TL;DR: Matchmaker working as advertised, even with streaks.

Sorry, that doesn't create streaks afai can see. By the same logic in your example, that you can get carried in a light to a win against a better team and gain elo, you could be almost exactly equally carried at a higher elo by your high elo teammates. What we are seeing on loss-streaks is not that one person has suddenly become a liability and cause such drag that the team must loose because they are fighting 11.5 vs 12.

No, it's more like you have win streaks where you will win without trying, followed by loss-streaks where you will loose no matter how hard your are fighting.

I know I have no reason to say it IS like this, but it FEELS like the MM gives you some 15-20 games where you have the chance to gain elo by facing better teams followed by 15-20 games where you have the chance to loose elo by facing worse teams. That's how it's perceived. I don't think the MM does that on purpose, it would actually be non-trivial to set up, but whatever it does makes it feel that way for many players in the solo queue. If there is any way to find out what is causing it and fix it it would be wonderful. It can be soooo frustrating.

[Sandpit]

LordKnightFandragon, on 07 December 2014 - 05:06 AM, said:

If MMs dont stack the deck, then why in normal pug play are there literally no players with less then about 45% win and no players with over 60 some % WR? In WoT, the only ones with massive win rates are those who Clan wars, where they literally do stack their own decks and can be in a position where they literally cant lose. But in pubbing, you dont see even the absolute worst player with less then maybe 45% win rate and even really good players dont have more then maybe 55%. Only in very rare occasions do you see a player with less or more then that, and its just 1 or 2 players every like 20 games.

If the MM didnt stack the deck, then the good players, even in solo pubbing would be able to absolutely dominate the bads and would have higher then average win rates, while the bads would be 20-30% or worse Win rates. BUt its always right around 45-55%. Ive sat and watched my dad in WoT over 13K battles using the XVM mod, and the trends are a disturbingly obvious. He gets a really good streak, where his team gets the greens, the unicums and the deck gets stacked in his favor, he has to do little to nothing and wins just drop in his lap, then like the last few days, its been him, the only green, then rows and rows of terrabad reds. Record the last few days is like 6/30+.

MM Doesnt stack the deck...riiiight.

Yea?
Interesting. ..

Where's your statistical data showing most?

[LordKnightFandragon]

Sandpit, on 07 December 2014 - 12:09 PM, said:

Yea?
Interesting. ..

Where's your statistical data showing most?

You could probably look up the game's average win rate and see for yourself.

WoT's was like 48% last time I knew. Only ones who are really over the top good are ones who play mostly fixed clan wars and the really terrabads, well they are prolly little kids who have no sense of anything or bots.

Edited by LordKnightFandragon, 07 December 2014 - 12:30 PM.

[generalazure]

Duke Nedo, on 07 December 2014 - 11:07 AM, said:

By the same logic in your example, that you can get carried in a light to a win against a better team and gain elo, you could be almost exactly equally carried at a higher elo by your high elo teammates.

The possibility to keep on winning despite being overrated is of course always there, but it should get more and more unlikely. The system is supposed to drag you down to your actual skill level and the only way to do that is to lose quite a few games.

But perhaps my example was not quite perfect, I was using just 1 guy as a score outlier in an otherwise balanced group (and for only one match) to keep things simple. In reality, you'll have slightly different scores for everyone, in every match, with the variance depending on how many suitable players the matchmaker can find quickly. The Elo system was never meant to deal with this spread efficiently (it was designed for 1on1), so imho it is not suprising if the results are bad.

My bet is that on average it still works, but you can get those weird streaks where you oscillate around your "correct" score. Maybe someone can dig up a paper on Elo in imhomogenous groups, there's gotta be something on this already...

[Davers]

Jman5, on 06 December 2014 - 09:58 AM, said:

There is one thing I would like to see matchmaker improve upon and that is Elo decay. I don't think players who have been gone from the game for a year should come back and find their Elo right back where they left it. I have seen this many times where players I knew used to be hotshots come back and frankly are incredibly rusty. Yet Matchmaker keeps treating them like nothing changed. Elo should steadily decay until it hits average.

Unfortunately this is exploitable. Better to force players to constantly make new smurf accounts, rather than just switching back and forth between 2 or three accounts.

Duke Nedo, on 06 December 2014 - 03:49 PM, said:

No need to be an ***. Like I wrote, I hope it's not like this. BUT, the MM creates more streaks than you would expect from a random series of coin-flips (which is what we should be having if every game was 50:50 chance of winning:losing before the match starts).

See, The Gambler's Fallacy. http://en.wikipedia....ler%27s_fallacy

Duke Nedo, on 06 December 2014 - 11:07 PM, said:

Ok mister, you really tick me off now. You blurt out "factual truths", when in fact you and me have access to exactly the same information. If you had been the coder of the MM you could make statements like that. Fact is, you don't know and I don't know anything more than what PGI has told us.. and that leaves out a lot of critical information. As long as we cannot use the information they have shared to fully reproduce the MM ourselves, they have not described how it works.

That does not make you enlightened or me uninformed. It just makes you arrogant and ignorant.

How Elo works is pretty open content. Lots of sites talk about it, and even give the percentage chance of winning based on the difference. For example, when the Clans won 65% of their matches with approximately 100 point Elo advantage, it matches up perfectly with their predicted chance to win. Thus proving Elo works in MW:O.

Duke Nedo, on 07 December 2014 - 02:18 AM, said:

Que????

http://mwomercs.com/...79-matchmaking/

This is an old post, but the prediction of the probable winner is central in Elo as far as I have understood this. That's how it changes Elo values after the match.

The bottom line is: Elo is really good for single or team ladder systems, It's about ranking players or team according to their ability to win. IF the matchmaker is doing a good job, both teams stand the same chance of winning, and thus the result cannot be predicted. There is no way to decide if the result should or should not reward a change in Elo.

If the aim of the MM is to create even matches and it can indeed predict the probable winner, then that match should not be launched. It should be rescrambled until no winner can be reliably predicted, then launched. Problem is, that breaks Elo.

So, it's actually not completely out of the blue to assume that PGI actually accepts a fair degree of imbalance because uneven Elo is needed to drive changes in Elo, and hence make Elo work. What I am trying to say is: if the outcome can indeed be predicted with any reliability, instead of launching that game and calculate Elo change from it, instead use that information to build a more even match.

Elo is the foundation of every MM out there. If there is a problem with the MM it is more likely due to the issues that PGI has brought up to fix things, things that the players demanded they not do. Which is iron out tonnage discrepancies and have players be flexible in game mode options. Basically players told PGI "We would rather have more imbalanced matches as long as we can drop in the mech we want, in the mode we want." PGI flat out said that for better MM to be possible one of those has to go.

Edited by Davers, 07 December 2014 - 01:20 PM.

[Jman5]

Davers, on 07 December 2014 - 01:19 PM, said:

Unfortunately this is exploitable. Better to force players to constantly make new smurf accounts, rather than just switching back and forth between 2 or three accounts.

In a typical game I would agree, but Mechwarrior Online is not a good game to smurf on. The cost of entry is too high to just replicate even a fraction of the resources a long-term account will have. Minor Elo decay is not going to be enough to get people to take such a big hit.

[Davers]

Jman5, on 07 December 2014 - 01:38 PM, said:

In a typical game I would agree, but Mechwarrior Online is not a good game to smurf on. The cost of entry is too high to just replicate even a fraction of the resources a long-term account will have. Minor Elo decay is not going to be enough to get people to take such a big hit.

How many people have alt accounts now? I know of quite a few. Now add in an actual in game advantage in switching accounts...

Now I personally would think it is silly, and it won't really have any effect on CW. But there will always be people who want that edge.

[Agent 0 Fortune]

For me it has gotten to the point where I have to be prepared to carry every match in the medium and heavy class. If I can't deliver 800 damage and 3+ kills (solo or most damage) we lose, it is that simple. And it usually takes a second person on our team doing to same. I can't remember the last time, I was on the winning team and the bottom half of the scroreboards in the same match.

I feel like I cannot purchase or play any new mechs, because I won't be able to carry those number while working on skill unlocks and figuring out a build. I probably wouldn't care as much if I felt like I was still making XP and C-Bill progress while being repeatedly violated every match.

Davers, on 07 December 2014 - 01:41 PM, said:

How many people have alt accounts now? I know of quite a few. Now add in an actual in game advantage in switching accounts...

You are actually forced to have two accounts if you want to play both clan and IS mechs during CW.

[Jman5]

Davers, on 07 December 2014 - 01:41 PM, said:

How many people have alt accounts now? I know of quite a few. Now add in an actual in game advantage in switching accounts...

Now I personally would think it is silly, and it won't really have any effect on CW. But there will always be people who want that edge.

I know a lot of people in this game and very few use alt accounts. A couple guys to temporarily get around bans, a couple for weird lore reasons, but that's about it. I could probably count on one hand the number of people I know who use an alt-account. It's not a problem because people don't want to re-grind all their mechs, cbills, modules. They don't want to purchase new mechbays.

[Davers]

Jman5, on 07 December 2014 - 01:49 PM, said:

I know a lot of people in this game and very few use alt accounts. A couple guys to temporarily get around bans, a couple for weird lore reasons, but that's about it. I could probably count on one hand the number of people I know who use an alt-account. It's not a problem because people don't want to re-grind all their mechs, cbills, modules. They don't want to purchase new mechbays.

Ok. Not trying to start a 'I know more people than you' thread here.

I know many games, not just MWO, where people make smurf accounts and trash noobs. Many times it's people who want to use their real money to give them advantages. Like in Neverwinter, where you often see low level characters running around with ultra powerful weapons in PVP one shotting everyone. Hell, I worked with this young guy for a while, and he spent more on that game in a month then I spent on MWO since Beta. Even he admitted it made the genuine new player's experience pretty crappy. But he did it, because he knew that others would be doing it too. So the whole 'players don't want to spend money on new things' isn't true for all players.

You don't think it is exploitable. But that is because you are probably a good guy who enjoys playing for fun. But as we have seen, every time there has been a bug, or an exploit, people FLOCK to it. It never occurred to me to not fully R&R my mech, and 'suicide grinding' just looked stupid to me. LOL, I had to have someone explain to me what they were doing. That's because I play for fun too. I don't calculate my Cbills/hour and spend my time looking at my stats page. But there are people that do. And those people will always look for an edge over everyone no matter how unscrupulous it might be.

[Davegt27]

Wins / Losses 2,969 / 2,965 (as of today)

[Razimir]

The bigger problem than MM is ping differences. Maybe PGI should put people with similar pings in the same matches. Usually games are fine around Europeans peat times, but when low ping US people gets online hit registration is getting very bad.

Like other day, I had very bad 270 ms ping and locust ran straight at me, he got 18 SRM missiles and ERPPC on his face, but nothing happened.

[GalmOne]

Bouncing between solo queue and group queue at a good enought high elo i noticed one thing

The problem is probably less related to elo itself and more about group coordination
For example: 2 teams of 6 players each move with alot more coordination than, say, 3 teams of 4 people each even if all players are at the same elo rate

What happens is that, if someone is leading properly, they will quickly sieze an advantage and before the other less coordinated team takes note of the situation will also loose some mechs and be unable to properly regroup and counterattack

If you notice in solo queue this is quite relevant and fights tend to be slower probably becouse everyone is just managing by themselves without any proper guidance

[Mystere]

LordKnightFandragon, on 07 December 2014 - 05:06 AM, said:

If MMs dont stack the deck, then why in normal pug play are there literally no players with less then about 45% win and no players with over 60 some % WR? In WoT, the only ones with massive win rates are those who Clan wars, where they literally do stack their own decks and can be in a position where they literally cant lose. But in pubbing, you dont see even the absolute worst player with less then maybe 45% win rate and even really good players dont have more then maybe 55%. Only in very rare occasions do you see a player with less or more then that, and its just 1 or 2 players every like 20 games.

If the MM didnt stack the deck, then the good players, even in solo pubbing would be able to absolutely dominate the bads and would have higher then average win rates, while the bads would be 20-30% or worse Win rates. BUt its always right around 45-55%. Ive sat and watched my dad in WoT over 13K battles using the XVM mod, and the trends are a disturbingly obvious. He gets a really good streak, where his team gets the greens, the unicums and the deck gets stacked in his favor, he has to do little to nothing and wins just drop in his lap, then like the last few days, its been him, the only green, then rows and rows of terrabad reds. Record the last few days is like 6/30+.

MM Doesnt stack the deck...riiiight.

The details of how Elo and MM work have been given by PGI. But if you insist on continuing your belief in this fantasy, then so be it.