Jump to content

A Proposal Of How To Fix The Match-Maker


51 replies to this topic

#21 P H O T O N

    Member

  • PipPipPipPipPip
  • Wild Dog
  • Wild Dog
  • 120 posts

Posted 22 January 2022 - 02:02 AM

View PostPotato Farmer, on 21 January 2022 - 05:15 PM, said:

You know MWO is dying when bunch of nobodies start claiming that they're top tier.


My claim is justified and proven.

What is more damning about the state of this game, is that in trying to make this game better and more fair, a player who has recently founded a MWO community in the UK to help keep this game alive, and who has spent the time to collect, present, and comment on the evidence that the match-maker is broken, is set upon by a very specific group of trolls. Rather telling if I do say so myself. Should I have expected any different - probably not!

The evidence is there for you to see in my original posts, take it or leave it.

As a closing remark from me, PGI now more than ever needs to publish exactly how the match-maker works.

Edited by Titan Prometheus, 22 January 2022 - 03:26 AM.


#22 SirVile

    Rookie

  • The 1 Percent
  • The 1 Percent
  • 2 posts

Posted 22 January 2022 - 03:15 AM

View PostI O O percent KongLord, on 21 January 2022 - 03:52 PM, said:

I love the "evidence" you post is basically you proving that you are the problem when you can't even reach 1kd on the mechs you claim to do good in.

Yes the matchmaker sucks but it should put noobs like OP with tier 5's and leave tier 1 for good people



Not very nice was it? He points out a problem with statistical evidence, but your response is "lool, get good n00b".

You're clearly not very nice. [redacted]

View PostPotato Farmer, on 21 January 2022 - 05:15 PM, said:

[redacted]


If he's tier 1 is he not top tier? Is there another tier above this? OMG? How did I not know?
[redacted]

Edited by Ekson Valdez, 25 January 2022 - 11:34 PM.
quote clean-up


#23 SirVile

    Rookie

  • The 1 Percent
  • The 1 Percent
  • 2 posts

Posted 22 January 2022 - 03:22 AM

I've been on the receiving end of a ROFLSTOMP and, occasionally, the giving end. I'm a tier 4 so essentially shocking. I play for fun, not points. Yet the fun is drained away if the game is a one sided clubbing fest. Like the Clan v IS campaign - which I found to be a frustrating time (12 clan mechs vs 12 IS mechs was no fun) - balance needs to be struck. If you're dropping in a mixed skill tier unit another way to match players to a game should be found. The game company has the states, they could do it.

#24 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 22 January 2022 - 03:44 AM

View PostGenOrtega, on 22 January 2022 - 01:57 AM, said:

Well I see some raw nerves have been struck. This was expected as some rather like the present eco-system as they gain to benefit the most at the expense of others and the health of the game itself. How selfish.

The evidence is there across the net of bad behaviour amongst some of the community that take advantage of a misfunctioning MM, do the research yourself its there and I'm not going to waste time or break rules pointing fingers. Want evidence, go out and look, it shouldn't take you too long, if you participate in this behaviour then you don't have to look far do you ehh?

Ohh and I see some of you want this thread locked judging by the responses, swept under the carpet ehh, you'd like that wouldn't you?


So you've made a broad and far reaching statement and when requested to even provide where your claim exists, you won't even answer that? Just "find it yourself"?

And my asking for even just the locations of your claims is me wanting it swept under the carpet?


Pretty safe to say with that type of response it quite simply doesn't exist and you're making it up. Good discussion.

#25 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 22 January 2022 - 04:37 AM

View PostTitan Prometheus, on 22 January 2022 - 01:13 AM, said:

This is a team game, how can I be underperforming to the extent that makes the team lose?


You showed how on Page 1 with your mech stats. For a Tier 1 player they are so far below the average I’m legitimately not sure how you’ve stayed in Tier 1.

As a Tier 1 player I can swing the battle all the time. If you are unable then do you think that is still the MMs fault for your lack of performance? C'mon...

View PostTitan Prometheus, on 22 January 2022 - 01:13 AM, said:

Spoiler


Doing damage is not the be all and end all – it is effective damage that wins matches.

The player doing 400dmg 2-3 kills/KMDDs etc is going to rank better than the guy spraying and praying doing 500dmg, no kills/KMDDs etc etc.

Unfortunately right now the delta between those two isn’t actually great enough – it should larger such that the guy that is more efficient with the damage should get even better rewards as that is clearly a showing of player skill – Remember those two words from my last post?

View PostTitan Prometheus, on 22 January 2022 - 01:13 AM, said:

As I mentioned in my post I am often dropping in mixed tier groups as I have just started a UK Discord for this game in an attempt to keep it alive - I drop with a lot of folk in tier 3 and below and that should not break the match-maker. That perhaps should factor into the reasoning you present, but all things combined clearly indicates to me and numerous others that the match-maker is at fault, not me.


Great! You started a discord and play in mixed groups, pat on the back for you sir.

As a T1 player in T3-T5 games, when grouping with lower tier friends, you should be over performing by a number of magnitudes. That clearly isn’t happening, you showed us. There is only one person responsible for that and it isn’t the MM.

I know when I play in mixed groups my stats shoot into the stratosphere because as a Tier 1 player, playing in a lower tier match (with friends) means I am at a significant advantage.

This to me just looks like you are just looking for a scapegoat here for your own poor play.


View PostTitan Prometheus, on 22 January 2022 - 01:13 AM, said:

Spoiler


Over the 500 games Jarls takes to rank people (which it states if you read the site) – You argument about Tiers not being considered becomes completely irrelevant. I say this as most players settle into their appropriate tier within 200-400 games for the most part.

The only real flaw is your understanding of it and how to read/use the data it gives out.

Check out the WLR vs Match Score Graph – HERE

With 280 AvgMS, you should be, around 1.5 WLR. Everyone else is… So if everyone else is able to do it over 100s, 1,000s of players, are you really going to continue blaming the MM?


Before you reply again as I can tell you won't be able to resist - Put your "I'm Tier 1" mindset aside and read what I've posted without your bias and you will see that everything I'm saying is making very good sense.

#26 P H O T O N

    Member

  • PipPipPipPipPip
  • Wild Dog
  • Wild Dog
  • 120 posts

Posted 22 January 2022 - 05:12 AM

View Postjustcallme A S H, on 22 January 2022 - 04:37 AM, said:


You showed how on Page 1 with your mech stats. For a Tier 1 player they are so far below the average I’m legitimately not sure how you’ve stayed in Tier 1.

I stay in tier 1 because generally I do well. I almost always get an up arrow even if the team loses! Which ever way you want to interpret that - it is clear the match-maker is broken.

The mech stats that I posted are not all of my mechs, clearly! I have others which when I pilot them the team does well in. Others still were the team does bad but I do good.

View Postjustcallme A S H, on 22 January 2022 - 04:37 AM, said:

This to me just looks like you are just looking for a scapegoat here for your own poor play.

You just seem to have an agenda against changing the match-maker. Just look around, everyone knows it is broken. You appear in all the threads with your straw-man arguments and strive for the last say on the matter. Why are you so biased?

View Postjustcallme A S H, on 22 January 2022 - 04:37 AM, said:

Over the 500 games Jarls takes to rank people (which it states if you read the site) – You argument about Tiers not being considered becomes completely irrelevant. I say this as most players settle into their appropriate tier within 200-400 games for the most part.

Not true. I have tier 5 people in my Discord who are going to post W/L ratios of nearly 2 and average match scores of around 350 this season. They stay tier 5 because of the broken match-maker. When we all drop together, they mostly get down arrows.

View Postjustcallme A S H, on 22 January 2022 - 04:37 AM, said:

The only real flaw is your understanding of it and how to read/use the data it gives out.

No. I understand statistics very will I would have you know. Let's just say it's part of my profession!

View Postjustcallme A S H, on 22 January 2022 - 04:37 AM, said:

Check out the WLR vs Match Score Graph – HERE

With 280 AvgMS, you should be, around 1.5 WLR. Everyone else is… So if everyone else is able to do it over 100s, 1,000s of players, are you really going to continue blaming the MM?

Clearly the match-maker does work for some (perhaps they know how to exploit it?), but it doesn't work for almost everyone else. It is well documented and you just need to acknowledge that - I have seen you post in other threads on the matter and you always try to shut people down.

View Postjustcallme A S H, on 22 January 2022 - 04:37 AM, said:

Before you reply again as I can tell you won't be able to resist - Put your "I'm Tier 1" mindset aside and read what I've posted without your bias and you will see that everything I'm saying is making very good sense.

It's funny, I have never actually seen you in game. For that matter, I have never seen you appear in games with any streamers (just for your info I have a fair amount) who are clearly at the top of the game. Obviously we are not in the same games or playing with the same people. What could that mean I wonder - perhaps there's a clue as to how the match-maker is working?

Edited by Titan Prometheus, 22 January 2022 - 05:28 AM.


#27 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 22 January 2022 - 05:38 AM

View PostTitan Prometheus, on 22 January 2022 - 05:12 AM, said:

You just seem to have an agenda against changing the match-maker. Just look around, everyone knows it is broken. You appear in all the threads with your straw-man arguments and strive for the last say on the matter. Why are you so biased?


Not at all accurate, look around. I've been regularly talking about secondary lobby balancing and other MM changes/issues for over a year. Additionally, Cauldron has had initial discussions around the MM with PGI.

You could probably google that and come up with dozens of results associated with my name across many mediums - I know - because I've been saying it constantly.

Again, you really, really need to look objectively. If you are seeing my posts "appear" yet you don't even know about the MM changes I regularly post about... Well... Posted Image

View PostTitan Prometheus, on 22 January 2022 - 05:12 AM, said:

Not true. I have tier 5 people in my Discord who are going to post W/L ratios of nearly 2 and average match scores of around 350 this season. They are going to be world champions on Jarl's - how does that make you feel?

No. I understand statistics very will I would have you know. Let's just say it's part of my profession!


The laws of averages will sort that out in time. If they are over performing, they will climb in Tier quite fast and then meet more skilled opponents and all will be right in the world. The laws of averages will sort that out in time as it always has. Nothing to worry about but then you should already know that file:///C:/Users/Ash/AppData/Local/Temp/msohtmlclip1/01/clip_image001.png

View PostTitan Prometheus, on 22 January 2022 - 05:12 AM, said:

Clearly the match-maker does work for some (perhaps they know how to exploit it?), but it doesn't work for others. It is well documented.


Yes, there are issues with the MM in some cases/matches, they are known, and no one has tried to deny it.

When there are 100,000 matches a month and there are generally a handful of issues posted that I see across the mediums I pointed out earlier - again just looking at the numbers - it's not anywhere near as large of a problem as people like to make out. When people do some of their own record keeping (dubstep albatros is one user that does) post up the stats/data the reality is nothing like the sensationalist corner makes out.

View PostTitan Prometheus, on 22 January 2022 - 05:12 AM, said:

It's funny, I have never actually seen you in game. Obviously we are not in the same games or playing with the same people. What could that mean I wonder - perhaps there's a clue as to how the match-maker is working?


Timezones do that. I rarely overlap with EU players. I've seen you on comms on a few streams I watch, and you are generally well and truly out of position and/or making bad calls on comms. So I kinda knew coming into it after reading the first post where the issues were and that was confirmed the more I read - your WLR issues are a direct reflection of your own performance as I've said numerous times and the public data backs that up.

#28 Aivazovsky

    Member

  • PipPipPipPipPipPipPip
  • The Dragoon
  • The Dragoon
  • 801 posts

Posted 22 January 2022 - 05:58 AM

I know this guy. The reason his team loses with him and he's on tier 1 is because he makes bad calls. He calls something in VOIP, for example "push now", but he does something completely different, using teammates as a meat shield. At the same time, his calls are really very bad, he does not understand what is happening on the battlefield. Аs an individual fighter he just mediocre player, it's a pity that the system allows such players to be on tier 1.

#29 John Bronco

    Member

  • PipPipPipPipPipPipPip
  • The Fighter
  • The Fighter
  • 966 posts

Posted 22 January 2022 - 06:33 AM

View PostTitan Prometheus, on 22 January 2022 - 12:47 AM, said:

And your evidence for this? Are you aware that this is a team game yet? Do you understand statistics? If not let's actually have a discussion.


Yes please continue questioning my understanding of statistics while posting shots of mechs with 13 games played and trying to argue the matchmaker is conspiring against you.

My evidence is actual good players win more than they lose, just check the leaderboards (and check them prior to May 2020 if you're going to use a group argument). Your win/loss is around 1 which means you neither help nor hurt your team. You're pretty much the definition of an average player in tier 1.

Edited by John Bronco, 22 January 2022 - 06:34 AM.


#30 Dogmeat1

    Member

  • PipPipPipPipPip
  • CS 2020 Gold Champ
  • CS 2020 Gold Champ
  • 123 posts
  • LocationButte Hold

Posted 22 January 2022 - 07:46 AM

View PostTitan Prometheus, on 22 January 2022 - 12:47 AM, said:


I think you are missing the point. I may not be piloting those mechs perfectly (who can?) but I am certainly not piloting them poorly. Look at the match stats. The team loses far more often than it wins - in most cases the losses are completely one-sided 12-0 type affairs. You try pulling more damage in those situations.

This is a team game. It is impossible to explain the W/L ratio unless the match-maker is broken. If the match-maker actually worked, there wouldn't be significant deviation from a W/L ratio of 1 in a team game such as this.


That is poor logic and an extreme amount of mental gymnastics to try to deflect responsibility away from yourself. Your individual efforts in each match have a large influence on your team's odds of winning and no MM system can change that. The matchmaker does have a lot of issues; predominantly the mixing of groups and solo players (which breaks weight class matching) and the fact that it forms mainly just 3 types of games, tier 1-3, tier 2-4 and tier 3-5 ones which results in large skill gaps. However those are the conditions that we all play under and your winrate is a direct result of your performance under those conditions.

Players who use their mechs well and get kills and damage efficiently will have consistently high winrates. They may occasionally have short-term losing streaks but long-term if they are playing well then their winrate will be well above 50%. Your very subpar winrates on the mechs you listed are a result of you using them poorly. Your damage on them is below average and your average kills per match is extremely bad. When you play these mechs you are not carrying your weight as a player and thus your team has a higher chance of losing. I haven't seen your builds on them so I can't comment if they are good are not but the fact is you don't do well in them despite your claims.

Edited by Dogmeat1, 22 January 2022 - 07:46 AM.


#31 Potato Farmer

    Member

  • PipPipPip
  • The Jaws
  • The Jaws
  • 60 posts
  • LocationSingapore

Posted 22 January 2022 - 09:19 AM

View PostTitan Prometheus, on 22 January 2022 - 01:01 AM, said:

To the ignorant trolls (no names mentioned) that doubted my claim (or even accused me of being a liar - shame on you buddy and if you wanted to devalue your alleged medal you managed to quite spectacularly even if I do say so myself! Well played old chap!) about being at the top of tier 1.

https://ibb.co/pRZ3RKv

I would have posted the image to this forum but I cannot for some reason.

p.s. All comments snipped and shared for posterity.



Ahhhhh the good ol EXP bar aye? Well, reality check for you: it means absolutely nothing that your bar is full. All it means is you do better than 50% of players in matches most of the time.

That does NOT mean that you do well. All it means is you are doing ENOUGH.

Sheesh. Way to mislead players into thinking they are better than they actually are. PSR system has its faults and this is one of them.

#32 P H O T O N

    Member

  • PipPipPipPipPip
  • Wild Dog
  • Wild Dog
  • 120 posts

Posted 22 January 2022 - 10:46 AM

View PostPotato Farmer, on 22 January 2022 - 09:19 AM, said:



Ahhhhh the good ol EXP bar aye? Well, reality check for you: it means absolutely nothing that your bar is full. All it means is you do better than 50% of players in matches most of the time.

That does NOT mean that you do well. All it means is you are doing ENOUGH.

Sheesh. Way to mislead players into thinking they are better than they actually are. PSR system has its faults and this is one of them.


Actually old bean, I'm afraid to say that by definition it does mean one does well in the games that one partakes in. Incidentally, nobody said having the bar all the way to the right in tier 1 meant anything other than being top-tier 1 - again by definition it is irrefutable. Let's now focus the discussion on the match-maker shall we and not partake any longer in logical fallacies, yours being ad hominem.

View PostDogmeat1, on 22 January 2022 - 07:46 AM, said:


That is poor logic and an extreme amount of mental gymnastics to try to deflect responsibility away from yourself. Your individual efforts in each match have a large influence on your team's odds of winning and no MM system can change that. The matchmaker does have a lot of issues; predominantly the mixing of groups and solo players (which breaks weight class matching) and the fact that it forms mainly just 3 types of games, tier 1-3, tier 2-4 and tier 3-5 ones which results in large skill gaps. However those are the conditions that we all play under and your winrate is a direct result of your performance under those conditions.

Players who use their mechs well and get kills and damage efficiently will have consistently high winrates. They may occasionally have short-term losing streaks but long-term if they are playing well then their winrate will be well above 50%. Your very subpar winrates on the mechs you listed are a result of you using them poorly. Your damage on them is below average and your average kills per match is extremely bad. When you play these mechs you are not carrying your weight as a player and thus your team has a higher chance of losing. I haven't seen your builds on them so I can't comment if they are good are not but the fact is you don't do well in them despite your claims.


This is a team game. I have posted statistics that are statistically significant and clearly show the match-maker is broken. I have argued why it is not me who is losing those games for my team, now you are telling me that it is me without even fully grasping the arguments and statistics I have put forward. Your argument is both a straw-man and ad hominem, congratulations! Why are you even arguing is beyond me, it's as if you just want to argue for the sake of it. Wait, oh I see!

View PostJohn Bronco, on 22 January 2022 - 06:33 AM, said:


Yes please continue questioning my understanding of statistics while posting shots of mechs with 13 games played and trying to argue the matchmaker is conspiring against you.

My evidence is actual good players win more than they lose, just check the leaderboards (and check them prior to May 2020 if you're going to use a group argument). Your win/loss is around 1 which means you neither help nor hurt your team. You're pretty much the definition of an average player in tier 1.


Focus on the other stats if you don't find a single one convincing. What about the TDR-5SS? Or the Urbie?
I don't disagree that good lances working on comms in a four-man will produce good stats, but players dropping solo or in mixed tier groups will probably not - as I have evidenced. If the match-maker worked properly, as the game is limited to four-man groups, by definition everyone's W/L ratio should be around 1.

View PostAivazovsky, on 22 January 2022 - 05:58 AM, said:

I know this guy. The reason his team loses with him and he's on tier 1 is because he makes bad calls. He calls something in VOIP, for example "push now", but he does something completely different, using teammates as a meat shield. At the same time, his calls are really very bad, he does not understand what is happening on the battlefield. Аs an individual fighter he just mediocre player, it's a pity that the system allows such players to be on tier 1.

I don't think we have ever had the pleasure of playing together and I certainly do not know you. If you were genuinely in those games that I lost for the team single-handedly by speaking, can I ask why you didn't communicate in the game yourself? Good try though old bean! Focus on the matchmaker please, yours is an ad hominem argument.

Edited by Titan Prometheus, 22 January 2022 - 11:56 AM.


#33 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 22 January 2022 - 11:59 AM

Have you thought that maybe he doesn't have a mic? Maybe English isn't his primary language and he doesn't feel comfortable?
For someone that claims to work in statistics - which requires highly logical thought - you are displaying nothing that resembles logic here at all.


https://en.m.wikiped.../wiki/Straw_man

Truths that your use comms poorly, run bad builds and make positioning mistakes - which are all impacting your WLR - are not strawman arguments at all as they go against the clear meaning of what that is.

End of the day the MM is not causing YOUR WLR to suffer, you are. Until you accept that then this discussion isn't going to progress.


#34 Potato Farmer

    Member

  • PipPipPip
  • The Jaws
  • The Jaws
  • 60 posts
  • LocationSingapore

Posted 22 January 2022 - 07:48 PM

Let's not go too deep into math territory guys.

Matchmaker for all it's faults, places players on random teams. If you play enough matches you should have even matches of bad or good players on your team.

Now the only non-variable in each match is YOU
Let's say you're performing well like you said, there is no reason for you to be losing more than winning in the long run. Stats do not lie.

I suggest you read up on dunning-kruger syndrome OP. Maybe you'll learn something about yourself. If you don't, well, it just proves my point.

#35 Arnetheus

    Member

  • PipPipPip
  • 69 posts

Posted 23 January 2022 - 10:44 AM

Tbh i'm not even sure what's the point of the initial suggestion is, after reading through some replies.
Considering OP seems to dislike Jarl, yet wants to make MM work off the in-game average MS between different players... which would require PGI creating a playerbase stat-checking script similar to Jarl.
Otherwise, people were asking for similar things/secondary lobby balancing ever since the PSR rework a year ago, nothing new here too.

Quote

Incidentally, nobody said having the bar all the way to the right in tier 1 meant anything other than being top-tier 1 - again by definition it is irrefutable.

Wait... you're not thinking that Tier 1 players with a maxed bar are equal to each other in "actual" skill, right?
I'm not sure if you do, so just gonna wait for an answer. It's just that "top of the Tier 1" thing you keep mentioning is a bit confusing.
Anyway, going back to the original post for posterity sake:

Quote

Match score should not be inflated if the team wins and should not be depressed if the team loses.

Agree.
People having a good match but getting no win and being punished for it is weird. And on the opposite, players who barely did anything but get free MS just for their team winning. Not fair.

Also, remove suicide MS penalty. A thing that should not exist in the first place. When people see someone throwing games on purpose, they see it. The rest is a matter of reporting/sending a video to PGI. Nine times out of ten, automated MS penalty punishes people either "going out in a blaze of glory" or just being Chrissed.

Quote

More match score should be awarded for capturing flags, being in the circle, spotting etc, and should not just prioritise inflicting damage.

Disagree, at least for the actions you mention. This is a session based FPS with matches lasting around 4-5 minutes on average (at least in T1-3 bracket). People were annoyed for a while now with 3-4 light mech capping premades trolling games as it is. Having to wait for a game up to 5 minutes or longer, only to drop on one of the assault matches and see it end in 2 minutes with nobody shooting, quite a waste of time, imo.
Capture points, domination circle, bases in assault. Those are just areas where fighting is supposed to take place, as opposed to the freedom of skirmish. That's all.
What should be actually raised though, is MS gain for kills, solo kills and kmdd.

Quote

This will give the lighter mechs a chance. Match score should be weighted by tonnage.

Um, it already is. Light and medium mechs get MS bonuses for protecting heavies and assaults. Also, you're basically talking Jarl again, but as official in-game version.

Quote

The tier system should be abolished and replaced with ranks for average match score instead [...] should be publicly visible in the loading screen and on player profiles

While i personally would be ok with this, something tells me the majority of the playerbase will not take it kindly. Mostly because it will lead to people seeing their "real" standing in the food chain. Again, a thing Jarl already does, unofficially.
And for all those types screaming how toxic MWO is, this is one of the few things that might unironically lead to toxicity increasing = )
With the years old rift between some of the more vocal parts in both higher and lower ends of skill spectrum, not sure how "healthy" this feature will be, in a game long past it's "esports" prime (if it ever had it).
Also:

Quote

If the match-maker worked properly, as the game is limited to four-man groups, by definition everyone's W/L ratio should be around 1.

The only scenario of it happening, is MWO having millions of players, so that variety of skill levels would be properly populated. Even then, games already having those millions don't even work that well. The closer you get to the extremes of high and low skill base, the further it deviates from 1.

#36 P H O T O N

    Member

  • PipPipPipPipPip
  • Wild Dog
  • Wild Dog
  • 120 posts

Posted 25 January 2022 - 05:55 AM

I have purposely given some time now for this thread to calm down. The focus of this thread is to discuss the match-maker and so I shall now return to discuss it. If you don't want to discuss it, or if you want to make straw-man or ad hominem arguments, don't post.

To add to my case that the match maker is broken, please consider the following game linked below and decide for yourself if you think it was fair. For what's it's worth I was dropping solo at the time but I believe the alpha lance and possibly the charlie lance on the red team were dropping as groups and were on comms (to substantiate this claim I played a few more games after this and the same people appeared together in lances). I will post the average Jarl's ranking of each player below and state my own conclusions.

https://ibb.co/5hD9P8z

Please note the performance of the charlie lance on the blue team which is indicative of a common issue that I am observing in that in one-sided games, one lance seems to perform at a lower level to the others (although in this case it admittedly was marginal!). In such cases the lower performing lance collapses first shortly followed by the rest of the team like dominos as they are then outnumbered - this seems to happen no matter how good the rest of the team are.

Edited by Titan Prometheus, 26 January 2022 - 08:28 AM.


#37 P H O T O N

    Member

  • PipPipPipPipPip
  • Wild Dog
  • Wild Dog
  • 120 posts

Posted 25 January 2022 - 06:21 AM

I have stated the many caveats of using Jarl's list as a measure of player skill in my previous posts on this thread and I will not repeat them again. Lacking any other publicly accessible measure of player skill, for the sake of the argument to follow let us commit to the null hypothesis that the Jarl list's average percentile is representative of a player's skill rating. In the case where the player is listed as retired on Jarl's I have taken their peak percentile for ease of computation, and duly noted when I have done this.

The team compositions are thus:

Red Team

Alpha
Navid A1 99.69%
Pelassaria B Wolf 99.4% (retired)
Temporary Axis 98.81%
justcallme A S H 99.89%

Bravo
A Disguised Urbanmech 82%
WolfAK80 72%
sandmann3225 87%
MasoXXX 79%

Charlie
Solovey Razbounik 99.76%
HairyTickles 99.56%
oultkat 70% (though their peak is 96.31%)
SirBustaNut 95.83% (retired)

Average player percentile 90.25%.

Blue Team

Alpha
Lokos 80%
Cyxov 50%
DeadWeight18 49%
crawlero 59%

Beta
NumberFive 68%
Titan Prometheus 92%
Drawing Flies 94% (retired)
consentualPPC 84%

Charlie
Jo farkonar 67%
Fleshgrind 71%
Carab Culln 61%
PrismArc 74%

Average player percentile 70.75%.

Edited by Titan Prometheus, 26 January 2022 - 08:44 AM.


#38 P H O T O N

    Member

  • PipPipPipPipPip
  • Wild Dog
  • Wild Dog
  • 120 posts

Posted 25 January 2022 - 06:32 AM

Conclusions
Clearly the game was a complete mismatch as the red team had an average percentile of 90%, whereas the blue team had an average percentile of 70.75% - which is a huge difference. Not only that but the red team had at least one lance dropping as a group and on comms.

Not knowing the player's tiers, as unfathomably that information is not publicly available (unless you post to the forum and enable the setting to show your tier), I would also assume that the vast majority of players in the red team were tier 1, whereas the vast majority of players on the blue team were probably around tier 3.

The crucial point is that it would have been easy for the match-maker to make the game fairer. By swapping both team's charlie lances the average percentiles for each team would then have been 82% and 78%, for the red and blue team respectively. Further optimisation may have been possible.

From this single example I therefore rest my case. Assuming the null hypothesis to be true, in that the Jarl list's average percentile is representative of a player's skill rating, the match-maker is clearly broken. No further examples need to be discussed as a single example proving to the contrary refutes the assertion that the matchmaker is fit for purpose. Reductio ad absurdum.

PGI it is over to you now to answer my request for information as to how exactly the match-maker works?
We are all eager to know if you are going to fix it?

Edited by Titan Prometheus, 26 January 2022 - 08:00 AM.


#39 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 25 January 2022 - 06:54 AM

Was fun to drop. I'm glad you've taken the time to have a think. Legit.

One has to remember that game - out of the 100,000 or so a month, the overall etc. How common is it?


Additionally I do believe we had waited quite some time for a match (5mins+ maybe more). So the gates are wide open I would imagine.

No one is saying MM is perfect and it has secondary balance issues among others that absolutely need to be addressed (hopefully). It is certainly on PGI radar, I can say at least that much.

#40 P H O T O N

    Member

  • PipPipPipPipPip
  • Wild Dog
  • Wild Dog
  • 120 posts

Posted 25 January 2022 - 06:57 AM

View Postjustcallme A S H, on 25 January 2022 - 06:54 AM, said:

Was fun to drop. I'm glad you've taken the time to have a think. Legit.

One has to remember that game - out of the 100,000 or so a month, the overall etc. How common is it?


Additionally I do believe we had waited quite some time for a match (5mins+ maybe more). So the gates are wide open I would imagine.

No one is saying MM is perfect and it has secondary balance issues among others that absolutely need to be addressed (hopefully). It is certainly on PGI radar, I can say at least that much.


Indeed, it was nice to be involved in a few games with you ASH and finally put a voice to the forum name. I can see you are a very good player and for that you have my respect. You also appear to be a reasonable (if a little antagonistic!) chap which means more to me in this instance!

Yes there was a reasonable wait (maybe a minute for me) but the crucial point is that game could easily have been made more fair once all the players were assembled. This is not a rare occurrence, it happens over and over again. After all is said and done, we will all improve as players if the games are fairer, and we will all have more fun to boot.

Edited by Titan Prometheus, 26 January 2022 - 08:30 AM.






3 user(s) are reading this topic

0 members, 3 guests, 0 anonymous users