69 replies to this topic

[100mile]

I love it when a match comes down to 12/11 or 12/10 makes for a really exciting watch....I hear a lot of talk about how "Fair" things should be...about tonnage etc...(coming soon by the way) ((soon being relative)) About how unfair the match maker is and yes like the rest of you i hate being on either end of a PUG stomp...boring...but i want you to ask yourselves how often it's tied 3/3 or 1/1 the match is sitting at 4/3 and you are several minutes in....when all of the sudden it's 7/3 or 8/4 or 8/2 and you are going WTF just happened...how'd we get rolled???
Fact is you reached the tipping point for your teams damage and turned into dominoes...the amount of fire power the other team can bring to bear increases exponentially as your team dies...This has nothing to do with being rolled or the other team being that much better or the tonnage limits etc....
Tactics have more to do with this than the matchmaker does...I have seen light teams totally destroy heavy teams and vice versa and everything in between....

My point is the match maker can't account for one very important thing.....The human factor...and the fact that once a team gets to a certain damage point it's chances of winning become way less and that is a fact of battle real or play.. These 2 things have more to do with the "unfair" play than the ELO matchmaker does...

[Roadkill]

Kay Wolf, on 11 December 2013 - 10:40 AM, said:

So, Roadkill, what you're saying is, the matchmaker and Elo actually work counter to one-another? If I have that right, what is the point of doing so?

I'm not sure what you mean by that. Elo is just a computation of your skill level, so it doesn't really work counter to anything. It just computes your skill level.

The matchmaker tries to pair up teams of approximately equal skill level, which it does by adding up the Elo scores of the players on each side.

If the teams are roughly equal - which is the goal - then the Elo delta will be the k-value and the predicted winner is mostly irrelevant. You can think of this being caused by the confidence being low - when the Elo values are similar, the prediction engine doesn't really know who's going to win.

On the other hand, if the Elo values are wildly different then the prediction engine can be reasonably confident about who's going to win. If it is right, then the Elo ratings don't change much because the expected happened. But if it is wrong the Elo ratings can change by a lot - up to double the k-value.

Does that help?

[Davinelulinvega]

hm...
so when the elo system tries to get your w/l ratio to about 1, does it put you in matches you're gonna loose on purpose to even it out if you had a winning spree?
sometimes feels like it...

[Kekkone]

Davinelulinvega, on 11 December 2013 - 01:20 PM, said:

hm...
so when the elo system tries to get your w/l ratio to about 1, does it put you in matches you're gonna loose on purpose to even it out if you had a winning spree?
sometimes feels like it...

No. Elo doesn't try to get your w/l anywhere. A w/l ratio of 1 is a byproduct of perfectly working Elo over a sufficiently large number of matches.

[Belorion]

Roughneck45, on 11 December 2013 - 10:23 AM, said:

QFT

Would be nice if the higher ELO's didn't have to fail to find match repeatedly too. Doesn't happen all the time, but happens way more often than it should, sometimes for 30+ minutes.

Presumably the bottom elo people have that problem too. Just saying.

[Ghogiel]

Crimson Angel, on 11 December 2013 - 04:34 AM, said:

Win/Lose 2,608 / 1,731.

Fair point, but bear in mind quite alot of that is from before ELO was introduced. After, It has kinda went downhill for me.

Do you know that the dates of the last stat wipe was AFTER this competition> http://mwomercs.com/...ts?t=201302hero

Do you know that the dates of Elo introduction started with 2 weeks of seeding on this date> Feb 5th 2013

(edit: actually nvm the full W/L was never wiped, only mech stats :\, hey but at least you can tally up your actual W/L since Elo went in.)

Edited by Ghogiel, 11 December 2013 - 03:57 PM.

[Sandpit]

Roadkill, on 11 December 2013 - 12:03 PM, said:

I'm not sure what you mean by that. Elo is just a computation of your skill level, so it doesn't really work counter to anything. It just computes your skill level.

The matchmaker tries to pair up teams of approximately equal skill level, which it does by adding up the Elo scores of the players on each side.

If the teams are roughly equal - which is the goal - then the Elo delta will be the k-value and the predicted winner is mostly irrelevant. You can think of this being caused by the confidence being low - when the Elo values are similar, the prediction engine doesn't really know who's going to win.

On the other hand, if the Elo values are wildly different then the prediction engine can be reasonably confident about who's going to win. If it is right, then the Elo ratings don't change much because the expected happened. But if it is wrong the Elo ratings can change by a lot - up to double the k-value.

Does that help?

You're going well above the common understanding of statistics. (No offense to anyone, not intended to be a slight)

Basically:

Team 1 = ELO 500
Team 2 = ELO 500
The predictive component of the MM says both teams are equal, the winner is irelevant, and ELO won't change much if at all regardless of which team wins or how badly a team loses

Team 1 = ELO 1000
Team 2 = ELO 500
The predictive component of the MM says team 1 will win. If team 2 wins by any means even a 12-11 then ELO will shift dramatically for both teams. If team 1 wins then ELO won't shift much if at all.

I think that's what you were trying to illustrate? If not just disregard lol

[Monsoon]

I don't really care one way or the other, my question is: Does ELO really work? I still drop in matches with beginning players still forced to pilot a champion mech and in matches with people who are supposed to be 'uber'. Uber being people who have won tournament events and who are also members of some competitive guild.

Otherwise, I can say, yes there are days when it seems I'm getting stomped every match, but then there are those days when I'm the one doing the stomping.

[Sandpit]

Monsoon, on 11 December 2013 - 04:16 PM, said:

I don't really care one way or the other, my question is: Does ELO really work? I still drop in matches with beginning players still forced to pilot a champion mech and in matches with people who are supposed to be 'uber'. Uber being people who have won tournament events and who are also members of some competitive guild.

Otherwise, I can say, yes there are days when it seems I'm getting stomped every match, but then there are those days when I'm the one doing the stomping.

ELO works. It jsut doesn't work perfectly with MM

[Roadbeer]

This is serious business and it needs a poll.

[Balanor ElMarr]

Is ELO working correctly? Perhaps since all my win/loss over the last month are close to 50/50 (closer to 0.9 wins to 1 loss but who's counting). During this time, however, it seems that either we roll the other team or we get rolled ourselves. Do the matches seem balanced?

Absolutely not. If they were balanced more matches would go down to the wire.

[TB Freelancer]

I just did a quick scan of some end of round screen shots, maybe skimmed through 50'ish before getting bored. I couldn't see a single match where at least one lance on one side or the other didn't use a pair of clone builds or at least one lance wasn't at one extreme end of the weight scale be it an assault lance or a light lance. The most outlandish one I saw had 9 of the same mech on one side and 3 of the same light, so 2 chassis types, and 2 variants between 12 players on one side. All 3 lances were of the same house.....

TheCaptainJZ, on 11 December 2013 - 11:19 AM, said:

ELO works great for me. Not every match is balanced, but my overall experience has been. The people having trouble with ELO are those at the top or bottom. Since the scores are distributed as a bell curve, those in the top or bottom have almost no one in the same bracket to match with so they either fail to find a match, or get paired up with a bunch of low ELO players to average things out. That is the biggest failing of ELO right now. Once they have weight restrictions in, it should even things out a little better, but there will always be problems with any kind of match making solution.

I'm with you on that mostly. The main problem is that organized groups are able to game the system by running lance configurations that wildly throw balance out of whack. Lets be honest here, its those of us who do play in groups that drive what way the meta game moves. Pugs are forced to play along or suffer.

Simple weight restrictions will only solve a small part of the problem. Without class/variant restrictions in conjunction with weight restrictions, we'd wind up with even fewer cookie cutter lance configurations than we do now. It will simply narrow the parameters of how a group can game the system. In mixed matches, unless PGI does something to impose balanced lance configurations on groups, we'll never see balanced lances.

Anyway, the point I'm getting at, is ELO can't do its job properly while groups are completely free to game the system to their own ends.

Edited by TB Freelancer, 11 December 2013 - 06:03 PM.

[Roughneck45]

Belorion, on 11 December 2013 - 03:47 PM, said:

Presumably the bottom elo people have that problem too. Just saying.

No, they don't, as far as I have read and witnessed.

Their complaints are usually about the matchmaker taking 3+ minutes instead of the normal 30 seconds. Certainly have not heard of any half hour failed to find match sessions from new players or the lower bracket.

Edited by Roughneck45, 11 December 2013 - 06:34 PM.

[OneEyed Jack]

Roadkill, on 11 December 2013 - 10:33 AM, said:

Only theoretically true. Not true in practice (meaning with the MWO matchmaker).

Elo 1601 vs Elo 1599. Difference engine predicts that 1601 is going to win. But because the two scores are almost the same, the delta is going to be the k-value no matter who wins. How much you win by is irrelevant - Elo does not take margin of victory into consideration.

What you say is only true if the starting Elo values are wildly different, which shouldn't happen very often since that's exactly what the matchmaker is trying to prevent.

No, what I say is entirely true, even if the scores are very close. With very close scores the amount of movement is very small. With larger differences, the scores will move not at all (or close to it) if the match follows the prediction, and by a significant amount if there's an upset.

A lot of small movements can and do often result in very little overall movement, even when the W/L ratio isn't terribly close to 1:1. Particularly when considering that a single upset can reverse the results of dozens (or more) of smaller gains/losses in score.

(I would have thought it was clear from context, but I guess I need to spell out that by "score" I'm referring to Elo score, not match score)

[Threat Doc]

It seems to me that ELO is rather self-defeating, now that I've read more about it, when it works with the MWO matchmaking system. ELO is the system that defines the possibilities of wins and losses based off the overall ELO ratings of each side, but if the matchmaker is using ELO to make teams as even as possible, and the results of a match apply to both teams, regardless of who does what in the game, then that will keep those with higher ELOs down with lower teams, right? Or, does ELO consider the team as well as the individual?

If it does not, then ELO does NOTHING except restrict motion and the potential for greatness of MechWarrior's, right?

By the way, don't explain ELO to me, again, I understand what it does, how it works, I've one my research, and at this point I'm not certain whether the matchmaker needs to go, or ELO does. However, it seems to me, from what I understand, now, that the two are not honestly compatible.

As well, I hope Commander's are allowed to select teams that will drop for various missions once Planetary begins, because it would be rather silly to force either the matchmaker or ELO on either side.

[Sandpit]

Kay Wolf, on 12 December 2013 - 08:02 AM, said:

It seems to me that ELO is rather self-defeating, now that I've read more about it, when it works with the MWO matchmaking system. ELO is the system that defines the possibilities of wins and losses based off the overall ELO ratings of each side, but if the matchmaker is using ELO to make teams as even as possible, and the results of a match apply to both teams, regardless of who does what in the game, then that will keep those with higher ELOs down with lower teams, right? Or, does ELO consider the team as well as the individual?

If it does not, then ELO does NOTHING except restrict motion and the potential for greatness of MechWarrior's, right?

By the way, don't explain ELO to me, again, I understand what it does, how it works, I've one my research, and at this point I'm not certain whether the matchmaker needs to go, or ELO does. However, it seems to me, from what I understand, now, that the two are not honestly compatible.

As well, I hope Commander's are allowed to select teams that will drop for various missions once Planetary begins, because it would be rather silly to force either the matchmaker or ELO on either side.

See, ELO is an equation that matches players with one another based on skill which is in turn based on in-game performance that matches up against.......... couldn't resist lol

[Kekkone]

Kay Wolf, on 12 December 2013 - 08:02 AM, said:

then that will keep those with higher ELOs down with lower teams, right? Or, does ELO consider the team as well as the individual?

In a perfect world with an abundant pool of players, no. With a small player pool things get wonky.

Quote

However, it seems to me, from what I understand, now, that the two are not honestly compatible.

I don't understand this statement. The matchmaker works off of Elo, the two are tied together.

[Threat Doc]

Kekkone, on 12 December 2013 - 09:20 AM, said:

I don't understand this statement. The matchmaker works off of Elo, the two are tied together.

Yes, and no. The matchmaker USES ELO, but ELO does not work to improve player's positions in the game. So, if you match players all off together, based on their ELO range, and Sandpit has actually laid this out, before, there will be no ELO movement for the individual. So, everyone in your ELO bracket, in the range of players who have ELOs similar to you, and because the matchmaker tries to match the teams by exact, or inconsequentially separated ELO, is everyone you will continue to play with throughout your MWO career, and the likelihood of real improvement is little to none.

Does that help?

If not, I'll sort of lay out what Sandpit has expressed before, but in the terms I understand it in...

MWO Matchmaker
ELO Team 1 = 500
ELO Team 2 = 500 (+/- 10%, inconsequential)

In this instance, there will be little movement, if any at all, not only for the teams, but for the individuals, since individual score adjustments are tied directly to the team win, in accordance with PGIs own statements in the past concerning ELO.

Now, if one team ROFLstomps another, when they were perfectly matched, the way ELO works there is NO movement, because the teams were evenly matched.

The way it SHOULD work
ELO Team 1 = 1000
ELO Team 2 = 500

If Team 1 wins the match, because the way ELO works there will be no score adjustment to either team, because ELO predicted that Team 1 would win, so it's nothing special, no change.

If Team 2 wins, even by the slightest of margins, it's a major scoring change up for the winning team, and a minor scoring change down for the losing team.

The matchmaker DOES use ELO, but it is ineffective, because the ELO used is for the overall team, not the individuals playing, and there is never an ELO boost or drop because both teams playing have a 50:50 chance of winning against the other.

Does that help?

[Roadbeer]

Sandpit, on 12 December 2013 - 08:52 AM, said:

See, ELO is an equation that matches players with one another based on skill which is in turn based on in-game performance that matches up against.......... couldn't resist lol

No...

This is ELO
This is Elo

You were thinking of the later.

[Red Klown X]

ELO it s the best thing ever made !!!

DONT LET THIS TROLL KILL THIS FEATURE , IT HELP THE GAME TO BE AT THE TOP !