147 replies to this topic

[Dimento Graven]

Mr Snrub, on 31 October 2017 - 08:53 AM, said:

This happened several times at the weekend:

Oh those 228 guys! They doing their "shenannigizing" again! Them bunch o' silly gooses!!!

[Mystere]

Khobai, on 31 October 2017 - 07:40 AM, said:

can we at least all agree that matchmaker should use matchscore instead of wins/losses for elevating players?

top8 match score advances in rank
middle8 stay the same
bottom8 go down in rank

that way players dont get into tiers they dont deserve to be in and the matchmaker might actually start working...

that would at least fix the issue of not having to carry as many terrible players in 12v12 because players would actually be in the tiers they belong in.

No we cannot agree on that as well. A win currently seems to contribute less to match score when compared to ineffective damage.

Khobai, on 31 October 2017 - 07:40 AM, said:

so change match score so it cant be gamed? pretty simple.

if you make it so every action awards an appropriate amount of match score, there is no way to game it.

it would just require adjusting the match score awarded for things like ams shooting down missiles, which is currently set way too high. you shouldnt get like 600 match score because you shot down 2000 missiles.

its an easy fix. its not really a problem to fix that at all.

It is not simple because there a a lot of in-game actions that contribute to a win but are very difficult to measure much less detect. Only wins -- and a whole lot over a long period -- hints at those intangibles.

Edited by Mystere, 31 October 2017 - 09:11 AM.

[Dimento Graven]

Mystere, on 31 October 2017 - 09:01 AM, said:

Most of the players I used to see doing just that I no longer see.

Heck, I myself no longer give a flying **** about doing the same. I just go "Is my HOTAS doing what I want or should I adjust its programming again" when dropping in MWO.

Yep. Too frustrating. It is actually worse than herding kittens in the solo queue. At best just call out targets with open components, that are easy kills. That is pretty much the only thing that gets anyone's attention.

[Khobai]

Quote

No we cannot agree on that as well. A win currently seems to contribute less to match score when compared to ineffective damage.

a win should not contribute to match score anyway

because you cant control if your team of potatos wins or not

all you can control is how well you do personally

and how well you do personally is ALL that should matter for increasing your rank

Edited by Khobai, 31 October 2017 - 09:22 AM.

[MischiefSC]

Mr Snrub, on 31 October 2017 - 08:53 AM, said:

This happened several times at the weekend:

Several times last weekend?

I've played a bit over 11,000 matches and I can count on one hand the number of times I've seen something like that. Every time I've screenshot it and sent in a support ticket. One of those times I can say with certainty (because I knew some of the people involved) it resulted in short term bans and a serious warning.

However people who think everyone else is cheating all the time and the people who beat them in matches are all using hax or doing it because they wake up every morning trying to think of ways to hold so-and-so back from the successes they richly deserve....

there's no fixing people like that. If someone objects to 8 v 8 because of an actual technical complication (they just like having more targets on the field, they like short TTK, they prefer matchmaking to have a bigger stretch between players, etc) that's one thing. However objections based on 'because all the good players will syncdrop like they always do and super-farm all the bad players' can be treated with the same level of concern as 'if you introduce 8 v 8 back in then ALIENS'.

8 v 8 will tighten matchmaking by 33%. It will also create more simultaneous matches.

And FFS if you think someone is sync dropping then report it. If you think reporting does nothing.... I guess you should loosen the fit on your tinfoil hat maybe? I dunno.

Khobai, on 31 October 2017 - 09:21 AM, said:

a win should not contribute to match score anyway

because you cant control if your team of potatos wins or not

all you can control is how well you do personally

and how well you do personally is ALL that should matter for increasing your rank

It's a 12 v 12 team game. How well you play with the team you have is the single most critical direct reflection of how you play.

Nothing in your score, nothing else you do in the game, is as relevant to your win/loss as... did it actively create a win.

Saying that your win/loss average is beyond your control is not just false but demonstratively so. Are you saying all the great players in the game with high W/L got that by, what, random chance? Consistently for the last 16 months? All the T5 people with sub-1.0 W/L, that's just random chance? Coincidence?

That's not just false it's easy to demonstrate that it's false.

[Xiphias]

Personally, I think it would be neat if the game had 8v8v8, 6v6v6v6, 4v4v4v4, free for all, etc. modes players could drop in. It would be interesting to see what that would bring about from a gameplay perspective, even if it was limited to private lobbies for league play. That said, I seem to remember hearing that it was a programming limitation from how the game is coded so it's unlikely that we'll ever get a chance to see it in game.

[Khobai]

Quote

It's a 12 v 12 team game. How well you play with the team you have is the single most critical direct reflection of how you play.

um no. you can do everything right and play exceptionally and even try to take command and give your team all the right orders and still lose because of potatos on your team.

team rankings and team ladders reflect how well you play as a team.

individual rankings and individual tier score reflects how well you play individually.

you have the two confused.

winning and losing should affect team rankings. but it should have no bearing on individual ranking. only match score should matter for that.

Quote

Saying that your win/loss average is beyond your control is not just false but demonstratively so. Are you saying all the great players in the game with high W/L got that by, what, random chance? Consistently for the last 16 months? All the T5 people with sub-1.0 W/L, that's just random chance? Coincidence?

were obviously talking about the pug queue not playing in organized groups.

and yes in the pug queue you cant really control if you win or not. you could have some super baked potatos on your team that just run out of bounds or start teamkilling eachother and cost you the game. you have zero control over that. you cant control stupid.

so I dont know what youre talking about but it has absolutely no bearing on what I was talking about. because im talking specifically about potatos ruining games. im not talking about highly organized teams that have no potatos.

the only way to keep potatos in the tier where they belong is to rank players based on how they individually perform in games. that prevents potatos from advancing into higher tiers by parasiting off better players. the current system allows even the worst players to eventually reach T1 just by playing enough games. thats unacceptable.

Edited by Khobai, 31 October 2017 - 10:53 AM.

[MischiefSC]

Khobai, on 31 October 2017 - 10:40 AM, said:

um no. you can do everything right and play exceptionally and even try to take command and give your team all the right orders and still lose because of potatos on your team.

You can also do everything wrong and still get carried. We're all swimming in the same ocean; the ebbs and swells raise and lower all equally. However how you play directly impacts the success or failure of your team. That's more than farming damage and ganking kills, it's coordination, supporting teammates and communicating.

Go to the leaderboard. Do you see where some people have a W/L better than others? Do you see where some people win more matches than other people?

You're trying to falsely claim that's random chance. It absolutely isn't. Your performance as a team player impacts your win/loss every single match and it does so more than your damage or kills.

Every single mechanic in the game can be gamed, except win/loss. It's the only accurate indicator of how likely you are to win or lose a match.

The only mechanic that has any real value on predicting your odds of winning a given match are.... how often you've won/lost matches before. That's it. Everything else is just semantics and peripheral data. Did you win and how good was the team you beat. That's 100% as relevant for QP as it is for comp queue.

[Khobai]

Quote

You can also do everything wrong and still get carried.

which is why match score is all that should matter. so that cant happen.

Quote

Go to the leaderboard. Do you see where some people have a W/L better than others? Do you see where some people win more matches than other people?

yeah because they play in groups with other good players. its already been established that if you play in groups you can better control your W/L record than if you pug.

Again I am not talking about playing in organized groups Im talking about pugging it. And you cant control what potatos do. It literally is RNG because you could end up with the teamkilling potato on your team.

Quote

You're trying to falsely claim that's random chance. It absolutely isn't.

of course it is. if you get a team killing potato on your team randomly and lose because of him thats RNG.

dont !@#4ing tell me potatos arnt RNG. theyre unpredictably stupid.

Quote

Every single mechanic in the game can be gamed, except win/loss. It's the only accurate indicator of how likely you are to win or lose a match.

obviously match score isnt perfect and needs some work if it can be gamed. if you fix it so it cant be gamed, then your whole point about being able to game it goes out the window. so well just make the necessary changes to match score. done. fixed.

wins/losses should only matter for team ratings. wins/losses reflect how the team performed as a whole. not how a single individual performed. someone can do jack **** and still win. or they can do everything right, get massive damage, and still lose. wins/losses has no bearing on individual performance.

match score is the only thing that should matter when assigning individual ratings to players. match score reflects how an individual performed.

youre getting getting team ratings confused with individual ratings. theyre two different things.

when assigning individual players to a tier, you want to use match score as the metric. players that get consistently higher match score belong in a higher tier than players that get consistently lower match scores. wins/losses have nothing to do with that.

when ranking teams on a ladder, you want to use win/loss ratio. wins/losses reflect how well the team works together and how good the team is compared to other teams

its the same thing in sports. when they rank players they dont rank them based on whether their team wins or loses. they rank them based on their own personal stats. its only when they rank teams they look at the wins and losses.

Edited by Khobai, 31 October 2017 - 12:01 PM.

[SFC174]

I do think the W/L ratio is a red herring as long as group queues are mixed with solo queue drops. Teamwork/communication is OP and groups tend to do it better - with organized groups being hard to beat.

That said, if we exclude group queue, the randomness of getting potatoes or mechgods on your team should balance out in the long run, which would theoretically leave your W/L as representative of your contribution to the team.

But......

It doesn't often feel that way when you multiple games in a row with 4,6 or 8 potatoes. Part of it is the nature of our perception, we notice the runs more than we notice the randomness. And given that the Tier system is so screwy, its possible that your sample size of games in a given season (for an individual) isn't large enough to ensure you get an even distribution of potatoes and gods. An event which encourages the play of a particular class of mech (or weapon type), might bring out a bunch of guys who play well in other classes/types but suck in that particular event's specialty and you might end up with more of them for that weekend.

We really just don't know enough. The Tier system is so bad (and so biased toward upward movement) that it isn't hard to get a bad run of players, and if you're good, but not good enough to carry, it can tank your W/L ratio.

I can see W/L being important, but not until we actually get zero-sum player ranking and Tiers that mean something.

[Ex Atlas Overlord]

I wonder when people will realize if they go back to 8v8, it won't be anything like it used to be.

The powercreep is real.

[Mr Snrub]

MischiefSC, on 31 October 2017 - 10:18 AM, said:

Several times last weekend?

I've played a bit over 11,000 matches and I can count on one hand the number of times I've seen something like that. Every time I've screenshot it and sent in a support ticket. One of those times I can say with certainty (because I knew some of the people involved) it resulted in short term bans and a serious warning.

You're barking at the wrong tree here. Someone was arguing synch drops don't happen and if so never on a big scale and smuggly asked for screenshots - so I provided one.

I'm not 'wearing a tinfoil hat', thank you, and never commented on the effectiveness of reporting - I didn't report anyone because it wasn't worth the effort.

[Dimento Graven]

Mr Snrub, on 31 October 2017 - 12:37 PM, said:

You're barking at the wrong tree here. Someone was arguing synch drops don't happen and if so never on a big scale and smuggly asked for screenshots - so I provided one.

I'm not 'wearing a tinfoil hat', thank you, and never commented on the effectiveness of reporting - I didn't report anyone because it wasn't worth the effort.

Any unit with a lot of players probably does this, both intentionally and accidentally.

I doubt it occurs very often though.

Yeah, I don't worry about it, it doesn't seem worth any sort of 'enforcement' effort in my opinion.

Eventually they all get bored anyway and either group up in the quick play queue or go do FW...

[ZippySpeedMonkey]

Actually what I would like to see is this . It would be more like the lore.

8 vs.8 ( IS vs. IS )
8 vs. 5 ( IS vs.Clan )
5 vs. 5 ( Clan vs. Clan )

( I know it’s not going to happen but a way to deal with the mech is to limit each side to 400t and then adjust the weight values up or down based on results after testing. )

Edited by ZippySpeedMonkey, 31 October 2017 - 02:32 PM.

[Dimento Graven]

ZippySpeedMonkey, on 31 October 2017 - 02:31 PM, said:

Actually what I would like to see is this . It would be more like the lore.

8 vs.8 ( IS vs. IS )
8 vs. 5 ( IS vs.Clan )
5 vs. 5 ( Clan vs. Clan )

( I know it’s not going to happen but a way to deal with the mech is to limit each side to 400t and then adjust the weight values up or down based on results after testing. )

Long wanted to have the following: 12IS vs 10Clan - remove all buffs, nerfs, and tweaks...

Yep, ain't gonna happen, but we can dream!

[MechaBattler]

Has anyone tried doing a stock IS versus stock clan 8v5 private match? Would love to see a video on that.

[ZippySpeedMonkey]

Dimento Graven, on 31 October 2017 - 02:36 PM, said:

Long wanted to have the following: 12IS vs 10Clan - remove all buffs, nerfs, and tweaks...

Yep, ain't gonna happen, but we can dream!

I picked 8 vs. 5 because that way it would be scalable. In future we could actually have 16 vs. 10, if PGI ever got ambitious enough to make it more lore friendly.

[MischiefSC]

Khobai, on 31 October 2017 - 11:44 AM, said:

which is why match score is all that should matter. so that cant happen.



yeah because they play in groups with other good players. its already been established that if you play in groups you can better control your W/L record than if you pug.

Again I am not talking about playing in organized groups Im talking about pugging it. And you cant control what potatos do. It literally is RNG because you could end up with the teamkilling potato on your team.



of course it is. if you get a team killing potato on your team randomly and lose because of him thats RNG.

dont !@#4ing tell me potatos arnt RNG. theyre unpredictably stupid.



obviously match score isnt perfect and needs some work if it can be gamed. if you fix it so it cant be gamed, then your whole point about being able to game it goes out the window. so well just make the necessary changes to match score. done. fixed.

wins/losses should only matter for team ratings. wins/losses reflect how the team performed as a whole. not how a single individual performed. someone can do jack **** and still win. or they can do everything right, get massive damage, and still lose. wins/losses has no bearing on individual performance.

match score is the only thing that should matter when assigning individual ratings to players. match score reflects how an individual performed.

youre getting getting team ratings confused with individual ratings. theyre two different things.

when assigning individual players to a tier, you want to use match score as the metric. players that get consistently higher match score belong in a higher tier than players that get consistently lower match scores. wins/losses have nothing to do with that.

when ranking teams on a ladder, you want to use win/loss ratio. wins/losses reflect how well the team works together and how good the team is compared to other teams

its the same thing in sports. when they rank players they dont rank them based on whether their team wins or loses. they rank them based on their own personal stats. its only when they rank teams they look at the wins and losses.

Again, flat out wrong.

I have played maybe 10 games in group queue this last month - in which, btw, I lose more than I do in group queue. I do better in solo pug queue than group queue, almost universally. Because in group queue there's no matchmaker at all and you can have huge swings in the caliber of who's in what group on what side. In qp since you're always in a random group within a matchmaking parameter (with very few exceptions) you have far, far more statistical impact on your win/loss. In group queue you can just get consistently carried by your teammates as they are a consistent factor. In pug queue you are the only 100% consistent factor in all your matches.

I absolutely am talking about pug queue. The math isn't hard and it's fundamental statistics.

You are the only consistent factor in all your matches. All the others are random, with a matchmaker. The exact same matchmaker everyone else gets. So any 'random swings' you get are exactly the same as everyone else. Since group queue matches group SIZES that's not the case - if you're in a 4man you're only matched with groups who add up to 12, where as a 10man team only gets matched with 2mans. Also there's the tonnage variable based on your team size. It is *not* the same, not even close.

How you play impacts your teams ability to win. Every single match. So if you are better at the behaviors that win matches (yes, damage and kills but also teamwork, coordination, situational awareness, communication) then you'll win more. That shows out in the law of averages and law of large numbers.

Match score is ****. It's worthless as a measuring tool. If I tried to use something that inaccurate for anything in my line of work to do something like, say, predict a given persons impact on their teams ability to win a match like exactly what a MM is supposed to do, I'd get laughed at and moved to another project or even line of business because I would have no place working in analytics.

The only useful indicator of your ability to win matches is your historic win/loss. Full stop. End of topic. If you want to break down WHY someone has a particular damage or match score or KDR and what impact that has for how that person plays in order to help them improve then yes. Those metrics can be studied.

However, and I don't know how more simply this can be put, for Quickplay there is no metric that is in anyway trustworthy or useful for predicting someones value to winning (or losing) a match than their W/L rate. Nothing else. At all. For any reason, in any way, shape or form, has any place or value in establishing your odds of helping win a match. You could break W/L down more by having a separate stat for individual mechs and loadouts but, again. To be perfectly clear. Only your W/L has any relevance at all for predicting your future W/L value. This applies more in QP than in any other mode in the game including FW and comp queue.

Damage is usually very useful in driving wins. Kills, maybe. Match score? It's modified by your wins but also includes stuff that is often completely irrelevant like component destruction. I maximize match score by taking Streaks or ATMs and by hiding at the end of a match to avoid dying, so if I blow some STs off, steal 2 kills and then go power down my match score average would be higher than if I squirreled 1/4 the enemy team or spent the match running caps.

Players in sports don't get their own Elo score because they don't play in random teams. If they did you would have to give them one and it absolutely would be Elo based because that's the only useful way to rank their ability to win matches.

ZippySpeedMonkey, on 31 October 2017 - 02:40 PM, said:

I picked 8 vs. 5 because that way it would be scalable. In future we could actually have 16 vs. 10, if PGI ever got ambitious enough to make it more lore friendly.

So you think if we just remove all quirks and such that would somehow be more balanced than now? Why do you think that's somehow going to balance the game?

Please. I'm begging here. Someone, please, help me understand where the idea comes from that makes people think that just removing all quirks and going 8 v 5 is somehow magically going to 'balance' Clan vs IS. I really want to understand where the logic there is coming from.

We can skip over the fact that the idea failed completely in tabletop and just go right to 'how is balancing 5 v 8 going to be easier than 8 v 8'?

[Mystere]

MischiefSC, on 31 October 2017 - 02:47 PM, said:

Only your W/L has any relevance at all for predicting your future W/L value.

I really cannot grasp why people find it very hard to realize this.

Edited by Mystere, 31 October 2017 - 05:28 PM.

[Humpday]

oOOOOoo +1 in 8v8

The beta test for the new weapons was a blast