78 replies to this topic

[anubis969]

Kekkone, on 10 January 2014 - 01:38 PM, said:

Wrong.



http://en.wikipedia....o_rating_system

Not exactly. PGI doesn't use the text book version of Elo but rather their own slightly modified version. As such your actually both right and both wrong.

Under PGI's version if you win your Elo score goes up, if you lose it goes down. How much your Elo score changes is dependent on the disparity between the two teams average Elo scores and whether the outcome of the game matched the MM's prediction. If the disparity is large enough and the MM predicted right then your Elo score won't change.

Source:
http://mwomercs.com/...65#entry1626065

Edited by anubis969, 10 January 2014 - 04:55 PM.

[DaZur]

Nicholas Carlyle, on 10 January 2014 - 04:08 PM, said:

Whether you mean too or not you are agreeing that the game is fundamentally flawed. Whatever bud, enjoy your 12-0 games. I think they suck. Until we get rid of them this game will never flourish.

And it sure as hell won't be an e-sport like they want it to be.

LOL! You can read into what I stated all you want... I simply explained in detail why the MM is not the roll-stomp boogieman it's portrayed as...

If the mechanics of game-play are fundamentally "broken" because roll-stomps occur with greater frequency than reasoned matches, then bowling is broken because even when I throw a perfect strike ball... I occasionally get splits...

Chaos theory my friend... chaos.

[Craig Steele]

@Dazur, the challenge though is that not everyone shares your faith in Chaos theory.

I don't get your bowling example, either it is a strike ball or it's not. You only get splits from balls that are not strikes?

But on subject there's many players with commitments and have only a limited time to unwind / relax / blow stuff up and what they want is a good gaming experieince, they want entertainment for their time.

The F2P model loves this type of player, cause they are (stereotype) the ones who pay real cash to accelerate their place in the game. They are simply not going to have the time to grind out C-Bills for 3 shiny mechs, but $50- is no biggy, wheres the Credit Card.

If MM (or an over powering meta) is degrading their entertainment, thats not good for the game and should be something PGI look closely at.

[DaZur]

Craig Steele, on 10 January 2014 - 09:13 PM, said:

I don't get your bowling example, either it is a strike ball or it's not. You only get splits from balls that are not strikes?

If MM (or an over powering meta) is degrading their entertainment, thats not good for the game and should be something PGI look closely at.

I would absolutely love every match to be Nic's ideal "fun" match... The problem is, the MM (which he's erroneously blaming for the abundance of roll-stomps).. does it's best to create a reasonably balanced match. After that... it's largely dependant upon the ability of these two teams to play in a manor that would result in said "fun" match... Which is unpredictable and outside of any external influence to lead them to that result. That's when my above explained force attrition -aka- "tipping point" premise kicks in... with it's statistically predictable results.

The only external way to influence the outcome of match results is for the active participants to avoid the scenarios that lead to combat group attrition... Which obviously is reliant upon the individuals participating cooperating without any deviation. Again... unenforceable.

This is why teams and groups and players on VoIP have more reasoned results than the pure PUG groups... They are practiced, seeking mutually agreed goals and communicating them, which results in a higher probability of the reasoned outcomes.

The bowling analogy is simple... In controlled tests, a machine can deliver a "perfect" strike ball which in theory should guarantee a strike every time.... but it doesn't. The slightest variance of where the pins were set, imperfections in the pins themselves as well as the ball and drying of the lane... all provided variances that are outside of control that influenced what one would presume to be a predictable result.

Chaos Theory...

The MM is like the pinsetter... It creates matches between teams it believes is as balanced as reasonably possible. Once the pins are set... it's ability to make sure all things necessary to obtain a perfect strike are outside of it's control and influence.

[Craig Steele]

DaZur, on 10 January 2014 - 10:39 PM, said:

I would absolutely love every match to be Nic's ideal "fun" match... The problem is, the MM (which he's erroneously blaming for the abundance of roll-stomps).. does it's best to create a reasonably balanced match. After that... it's largely dependant upon the ability of these two teams to play in a manor that would result in said "fun" match... Which is unpredictable and outside of any external influence to lead them to that result. That's when my above explained force attrition -aka- "tipping point" premise kicks in... with it's statistically predictable results.

The only external way to influence the outcome of match results is for the active participants to avoid the scenarios that lead to combat group attrition... Which obviously is reliant upon the individuals participating cooperating without any deviation. Again... unenforceable.

This is why teams and groups and players on VoIP have more reasoned results than the pure PUG groups... They are practiced, seeking mutually agreed goals and communicating them, which results in a higher probability of the reasoned outcomes.

The bowling analogy is simple... In controlled tests, a machine can deliver a "perfect" strike ball which in theory should guarantee a strike every time.... but it doesn't. The slightest variance of where the pins were set, imperfections in the pins themselves as well as the ball and drying of the lane... all provided variances that are outside of control that influenced what one would presume to be a predictable result.

Chaos Theory...

The MM is like the pinsetter... It creates matches between teams it believes is as balanced as reasonably possible. Once the pins are set... it's ability to make sure all things necessary to obtain a perfect strike are outside of it's control and influence.

Agree with your statements, just not as to the competency of MM. I don't share your faith in it's basic premise.

Up until the last couple of days my experience from MM / Meta was poor for about 4 weeks in the majority of matches. Whether my experience is the same for everyone else, or whether its actually 40% and my mind is stretching it out in frustation is irrelevant, my feeling / experience is such that I am not going to spend anymore money here and I play a lot less than before.

F2P models build a massive player attrition into the model (same as shops build the cost of shop lifting into their prices) so yes I know I won't be missed, but forums are also for feedback and my feedback to the Devs is the MM / Meta is killing my game off.

[War Beast]

Texas Merc, on 04 January 2014 - 12:42 AM, said:

Its a numbers game. Once your team is down (or up) by 2-3 mechs things tend to snowball.

Which makes the rationale for 12 man vs 8 (less snowballing) a very bad move for having close games. It makes things quite a bit worse.

[DaZur]

Craig Steele, on 10 January 2014 - 10:55 PM, said:

Agree with your statements, just not as to the competency of MM. I don't share your faith in it's basic premise.

Up until the last couple of days my experience from MM / Meta was poor for about 4 weeks in the majority of matches. Whether my experience is the same for everyone else, or whether its actually 40% and my mind is stretching it out in frustation is irrelevant, my feeling / experience is such that I am not going to spend anymore money here and I play a lot less than before.

F2P models build a massive player attrition into the model (same as shops build the cost of shop lifting into their prices) so yes I know I won't be missed, but forums are also for feedback and my feedback to the Devs is the MM / Meta is killing my game off.

Understood... I too had my moments of "this is unreasonable and unbearable". I just kept plugging away... well, because I'm stubborn I guess.

I think the hardest hurdle folks have is the presumption that the MM is matching them individually... when in reality it's matching a team. As such, it's not guaranteeing each player a 50/50 chance of winning... it's guaranteeing the "team".

If the 'team" fails... so goes the perceived failure of the MM.

now, that said... I'd be a raving fool if I said the MM was "perfect"... which it's clearly is not. But again... this is the end result of chaos. The MM cannot control the number of players available for the pool, cannot control their Elo and cannot control the class of mech nor the meta they all subscribe to. In short... the MM does the best it can, which what it has to work with...

Edited by DaZur, 10 January 2014 - 11:11 PM.

[War Beast]

Craig Steele, on 10 January 2014 - 10:55 PM, said:

F2P models build a massive player attrition into the model (same as shops build the cost of shop lifting into their prices) so yes I know I won't be missed, but forums are also for feedback and my feedback to the Devs is the MM / Meta is killing my game off.

Actually the sad thing is you will be missed. Cause as you and all of the new players leave, it leaves only a handful of core players to keep playing against each other in a rather tiny player population (compared to what MM really needs to function), making matchmaking worse and worse. Leaving the game with a small income base. Meaning that understaff situation development suffers from now only gets worse. Production gets slower and slower (as if it isnt f'ing ridiculously slow as it is). And the game simply dies.

Thats the future as it stands.

[Kekkone]

anubis969, on 10 January 2014 - 04:41 PM, said:

Not exactly. PGI doesn't use the text book version of Elo but rather their own slightly modified version. As such your actually both right and both wrong.

Under PGI's version if you win your Elo score goes up, if you lose it goes down. How much your Elo score changes is dependent on the disparity between the two teams average Elo scores and whether the outcome of the game matched the MM's prediction. If the disparity is large enough and the MM predicted right then your Elo score won't change.

Source:
http://mwomercs.com/...65#entry1626065

Thank you for the link, i stand corrected!

Some contradicting info on those posts:

Quote

A player’s rating will only go down if they are beaten by a player who has a lower rating than theirs.

vs

Quote

If the Match Maker determines that you're going to win and you actually win, then your Elo score isn't going to change very much (if at all). The same applies to a prediction of loss and you actually lose, your score may drop but it will be slight.

[Joseph Mallan]

Khobai, on 10 January 2014 - 10:36 AM, said:

This. Lack of respawns causes an avalanche effect. Generally once a team gets down by 4 mechs theres almost no way they can come back.

As should happen when you are fighting. Or even playing at fighting. Not everyone deserves a trophy just because they showed up. If you don't play well you die. Sit on the sidelines and try to figure out your mistakes. It's what I do.

[Craig Steele]

DaZur, on 10 January 2014 - 11:09 PM, said:

In short... the MM does the best it can, which what it has to work with...

Which probably concerns me the most, that if the system is optimal (and lets face it, PGI aren't going to deliberatly roll out a bucket of proverbial) then the MM experieince is driven by poor populations, which deflates me somewhat about what I hoped to see from this game.

It's not dead yet, but gee they have their work cut out for them. Such is life, but it's not over yet.

[Joseph Mallan]

Craig Steele, on 11 January 2014 - 04:49 AM, said:

Which probably concerns me the most, that if the system is optimal (and lets face it, PGI aren't going to deliberatly roll out a bucket of proverbial) then the MM experieince is driven by poor populations, which deflates me somewhat about what I hoped to see from this game.

It's not dead yet, but gee they have their work cut out for them. Such is life, but it's not over yet.

Or there are not as many players in/near your Elo an the game needs to stretch out to fill your games.

Dammit Jim It's a Number cruncher not a Doctor!

[Craig Steele]

Joseph Mallan, on 11 January 2014 - 04:53 AM, said:

Or there are not as many players in/near your Elo an the game needs to stretch out to fill your games.

Dammit Jim It's a Number cruncher not a Doctor!

Which is still the same argument, If I am a high ELO player and their are not enough players, not great that all the high skilled players have left. If I am mid range, not great either, and is I am low, not great.

As a company PGI are (hopefully) trying to keep populations high across the board (because skill is not a driver of revenue as we know).

If MM is doing it job whatever my ELO, and there is significantly less population, I stand by my statement.

Or MM is bust, one or the other.....

Forgive speeling, too many sherberts

Edited by Craig Steele, 11 January 2014 - 04:59 AM.

[anubis969]

Kekkone, on 11 January 2014 - 04:21 AM, said:

Thank you for the link, i stand corrected!

Some contradicting info on those posts:



vs

You're welcome .

Yer they didn't do a very good job of explaining who Elo works first time which is why they had to go back and clarify things. It's the second statement that is correct.

I probably should also point out that with the post being old there are a few minor details that are now out of date. For example, and correct me if I'm wrong but I believe new players now start with an Elo score of 1100 rather than 1300.

[Joseph Mallan]

Craig Steele, on 11 January 2014 - 04:58 AM, said:

Which is still the same argument, If I am a high ELO player and their are not enough players, not great that all the high skilled players have left. If I am mid range, not great either, and is I am low, not great.

As a company PGI are (hopefully) trying to keep populations high across the board (because skill is not a driver of revenue as we know).

If MM is doing it job whatever my ELO, and there is significantly less population, I stand by my statement.

Or MM is bust, one or the other.....

Forgive speeling, too many sherberts

You are not reading that right. There could be thousands of high skill players out in drops all around you, but that doesn't mean they are available when you hit Launch.

I get the feeling that to many people are thinking they were born in the Al Franken Era.

[Craig Steele]

Joseph Mallan, on 11 January 2014 - 05:09 AM, said:

You are not reading that right. There could be thousands of high skill players out in drops all around you, but that doesn't mean they are available when you hit Launch.

I get the feeling that to many people are thinking they were born in the Al Franken Era.

Maybe I shoudl have coffee?? lolol

I still don't get your point. Albeit I think we are talking about the same thing at tangents.

If I press launch the game searches for 60 seconds (or whatwever it is) for players in my ELO. Finds 23 others and throws us into a game.

If it doesn't find 23 of my ELO it throws me into a game regardless.

If there are "thousands" of players in my ELO pressing the button around my time, I am going to be dropping with them (60 seconds either side). The chances of being the 25th person every second for 60 seconds when "thousands" are pressing the button are astronomical no? My assessment is not based of pressing the button once, but repeated (imo) failures of MM over a 4 week period playing 3 / 4 hours a day minimum. It just defies the odds?

But if there is not population in my ELO, then I get dumped into another ELO with either a stomp or a wipe accordingly.

Which was the point of my orginal post, that stomps and wipes are happening all to frequently and either MM is bust or populations have degraded to a point that MM cannot function as intended.

[Joseph Mallan]

How many of those thousands are already in a drop when you hit Launch though? ELo will use a a Bell curve to fill your game and Elo can be a couple hundred points in either direction.

[Craig Steele]

Joseph Mallan, on 11 January 2014 - 06:06 AM, said:

How many of those thousands are already in a drop when you hit Launch though? ELo will use a a Bell curve to fill your game and Elo can be a couple hundred points in either direction.

But 5 out of 6 matches? sometimes I DC, it's possible sure, but pretty heavy odds.

Either way, I am still here waiting for MM fix to roll out, hopefully my experience is the minority and life goes on right

[MustrumRidcully]

No matter how equal two sides in a team are - it is the nature of this type of attritional warfare to end one-sided. Bad matchmaking can make this worse if teams aren't equally strong to begin with, but even if they are, the first side to lose a unit is likely to lose the next one, too.

Only if both sides would be perfectly in focusing fire and every one always can shoot a the same targets (so no cover).

Side A is fighting Side B. ONe mech on each side is close to destruction and thus tries to retreat. But Side B's mech doesn't make it in time and is dead. Now both sides must target a new, fresh target, but during this time, the mech that retreated in time can return to join Side A, giving his team a firepower advantage. The enemy now would have to react very quickly to realize the weakened target is back - if they don't, the firepower advantage will live to full and the second mech of Side B is gone, and the second damaged mech on Side A has even more time to retreat.
And that is just an assumption were both sides are engaged with each other the entire time, basically. Already pretty optimistic.