55 replies to this topic

[Scandinavian Jawbreaker]

So, today I was sipping my morning coffee and humming old Black Sabbath songs as usual when I saw this on reddit:



Yeah. I see what you're getting at there Paul, but there's couple of problems.

1. PSR should not be affected if the game mode does not use PSR for matchmaking

Seriously, I really thought this was a no brainer. From an experience (being top units in Tukayyid with >90% winning rate) without matchmaking the scores in CW do not represent individual skill. We did speed runs in Tukayyid, that's not skill, that's abusing solo players to max the wins per hour.

Which brings me to the next concern

2. PSR will be easily manipulated

Yes, very easily. It can be manipulated to either way. If you want to drop you PSR all you have to do is drop to a CW match, do absolutely nothing, exit with 0 match score. Yes, you heard it. CW matches affect only your PSR - no stats tracking so your KDR is just fine.

Works other way around too. Want those precious T1 bragging rights? Drop in CW with a good unit for a week. Just make sure you don't die while still doing damage, because the CW match score is calculated the same way than in PUGs, except it is divided by mechs used - just like XP. Sooo... If you don't die, you get ridiculous match score, check this for a preference: http://mwomercs.com/...tion-bug-in-cw/

As you can see the manipulation is real. It is BAD for the new system and coming tournaments. Before a solo tournament I can just drop my PSR so low in CW that I will be #1 in my first 10 matches with insane score cus of seal clubbing. With no effect on my stats.

3. Inflated PSR for CW units

Ok, let's say you are part of the unit Teenage Mutant Smoke Jaguars and you field a full 12-man every day because you enjoy playing CW. You attack Kurita territory fiercly because Jaguars are fierce and all especially teenage ones so you win quite alot of matches. Most of these matches being pretty much stomps, because defenders are made of solo players who don't have what it takes to beat an organised 12-man in skill or equipment.

Now, this happens for a week. A member from TMSJ then buys a new mech and goes to the public queue to level it up. And gets absolutely obliberated. He must think, what is happening I cannot understand, but here's what's happening: That guy has now been carried to T1-T2 matches where in fact he has no business being in. He does not enjoy the situation nor does his team mates appreciate his presence in those matches. Everyone loses.

4. You bring back the exact thing you wanted to prevent

And that is allowing someone to be carried to a certain level they don't belong. Except now you widen the option. You're allowing someone to DROP DOWN to a tier they don't belong, thus making the possibility for exploits. As I stated, CW doesn't gather stats, it does nothing, no one has nothing to lose when the score is being manipulated.

I hope I covered all the crucial things why this is overall a BAD IDEA and why it should be taken off from the equation as soon as possible. Players will exploit and it will mess up the matchmaking, tournament scores and tiers.

Have a good day, PGI! Sincerely yours,

ugrakarma

[Fate 6]

Yup

[Iqfish]

Fate 6, on 24 August 2015 - 10:58 AM, said:

Yup

qft

[Nothing Whatsoever]

What are the devs thinking?

I want to know what they are evaluating to think this was good to cross-associate. If anything there should be one score for solo, then one score for groups, with maybe CW tossed in if CW would then use PSR, otherwise CW should be separate from quick group drops.

[FupDup]

Dafuq...

[Rasc4l]

OP is 100 % correct.

[stjobe]

[Dimento Graven]

ugrakarma, on 24 August 2015 - 10:52 AM, said:

So, today I was sipping my morning coffee and humming old Black Sabbath songs as usual when I saw this on reddit:



Yeah. I see what you're getting at there Paul, but there's couple of problems.

1. PSR should not be affected if the game mode does not use PSR for matchmaking

Seriously, I really thought this was a no brainer. From an experience (being top units in Tukayyid with >90% winning rate) without matchmaking the scores in CW do not represent individual skill. We did speed runs in Tukayyid, that's not skill, that's abusing solo players to max the wins per hour.

Which brings me to the next concern

I disagree.

PSR is your PILOT SKILL RATING. You are theoretically using PILOT SKILL in any game mode, so it should absolutely be affected, ANY TIME, you're out on the field (expect for private matches of course).

Quote

2. PSR will be easily manipulated

Yes, very easily. It can be manipulated to either way. If you want to drop you PSR all you have to do is drop to a CW match, do absolutely nothing, exit with 0 match score. Yes, you heard it. CW matches affect only your PSR - no stats tracking so your KDR is just fine.

Works other way around too. Want those precious T1 bragging rights? Drop in CW with a good unit for a week. Just make sure you don't die while still doing damage, because the CW match score is calculated the same way than in PUGs, except it is divided by mechs used - just like XP. Sooo... If you don't die, you get ridiculous match score, check this for a preference: http://mwomercs.com/...tion-bug-in-cw/

As you can see the manipulation is real. It is BAD for the new system and coming tournaments. Before a solo tournament I can just drop my PSR so low in CW that I will be #1 in my first 10 matches with insane score cus of seal clubbing. With no effect on my stats.

How do you know this, and give that, at least as far as I know, NO ONE knows what their current Tier is, or their PSR skill, how can you know HOW MUCH the PSR value drops in a '0' skill CW/Public queue drop?

Does it only take one loss to drop a full tier, or 100?

We don't know, so your assumption is, at best, a bit melodramatic...

Quote

3. Inflated PSR for CW units

Ok, let's say you are part of the unit Teenage Mutant Smoke Jaguars and you field a full 12-man every day because you enjoy playing CW. You attack Kurita territory fiercly because Jaguars are fierce and all especially teenage ones so you win quite alot of matches. Most of these matches being pretty much stomps, because defenders are made of solo players who don't have what it takes to beat an organised 12-man in skill or equipment.

Now, this happens for a week. A member from TMSJ then buys a new mech and goes to the public queue to level it up. And gets absolutely obliberated. He must think, what is happening I cannot understand, but here's what's happening: That guy has now been carried to T1-T2 matches where in fact he has no business being in. He does not enjoy the situation nor does his team mates appreciate his presence in those matches. Everyone loses.

I completely disagree again, the scoring is theoretically based on how much you're doing for the entire team. This is the same thing for both CW and public scoring. If you're not doing much for the team, your resulting PSR score is going to be low. If you're doing a LOT for the team, then your score will be higher. There's no artificiality, other than intentionally losing a bunch matches, that can be used to manipulate it.

You stick a noob with 11 elite vets, the noobs score is going to be low, and have little affect on his tier movement. This is the same in CW as it is in the public queues.

Again, no need to panic.

Quote

4. You bring back the exact thing you wanted to prevent

And that is allowing someone to be carried to a certain level they don't belong. Except now you widen the option. You're allowing someone to DROP DOWN to a tier they don't belong, thus making the possibility for exploits. As I stated, CW doesn't gather stats, it does nothing, no one has nothing to lose when the score is being manipulated.

I hope I covered all the crucial things why this is overall a BAD IDEA and why it should be taken off from the equation as soon as possible. Players will exploit and it will mess up the matchmaking, tournament scores and tiers.

Have a good day, PGI! Sincerely yours,

ugrakarma

Again, you're wrong. Regardless of how skilled the other 11 people in the match are, if you don't do much to contribute to the outcome of the match, your score is going to reflect that.

Hell, there are times when I'm in with what is obviously a VERY skilled group of individuals in the group queue, and it's damn hard to 'stand out' with more than one or two kills and more than 400 damage, because there's 11 other individuals with equal or better skill out there on the field. In those matches, interestingly enough, every one's score is relatively similar, maybe at most a 100 points difference between the top scorer and the bottom.

Don't panic.

It's been only ONE week...

Let it ride a bit.

[Clownwarlord]

This explains why my solo and group matches are such trash and I feel like I am getting my ass handed to me every time. I play way to much CW

[Lily from animove]

already said, that, also belongs to suggestions

[Scandinavian Jawbreaker]

Dimento Graven, on 24 August 2015 - 12:26 PM, said:

I disagree.

PSR is your PILOT SKILL RATING. You are theoretically using PILOT SKILL in any game mode, so it should absolutely be affected, ANY TIME, you're out on the field (expect for private matches of course).

PSR is a match making algorithm to give you better match quality. If a game mode does not use that algorithm for match making, it shouldn't affect it at all. Tiers 1-5 are only (soon to be) visual manifestations of this algorithm, there is still a varied numerical value for each player only for the match maker to use. While I am using personal pilot skill in CW, I am not matched against opponents with similiar skill thus messing up the match making algorithm PSR is used for.

Quote

How do you know this, and give that, at least as far as I know, NO ONE knows what their current Tier is, or their PSR skill, how can you know HOW MUCH the PSR value drops in a '0' skill CW/Public queue drop?

Does it only take one loss to drop a full tier, or 100?

We don't know, so your assumption is, at best, a bit melodramatic...

We don't know the exact values, but it is not about the exact values. It is about the fact that there is the possibility it can be manipulated both ways. When tiers go opt in I will do extensive testing with an alt account to see the true effect if this is still in place.

Quote

I completely disagree again, the scoring is theoretically based on how much you're doing for the entire team. This is the same thing for both CW and public scoring. If you're not doing much for the team, your resulting PSR score is going to be low. If you're doing a LOT for the team, then your score will be higher. There's no artificiality, other than intentionally losing a bunch matches, that can be used to manipulate it.

I guess you have no problem mixing up group queue with solo queue then. Would have the same effect. Oh yeah and take out any kind of match making. It is not about what you do, it's about who you do it against. When I started with my alt the first match I played I got Ace of Spades and 900dmg in a trial Hunchback. There's nothing to brag about that neither does it tell anything about my skill against equally skilled opponents that the PSR match making is designed for.

Quote

You stick a noob with 11 elite vets, the noobs score is going to be low, and have little affect on his tier movement. This is the same in CW as it is in the public queues.

I'm not talking about noobs.

Quote

Again, you're wrong. Regardless of how skilled the other 11 people in the match are, if you don't do much to contribute to the outcome of the match, your score is going to reflect that.

I will demonstrate this with a smurf account while doing testing when the tiers go public.

Quote

Don't panic.

I'm not panicing. I'm merely bringing out the possible errors in the thought process of bringing CW results into the equation for public queue match making.

[Mechwarrior Buddah]

PSR should not be manipulated by a BETA game mode

Hell the live server shouldnt be affected by the BETA test of another game mode. Shows Beta means nothing

[Dimento Graven]

ugrakarma, on 24 August 2015 - 01:02 PM, said:

PSR is a match making algorithm to give you better match quality. If a game mode does not use that algorithm for match making, it shouldn't affect it at all. Tiers 1-5 are only (soon to be) visual manifestations of this algorithm, there is still a varied numerical value for each player only for the match maker to use. While I am using personal pilot skill in CW, I am not matched against opponents with similiar skill thus messing up the match making algorithm PSR is used for.

Let's get a little more specific here:

PSR is your score as a pilot. It ranks you based on your performance, and that rank relates to how much you've used your skills in ways that benefits the team you were placed on.

The MM, uses an algorithm that includes PSR, among other things to create a balanced match.

PSR is NOT the algorithm itself, just a component of it.

Quote

We don't know the exact values, but it is not about the exact values. It is about the fact that there is the possibility it can be manipulated both ways. When tiers go opt in I will do extensive testing with an alt account to see the true effect if this is still in place.

You could manipulate elo too, and there were the occasional rumors of top tier clans/units doing just that. Intentionally losing, etc.

There's nothing you can do to stop that, IF, people are going to stoop to that level of douchebaggery. It won't matter if you use elo, PSR, freckle counts, or what have you...

Better to create a system that works for the 'average Joe with honorable intent' than to try and design around the douchebags...

Quote

I guess you have no problem mixing up group queue with solo queue then. Would have the same effect. Oh yeah and take out any kind of match making. It is not about what you do, it's about who you do it against. When I started with my alt the first match I played I got Ace of Spades and 900dmg in a trial Hunchback. There's nothing to brag about that neither does it tell anything about my skill against equally skilled opponents that the PSR match making is designed for.

I play in all queues quite a bit, and don't have a problem, no.

Your example is extreme. You're apparently a highly skilled pilot that started a fresh/rarely used(?) account that is by default ranked in at the middle/top end of Tier 4, and you were apparently matched against others who were ACTUALLY at an appropriate level.

You're going to have to tell me how ANY algorithm is going to handle that situation. A 'new' account is defaulted to a certain level and is raised or lowered based off subsequent performance. I imagine that if you play that alt account a lot, you're going to quickly rise out of the Tier 4 and be placed at the rank you actually deserve.

At that point, you could start a whole new alt account and again, no matter what changes are made, the system is designed to assume a new player is going to be AT BEST, middle/top of Tier 4.

There's no reasonable method around this, and SHOULD have been the same problem experienced in elo, BUT BECAUSE elo was broken in MM, noobs and unskilled players were regularly tossed in with the sharks.

The changes in MM and use of PSR will reduce that scenario significantly (unless of course we have a lot of top tier players creating new accounts just so that they can go kick noob ass, but again, if/when that happens it's NOT because MM or PSR is broken).

Quote

I'm not talking about noobs.

Wasn't that the whole point of the example you just gave?

Quote

I will demonstrate this with a smurf account while doing testing when the tiers go public.

What? You'll demonstrate that a person who is highly skilled at the game, but creates a new account to receive an artificially low PSR rank will kick new and unskilled players asses?

I'm sorry, there's a logical disconnect I'm missing here.

You apparently are mad that because the PSR (or elo for that matter) can't truly know a brand new player's skill, that it's broken. If that's not your point, please explain it to me, again, I'm missing something here.

Quote

I'm not panicing. I'm merely bringing out the possible errors in the thought process of bringing CW results into the equation for public queue match making.

I dunno, you seem to verging on histrionics. Your logic, currently, seems faulty (I'm waiting for you to fill in the gaps for me, you may have a valid point and I just missed it).

[Scandinavian Jawbreaker]

Dimento Graven, on 24 August 2015 - 01:20 PM, said:

Let's get a little more specific here:

PSR is your score as a pilot. It ranks you based on your performance, and that rank relates to how much you've used your skills in ways that benefits the team you were placed on.

The MM, uses an algorithm that includes PSR, among other things to create a balanced match.

PSR is NOT the algorithm itself, just a component of it.

The only "other things" is weight class, so in the end PSR value is the MM algorithm. I don't know where you're getting at with this. Might be misunderstanding cus English is not my native language and I can be using terms wrong.

Quote

You could manipulate elo too, and there were the occasional rumors of top tier clans/units doing just that. Intentionally losing, etc.

There's nothing you can do to stop that, IF, people are going to stoop to that level of douchebaggery. It won't matter if you use elo, PSR, freckle counts, or what have you...

Better to create a system that works for the 'average Joe with honorable intent' than to try and design around the douchebags...

Yes, but tanking your Elo meant also tanking your KDR and WLR. With CW you can tank PSR without taking hits to those stats. You can also (according to latest poll result) opt out when not T1 ot T2 (tanking your tier for a tournament for example) and opt back in when you are at appropriate level for you liking. I agree, far fetched, but still a possible exploit.

Quote

I play in all queues quite a bit, and don't have a problem, no.

Your example is extreme. You're apparently a highly skilled pilot that started a fresh/rarely used(?) account that is by default ranked in at the middle/top end of Tier 4, and you were apparently matched against others who were ACTUALLY at an appropriate level.

You're going to have to tell me how ANY algorithm is going to handle that situation. A 'new' account is defaulted to a certain level and is raised or lowered based off subsequent performance. I imagine that if you play that alt account a lot, you're going to quickly rise out of the Tier 4 and be placed at the rank you actually deserve.

At that point, you could start a whole new alt account and again, no matter what changes are made, the system is designed to assume a new player is going to be AT BEST, middle/top of Tier 4.

There's no reasonable method around this, and SHOULD have been the same problem experienced in elo, BUT BECAUSE elo was broken in MM, noobs and unskilled players were regularly tossed in with the sharks.

The changes in MM and use of PSR will reduce that scenario significantly (unless of course we have a lot of top tier players creating new accounts just so that they can go kick noob ass, but again, if/when that happens it's NOT because MM or PSR is broken).

My example was to compare a CW setting in PUG environment. ATM T1s are fighting T5s in CW, but not in public queue, yet the result is taken account for PSR that should prevent this from happening.

Quote

Wasn't that the whole point of the example you just gave?

No. The quote was a response for my point 3 in OP and I wasn't talking about noobs in my OP.

Quote

What? You'll demonstrate that a person who is highly skilled at the game, but creates a new account to receive an artificially low PSR rank will kick new and unskilled players asses?

No. Demonstrate how PSR can be manipulated using CW. I hope they fix this by then though.

Quote

You apparently are mad that because the PSR (or elo for that matter) can't truly know a brand new player's skill, that it's broken. If that's not your point, please explain it to me, again, I'm missing something here.

I dunno, you seem to verging on histrionics. Your logic, currently, seems faulty (I'm waiting for you to fill in the gaps for me, you may have a valid point and I just missed it).

I don't know where you got the idea I am talking about new players. I'm not. I'm not mad either, just pointing out possible exploits and problems when including CW stats to public queue matchmaking. I love PSR and it's improved match quality alot compared to Elo and I want to see it work properly.

[H I A S]

I disagree at one Point with ugra: I think ELO was much better than PSR, if we use different ELO's for Solo and Group. But thats only my opinion.
To the Rest: ugra is right and his Points are obviuse. CW is most of the Time seal clubbing and it should not have an impact for PSR.

[Ace Selin]

Totally agree with OP, remove CW from PSR equation

[Dimento Graven]

ugrakarma, on 24 August 2015 - 01:56 PM, said:

The only "other things" is weight class, so in the end PSR value is the MM algorithm. I don't know where you're getting at with this. Might be misunderstanding cus English is not my native language and I can be using terms wrong.

Then we're disconnecting, PSR is NOT the MM algorithm, and yes, 'mech weight IS another major component and taken together ensure the matches your drop in have the best chance to be balanced.

Quote

Yes, but tanking your Elo meant also tanking your KDR and WLR. With CW you can tank PSR without taking hits to those stats. You can also (according to latest poll result) opt out when not T1 ot T2 (tanking your tier for a tournament for example) and opt back in when you are at appropriate level for you liking. I agree, far fetched, but still a possible exploit.

I'm sorry but only fools and epeen obsessive idiots worry about KDR and WLR. Those are truly meaningless stats, and you can't tell me you still lose in a big way without dying, and not killing (thus affecting your KDR and WLR) with PSR.

This is absolutely much ado about ZERO.

Quote

My example was to compare a CW setting in PUG environment. ATM T1s are fighting T5s in CW, but not in public queue, yet the result is taken account for PSR that should prevent this from happening.

The primary difference between CW and the public queues is that the MAJORITY of regular players understand this is a 'team oriented' game and are communicating at a much more cooperative and detail oriented level.

My experience is that the typical public queue player, when playing CW with the appropriate attitude ends up doing much better in CW. NOW, you can have the recalcitrant ******** who will invariably still try and 'Rambo' their way to a solo win in CW, those fools will lose, lose big, and probably garuntee your side has lost, BUT BECAUSE the 11 other people are playing intelligently and for a 'team' goal, their scores will typically reflect that, and the affect on Tier status will most probably be neutral at its worst. You'd have to play spectacularly badly to negatively affect your Tier ranking in CW.

Quote

No. The quote was a response for my point 3 in OP and I wasn't talking about noobs in my OP.

Then there's a point I'm missing. You can't be concerned for and dismissive of the affect of PSR on noobs.

Quote

No. Demonstrate how PSR can be manipulated using CW. I hope they fix this by then though.

No you demonstrate that a new account created and played, regardless of CW or public queue will quickly be placed in the Tier they deserve through their play. Noob, fake noob, or plain ol' ain't got no skill player alike. It won't matter.

So you go into CW with a new character, get some great scores and then go into the public queue, you'll find you'll be dropping against players with similar skills.

This is win-win as far as I'm concerned, and proof positive that absolutely CW scores should affect PSR.

Quote

I don't know where you got the idea I am talking about new players. I'm not. I'm not mad either, just pointing out possible exploits and problems when including CW stats to public queue matchmaking. I love PSR and it's improved match quality alot compared to Elo and I want to see it work properly.

I got that idea from your example referencing what I understood to be a 'new' or at least 'rarely played' account, which is exactly like a "NOOB".

I'm sorry, but nothing you've stated shows me that there's any real problem with CW scoring affecting PSR. PSR is based off your in game actions that theoretically benefit your team. The same actions that will be taken regardless of whether or not your playing the solo queue, the group queue, the CW queue in a pre-made group, a CW pick up group.

If your play style is completely solo oriented your PSR, regardless of which mode you primarily play in, will reflect that. If your play style is completely oriented around cooperating with 11 other people, your PSR will reflect that. The degree of affect on PSR those play styles have will be based on how skilled you are at those play styles. If you're good enough to go off on your own ignoring 11 other people but still crank out tons of points in damage, assists, skills, spotting, component loss, etc. etc. etc., you're going to have a good PSR. If your skills with a solo play style mean you're only getting lots of kills, very few assists, no spotting, etc., then your PSR will be low.

And that's going to be true in any game mode.

And that's how it should be.

[bad arcade kitty]

their elo remake is incredibly unprofessional

[Dimento Graven]

bad arcade kitty, on 24 August 2015 - 02:16 PM, said:

their elo remake is incredibly unprofessional

How so?

[Escef]

ugrakarma, on 24 August 2015 - 01:56 PM, said:

The only "other things" is weight class, so in the end PSR value is the MM algorithm. I don't know where you're getting at with this. Might be misunderstanding cus English is not my native language and I can be using terms wrong.

Well, you ARE on the internet. The first thing I suggest you do is look up what the definition of an algorithm is. Because if you are seriously saying that the PSR value is the match maker algorithm than you have no idea what the words you are saying actually mean.