184 replies to this topic

[Xavori]

A Headless Chicken, on 04 December 2017 - 06:11 PM, said:

Kay, I have been asking many questions over and over and found I have been repeating myself and becoming more snarky as it goes.

Honestly, all it really takes to turn the tide of battle is that one player. Seen it happen more often than not. Sure, having a bunch of potatoes in your team may hinder you but it is entirely possible.

More importantly, what makes you so sure that the people who are "potatoes, "good" and "legendary" will not end up in the same game after implementation of tr00sk1ll?

Note that we have a small playerbase - I easily see the same names in my games throughout the night. Right now your top 20% of the playerbase has a mix of the above tiers simply because they are active.

Even if PGI develops some amazing magical formula to slot people where their deserved position on the bell curve is (through what metrics is questionable and wow we r goin' 4 real esports bois), nothing much is going to change when only the top 20% of the bell curve plays regularly, which is not many people at all. They'll all either get forced into a game and you whine "potatoes!" which is this thread all over, or nobody gets a game and we all whine and start another rage post or quit.

I have been asking this with no solid answer, and all I keep hearing is TrUeSkIlL wIlL sOlVe EvErYtHInG.

For starters, there is no bell curve for player skill. There is no curve at all. We have no idea what the relative skills of pilots are compared to each other, nor do we have any idea what kind of distribution there is. For all we know, out of the ~30k active players, there are only 100 superstar pilots and 1000 guys who still haven't figured out how to press W to make their mech go forward with another 9000 who are as dangerous to their own team as the enemy. But the key point is we have no idea.

As for your question: "More importantly, what makes you so sure that the people who are "potatoes, "good" and "legendary" will not end up in the same game after implementation of tr00sk1ll?"

They WILL end up in the same game. The goal is not to keep potatoes and legendary players apart. The goal is to make sure the two teams have approximately the same total player rating. The matchmaker isn't trying to keep tiers separate. It's trying to make quality matches where the two teams have a roughly equal chance to win.

So the potatoes and legendaries not only will end up in the same game, they'll likely be on the same team in order to balance each other out. The highest rated player in a match will almost always end up with the lowest rated player on their team. The matchmaking algorithms out there literally start with this in many cases. It's the obvious place to start. Best-worst on team A. 2nd best and 2nd worst on Team B, then compare the two teams and the lower total skill gets the third best and so on. This isn't anything new, and it's been used for a very long time.

It's why when you go back and look at PGI's attempt at ELO, you just have to keep scratching your head how they broke it. It isn't like they had to start from scratch, and yet, apparently they had all kinds of problems.

So my suggestion is to not make up their own, but to take an existing, and known to work, player rating system and give that to us.

[The Lighthouse]

A Headless Chicken, on 04 December 2017 - 06:21 PM, said:

Kay, I have been asking many questions over and over and found I have been repeating myself and becoming more snarky as it goes.

More importantly, what makes you so sure that the people who are "potatoes, "good" and "legendary" will not end up in the same game after implementation of tr00sk1ll?

Note that we have a small playerbase - I easily see the same names in my games throughout the night. Right now your top 20% of the playerbase has a mix of the above tiers simply because they are active.

Even if PGI develops some amazing magical formula to slot people where their deserved position on the bell curve is (through what metrics is questionable and wow we r goin' 4 real esports bois), nothing much is going to change when only the top 20% of the bell curve plays regularly, which is not many people at all. They'll all either get forced into a game and you whine "potatoes!" which is this thread all over, or nobody gets a game and we all whine and start another rage post or quit.

I have been asking this with no solid answer, and all I keep hearing is TrUeSkIlL wIlL sOlVe EvErYtHInG.

The goal of better matchmaking is not separating good players and bad players at this point. With current population it is just impossible.

What good matchmaker should do is mix good players and bad players in balanced way for both teams.

Right now, current matchmaker does like this.


Potato Good
Good Good
Potato vs Good
Potato Legendary
Potato Legendary
Legendary Legendary

Of course no one is going to have good time with this.

Just by matchmaker to do....

Potato Potato
Potato Potato
Good vs Good
Good Good
Legendary Legendary
Legendary Legendary

Just balance skill alone would be tremendously helpful for quality game. Currently MWO's matchmaker even does not do this simple balance at all. We literally do not have any matchmaking.

[A Headless Chicken]

Point taken. Opening the tier gates to the "lower" tiers would actually destabilize the games I think, what with the lesser potatoes and legendary people, meta and derp loadouts, mismatched tonnages, stomps are going to happen more often.

Actually now that I think about it, I don't think that stomps mean "low quality games". It just means one team is more aggressive than the other. You can't measure that. It also could mean one guy was caught with pants down and picked off effectively, and the deathball kept rolling. For all you know its a 12-0 and the survivors all have red CTs. Or maybe the losing team had only Nighstars and Thanatos'

Now, I am very interested to know, what constitutes a 'quality game' to you?

Edited by A Headless Chicken, 04 December 2017 - 07:00 PM.

[Khobai]

if you just take 24 random people and stick them randomly on two teams you should statistically get two balanced teams. thats not where MWO fails at matchmaking.

where MWO fails at matchmaking is that it will stick an all flamer boars head on one team and a super meta dual gauss/laservomit deathstrike on the other team and call it even.

the average skill of the players might be close enough, but if matchmaker doesnt also balance the quality of mechs for both teams, then its all for naught.

because player skill should follow a bell curve. but mech quality is all over the place and doesnt follow a bell curve like player skill does. so you cant just stick random mech builds on both teams and expect it to balance out.

Edited by Khobai, 04 December 2017 - 07:00 PM.

[Xavori]

A Headless Chicken, on 04 December 2017 - 06:53 PM, said:

Point taken.

Now, I am very interested to know, what constitutes a 'quality game' to you?

50/50 chance for either team to win.

Khobai, on 04 December 2017 - 06:56 PM, said:

if you just take 24 people and stick them randomly on two teams you should statistically get a balanced team. thats not where MWO fails at matchmaking.

where MWO fails at matchmaking is that it will stick an all flamer boars head on one team and a super meta dual gauss/laservomit deathstrike on the other team and call it even.

That is not how statistics work AT ALL. You cannot just take random numbers and get meaning from them. Random numbers are random.

And there is absolutely nothing that can be done about bad builds finding their way into matches. But if you have balanced rated teams, odds are you won't get stuck on an entire team of potatoes every match all night (or in my case, it's usually the mornings when I get stuck on Oceanic servers that things go terribad in long stretches)

[Khobai]

Quote

That is not how statistics work AT ALL. You cannot just take random numbers and get meaning from them. Random numbers are random.

its exactly how it works

because player skill follows a predictable bell curve with most people in the middle of the bellcurve for skill

so over thousands of games, if you take 24 random people and split them up, both teams should come out roughly even in terms of skill level.

yeah not all games will be perfectly balanced. but over thousands of games it will even out. thats how a statistical average works.

most games dont have matchmakers at all. they just pick random people out of a high population and let the teams statistically balance themselves. And most of the time IT WORKS.

but most games dont let you take highly customized mechs either, and thats where the problems happen for MWO. because the potential disparity between a bad custom mech and a good custom mech is huge.

If MWO has a matchmaker for anything, it should be making sure both teams have the same quality of mechs.

Edited by Khobai, 04 December 2017 - 07:08 PM.

[The Lighthouse]

A Headless Chicken, on 04 December 2017 - 06:53 PM, said:

Point taken.

Now, I am very interested to know, what constitutes a 'quality game' to you?

"Quality game" by means that game fairly composes teams that each individual can do their work at near-max capacity without hindered by internal team issues. If that is not possible, both teams should have similar amount of internal team issues.


From good player's perspective, it's bad players playing bad and forced him/her to carry his/her team. Idealy, there should be no bad players, but that's just impossible. Next thing is that opposite team's good player should have same difficulty.


From bad player's perspective, it's good players stomping him/her. Idealy, he/she should not go against good players, but again it's not possible. Next thing we can do is that he/she should have same amount of good players against opposite team's good players, so hopefully it will cancel out.


Ironically, in perfect quality game no one gets "fun" because they are all fairly challenged. So this is where the matchmaking problem ends and gameplay design problem begins here, in order to solve this "fun" issue.


I highly recommend you to watch videos from Extra Credit on youtube. They have a lot of excellent gameplay design principle videos.

Edited by The Lighthouse, 04 December 2017 - 07:06 PM.

[sycocys]

Xavori, on 04 December 2017 - 07:01 PM, said:

And there is absolutely nothing that can be done about bad builds finding their way into matches.

There sure is - calculate a BV for the mechs + their loadouts + their skill tree - and use that as your primary weighting system.

*edit* While it won't stop terrible builds from getting into matches, you'll have an equally terrible build (or builds) on the other side of the match.

Edited by sycocys, 04 December 2017 - 07:09 PM.

[Xavori]

Khobai, on 04 December 2017 - 07:01 PM, said:

its exactly how it works

because player skill follows a predictable bell curve with most people in the middle for skill

so over thousands of games, if you take 24 random people and split them up, both teams should come out roughly even in terms of skill level.

yeah not all games will be perfectly balanced. but over thousands of games it will even out. thats how a statistical average works.

No. Player skill does not follow a bell curve. Player skill is games is usually very bottom heavy because the number of players who put in the time and effort to master a game is a tiny fraction of the number of players who just pick up a game casually for fun.

But worse than that, you have no way of knowing what the possible range of player skills is. That means you cannot possibly make any kinds of analysis about what will happen. You have to eliminate as much uncertainty as you can. You can't fix builds, but you can at least say that usually Player A plays better than 75% of the other players in MWO, so we'll give him a rating that reflects that and then build teams on that assumption.

It won't be perfect, but it will be so much better than what we have now which actually is your random teams because PSR is garbage and the matchmaker gives up on it if it can't make a match quickly enough.

[Khobai]

Quote

No. Player skill does not follow a bell curve. Player skill is games is usually very bottom heavy because the number of players who put in the time and effort to master a game is a tiny fraction of the number of players who just pick up a game casually for fun.

yes. what you just described is called a bell curve

in a bell curve, the number of highly skilled players drops off because theres less of them

Quote

There sure is - calculate a BV for the mechs + their loadouts + their skill tree - and use that as your primary weighting system.

this.

essentially you assign a value to each mech based on how good it is.

an all flamer boar's head wouldnt be worth the same as a laser vomit boar's head for example.

and all boar's heads would be worth less than laser vomit battlemasters, which is inarguably one of the best IS assaults.

it doesnt even have to be a complicated system. even something rudimentary that can just tell the difference between a good mech and a bad mech would literally be better than nothing. because what we have now is NOTHING.

Edited by Khobai, 04 December 2017 - 07:15 PM.

[UnofficialOperator]

A Headless Chicken, on 04 December 2017 - 06:21 PM, said:

More importantly, what makes you so sure that the people who are "potatoes, "good" and "legendary" will not end up in the same game after implementation of tr00sk1ll?

How are you being objective when you are the first to start with this tr00sk1ll nonsense? How is this not being insulting and derogatory? You can give the insults but can't take it eh? Just giving a little taste of your own medicine. LOL OWNED

Quote

I don't check Oceanic because no one plays there. Do you? Are you salty because of it?

LOOOOOOOL I play all servers so I don't really care either way.

Quote

I play noon (NA primetime) and nights (OC primetime) on weekends. And yes I see the same names constantly during either session. Problem?

Lol and now you change your previous answer. You said

Quote

I easily see the same names in my games throughout the night

And you play Oceanic nights.... therefore of course you will see the same names due to small player base! Wow are you that dense and dishonest? I answered your original assertion, don't try to dodge now by changing your statement.

Quote

And what point is there to implement a system that will not change anything? If anything, it maintains the status quo and makes wasted effort and little details on the devs. You can reduce the game sizes, but the distribution of players within a certain range (top 80-100%) will still be matched together, unless you want to tier it so stringently that you can only have games with people +/-5% of your tr00sk1ll. Then nobody gets games.

And here you are contradicting yourself. Ok so if you are right, it maintains status quo. So why should you care?

And since you are playing on NA server, the server is big enough for further segregation. Why? Are you worried that your WLR would drop because you can't feed on lower skilled players anymore?

Why are you advocating against improvements to the system?

Quote

EDIT: If you want to make this personal, by all means, go ahead. I have been trying to be objective, and in true Singaporean style, you come in, rebut my arguments and top it off with a few personal insults. Did you honestly think every human being that is not you is stupid? Thank you for for your opinions. I am ashamed to be YOUR countryman.

Sorry I don't pander to strawberry generation and I believe as I've shown you above, you were the first to start the belittling of others first with your troosk1ll BS

Your insistence on sticking to status quo and not trying to improve everyone's situation betrays your hidden motivations. Why are you advocating against playing only vs the best of the best since you are T1 with a 2 WLR?

[A Headless Chicken]

Xavori, on 04 December 2017 - 07:01 PM, said:

50/50 chance for either team to win.



That is not how statistics work AT ALL. You cannot just take random numbers and get meaning from them. Random numbers are random.

And there is absolutely nothing that can be done about bad builds finding their way into matches. But if you have balanced rated teams, odds are you won't get stuck on an entire team of potatoes every match all night (or in my case, it's usually the mornings when I get stuck on Oceanic servers that things go terribad in long stretches)

Uncheck Oceanic. you're playing with Pseudo-Tier 1s and blighted potatoes because there are only about 0.3 games going on there at a time. Hyperbole intended.

The statistics part is true though, player skill among the active base does follow a bell curve, because anyone can pick it up and play till they are used to the game (A day 1 player can be at the 50%th percentile). Learning the mechanics and playing the meta gets you higher. Having #gamesense and skills makes you top. If you can find a distribution in life that does not adhere to it then I applaud you.

More than that, most players are more or less at the same level in the games I play (not that I'll say anything about that level) - discounting LRM boats and people playing with their feet. Outliers rarely appear, but when they do they are usually alone.

Now, if you have a match where the skill variance is not just 'wide' but 'massive', I think that game quality would end up dropping because the good players would always end up feasting on the bads. Potshotting things standing still in the open is fun. these will always be the first to go. Poor bads are both challenged and get no fun, at worst getting instagibbed. Once a few of them go, the push begins and a stomp occurs. Yes, win rate may be 50% but at the expense of bads dropping in for 1 min and dying, and goods inflating their stats. Is that the 50-50 win game you want?

[Khobai]

I would contend that mech variance is a bigger issue than skill variance. All it takes is one bad assault mech build for one team to have an overwhelming advantage over the other team.

At least with players, the bell curve distribution usually guarantees that both teams get roughly the same number of bad players. but nothing guarantees both teams will get the same number of bad assault mechs, since that doesnt follow a bell curve distribution.

Edited by Khobai, 04 December 2017 - 07:36 PM.

[sycocys]

Khobai, on 04 December 2017 - 07:22 PM, said:

I would contend that mech variance is a bigger issue than skill variance. All it takes is one bad assault mech build for one team to have an overwhelming advantage over the other team.

One bad mech in any of the weights instantly makes it 11 vs 12 any way you look at it. Basically it is as bad as dropping an having one person d/c straight off.

[A Headless Chicken]

UnofficialOperator, on 04 December 2017 - 07:14 PM, said:

How are you being objective when you are the first to start with this tr00sk1ll nonsense? How is this not being insulting and derogatory? You can give the insults but can't take it eh? Just giving a little taste of your own medicine. LOL OWNED



LOOOOOOOL I play all servers so I don't really care either way.



Lol and now you change your previous answer. You said



And you play Oceanic nights.... therefore of course you will see the same names due to small player base! Wow are you that dense and dishonest? I answered your original assertion, don't try to dodge now by changing your statement.



And here you are contradicting yourself. Ok so if you are right, it maintains status quo. So why should you care?

And since you are playing on NA server, the server is big enough for further segregation. Why? Are you worried that your WLR would drop because you can't feed on lower skilled players anymore?

Why are you advocating against improvements to the system?


Sorry I don't pander to strawberry generation and I believe as I've shown you above, you were the first to start the belittling of others first with your troosk1ll BS

Your insistence on sticking to status quo and not trying to improve everyone's situation betrays your hidden motivations. Why are you advocating against playing only vs the best of the best since you are T1 with a 2 WLR?

Ah, I see UnofficialOperator replied. I may be a strawberry, sure. I'd actually have assumed the one typing LOOOOOLs and LOL OWNED is the little boy who hasn't finished NS. Quite usually, actually. If you're older then write like it, not some EDMW forum Ah Bang. Unless you are one.

Tr00sk1ll is what it is: a word with no metric and no yardstick behind it. No game studio has ever been able to compute it. therefore it's fitting that I use a derogatory term for it. Too bad you took offense at my deconstruction of the word because maybe you needed something to prove your manhood. Do you feel manlier now? Is your chest hair growing at an alarming rate?

Next, I edit my words to qualify things and spell out information for my lesser counterparts. Of course you would assume I'm trying to do a 180. Figured you could use some help understanding it if I spoon-fed it to you. Or would you prefer Ensure through a tube? Maybe you need an Encik to give you orders?

Also, you can put me in any game with both goods and bads. It doesn't matter. I'm doing well in status quo with and without potatoes on my team, and I'd be happier if I get a challenge.

Lastly if the Isengrim board can count my 0.00 WLR seasons where I didn't play and still rate me at 2 average, I think it explains enough. I like to think I'm a decent player.

Sorry that I don't respect my elders, but if anything you have proven yourself to be unworthy of any.

OP, sorry for derailing. Some things need to be given hard and good. I'm going silent before the Modhammer drops on the thread.

[Deathlike]

Regardless of how you want to do the matchmaking, people not wanting to work together is why those teams lose. Even if you stack up to a certain # of great players together, there is a tipping point where they can still lose as rare as that can be.

Still, the majority of you that clamor for a change in MM probably are in the group that never really want to work together, die, and lose often.

You don't need a perfect MM to see this happen more often than not.

To think the days were the actual really good players were complaining about wait times? Let's just have everyone suffer instead. That's what you have now.

[The Lighthouse]

A Headless Chicken, on 04 December 2017 - 07:15 PM, said:

Now, if you have a match where the skill variance is not just 'wide' but 'massive', I think that game quality would end up dropping because the good players would always end up feasting on the bads. Potshotting things standing still in the open is fun. these will always be the first to go. Poor bads are both challenged and get no fun, at worst getting instagibbed. Once a few of them go, the push begins and a stomp occurs. Yes, win rate may be 50% but at the expense of bads dropping in for 1 min and dying, and goods inflating their stats. Is that the 50-50 win game you want?

Nothing you can do about it, this is the inherent problem of all pvp games with low population.

As I said before, this is where matchmaking problem ends and gameplay design problem begins. Supposedly there should be a weapon or mechanism that...


1) Very powerful that gives bad players chance to win against good players.
2) But with definite limit on its power that it is not appealing to good players.

The best example is the noobtube from Call of Duty series. MWO does have such weapon too, it's called LRMs.

Despite hate from pretty much all competent players, LRMs is necessary evil and should be kept unappealing for good players but usable for bad players, otherwise this game becomes predictable game where good players always win against bad players....

Sounds un-intuitive, but it really harms the game. But think about it; after all people play games to have fun. When the game stops being fun, the game dies.

[Khobai]

Quote

Despite hate from pretty much all competent players, LRMs is necessary evil and should be kept unappealing for good players but usable for bad players, otherwise this game becomes predictable game where good players always win against bad players....

LRMs are stupid mostly because of the lack of destructible terrain

LRMs should tear right through rocks and buildings

But for some dumb reason they provide better protection than mech armor

PGI needs to do a pass on every map and make most of the things that stick out of the ground destructible like the walls in incursion. hard terrain like hills, the walls in HPG, or the ground itself shouldnt be destructible. but all trees, buildings, and many of the rocks that jut out of the ground should be destructible.

LRMs would be a more competitive weapon if people couldnt hide behind stupid crap that shouldnt be invincible. But I think indirect LRMs should probably require NARC or TAG or a UAV in that case too. That would increase the skill level to use indirect LRMs.


As for 8v8... Im of the opinion that 8v8 was way better

But id be okay with some maps being 8v8 and some being 12v12

Thats better than everything being 12v12 at least

Edited by Khobai, 04 December 2017 - 08:02 PM.

[WarHippy]

Wintersdark, on 29 November 2017 - 05:29 PM, said:

You had me till the last

12v12 has never been a good change. In theory it's awesome, in practice it's just not. It's not about "giving up"; it's that in all these years quickplay 12v12 has never worked out well, for all the reasons that detractors warned when it was first introduced.

There comes a point where you need to accept that something you wanted to do is simply not going to work, and to continue tilting at windmills isn't going to be productive. It's great for quasi-respawn Faction Play, but much less so for quickplay.

I just don't see where you are coming from and have to agree with Mystere. 12v12 was fine and still is. In fact it is probably better now than it was because all of the smaller maps where it was a bit overcrowded are gone(sadly). The games during 8v8 and 12v12 played out more or less the same and stomps were still prevalent. What was different was the mechs and build options available during 8v8 that gave everyone a skewed opinion of what 8v8 was like. The change they should have made was for certain maps to be 8v8 and others 12v12 instead of all of either option.

[UnofficialOperator]

A Headless Chicken, on 04 December 2017 - 07:33 PM, said:

Ah, I see UnofficialOperator replied. I may be a strawberry, sure. I'd actually have assumed the one typing LOOOOOLs and LOL OWNED is the little boy who hasn't finished NS. Quite usually, actually. If you're older then write like it, not some EDMW forum Ah Bang. Unless you are one.

Tr00sk1ll is what it is: a word with no metric and no yardstick behind it. No game studio has ever been able to compute it. therefore it's fitting that I use a derogatory term for it. Too bad you took offense at my deconstruction of the word because maybe you needed something to prove your manhood. Do you feel manlier now? Is your chest hair growing at an alarming rate?

Next, I edit my words to qualify things and spell out information for my lesser counterparts. Of course you would assume I'm trying to do a 180. Figured you could use some help understanding it if I spoon-fed it to you. Or would you prefer Ensure through a tube? Maybe you need an Encik to give you orders?

Also, you can put me in any game with both goods and bads. It doesn't matter. I'm doing well in status quo with and without potatoes on my team, and I'd be happier if I get a challenge.

Lastly if the Isengrim board can count my 0.00 WLR seasons where I didn't play and still rate me at 2 average, I think it explains enough. I like to think I'm a decent player.

Sorry that I don't respect my elders, but if anything you have proven yourself to be unworthy of any.

OP, sorry for derailing. Some things need to be given hard and good. I'm going silent before the Modhammer drops on the thread.

Lol I think you are confusing being able to play a game well vs being able to think critically or argue well. You answered nothing, dodged a lot and only looked worse