Jump to content

Psr Update And Hold On Patch.


713 replies to this topic

#521 Jay Z

    Member

  • PipPipPipPipPipPip
  • Deadset Legend
  • Deadset Legend
  • 436 posts
  • Twitch: Link
  • LocationAustralia

Posted 13 June 2020 - 03:15 PM

Just an FYI. I am fully aware of the weaknesses of a [Single Match result] -> [PSR Shift] calculation type, in particular uncontrolled divergence.

However, as a professional engineer and physicist IRL, I must follow the scope guidlines set out by the project lead: Paul.

Would 1. Rolling Averages, 2. Global Stats, 3. pure WLR variables be good to use or incorporate in calcs? Sure, I would love to have access to those variables. However these were, as stated by Paul, OUTSIDE THE SCOPE.

The solution you see presented is simplified down for ease of implementation. As a professional, staying within the scope and ensuring ease of implementation is how I keep my clients happy and contracts rolling in.

#522 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 13 June 2020 - 03:23 PM

View PostJay Z, on 13 June 2020 - 03:15 PM, said:

The solution you see presented is simplified down for ease of implementation. As a professional, staying within the scope and ensuring ease of implementation is how I keep my clients happy and contracts rolling in.


I understand this of course, however it is different for a me. I work in medical R & D. My analysis proves the safety and effectiveness of products, which all of my clients want to see be successful. It is within my power to tweak the result, but I never do, because my clients' interests do not outweigh the interests of tens of thousands of patients who would receive ineffective or unsafe treatments.

I ask you, professional to professional, do you believe your cure to the PSR and MM will benefit the playerbase? Or is it, as you say, just to deliver a project?

#523 Xiphias

    Member

  • PipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 862 posts

Posted 13 June 2020 - 03:26 PM

View PostJay Z, on 13 June 2020 - 03:15 PM, said:

Just an FYI. I am fully aware of the weaknesses of a [Single Match result] -> [PSR Shift] calculation type, in particular uncontrolled divergence.

However, as a professional engineer and physicist IRL, I must follow the scope guidlines set out by the project lead: Paul.

Would 1. Rolling Averages, 2. Global Stats, 3. pure WLR variables be good to use or incorporate in calcs? Sure, I would love to have access to those variables. However these were, as stated by Paul, OUTSIDE THE SCOPE.

The solution you see presented is simplified down for ease of implementation. As a professional, staying within the scope and ensuring ease of implementation is how I keep my clients happy and contracts rolling in.

I appreciate your contributions and not trying to attack you/your approach in particular. Just missed the issues on my first pass (been awhile since I've had to deal with any significant statistics and I'm a bit rusty) and wanted to make it more clear for others who might not understand the issues that Nighbird has been bringing up (took me a minute to realize the problem and I have some background in stats).

I appreciate how much you've done with how little you've been given. That said, if PGI is serious about fixing the matchmaker I think they need to go back and redraw the scope, because otherwise they aren't going to fix the problem.

#524 Jay Z

    Member

  • PipPipPipPipPipPip
  • Deadset Legend
  • Deadset Legend
  • 436 posts
  • Twitch: Link
  • LocationAustralia

Posted 13 June 2020 - 03:30 PM

The best net benefit within the scope. Going outside the scope would simply write off the proposal and so we are stuck with the current system with way way more flaws. Also, no one is medically at risk here Posted Image

Edit: There is no cure without significant changes. However, a small yet tangible improvement is a good thing.

Edited by Jay Z, 13 June 2020 - 03:32 PM.


#525 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 13 June 2020 - 04:08 PM

deleted - sorry I said I'd stop so I'll stop

edit 2: I found the words I really wanted to say. If you truly believe the improvement from the current proposal is small, and that a significant improvement can be made with a rolling average MS or WLR. Then just say so, at the front of the proposal. What do you have to lose, and what's the upside? (I also made a post about how to do a rolling average without a database query, which is actually what Paul said is impossible.

Edited by Nightbird, 13 June 2020 - 05:33 PM.


#526 General Solo

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • 3,625 posts

Posted 13 June 2020 - 04:52 PM

I'm no maths whizz but here is one I stole earlier.

Posted Image
Simple, minimal effort that should work with yearly resets, no worries mate hey Fahad Posted Image
The gradual change point can be adjusted of course, like to the average matchscore perhaps.

Edited by OZHomerOZ, 13 June 2020 - 04:55 PM.


#527 OneTeamPlayer

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • 399 posts

Posted 13 June 2020 - 05:32 PM

Whatever happens is it possible to blanket vote no on an idea?

Like a flat no to any proposal that suggests that the team means nothing and everyone should be competing against each other no matter the grouping.

Not to be rude, but i am 75% certain that if a system in which players are directly competing with their own team in addition to the other team for PSR this game is going to see a sharp rise in anti-team behavior and a sharp decline in population as people get fed up with watching teammates avoid damage, avoid helping strip armor, and waiting in cover until they can get crits, kmdd, and other behavior that will lead directly to a mass exodus from this game.

I'd imagine very few people want to play a team game where they are in direct competition with their own team- if we're going to be competing with the mech next to us, just remove the lances and blue diamonds, make everyone red, and make MWO a constant free-for-all.

#528 General Solo

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • 3,625 posts

Posted 13 June 2020 - 05:36 PM

Its called Pilot Skill Rating PSR.
Not Pilots Team Skill Rating
So yes pilots must compete with each other for PSR ranking

#529 Xiphias

    Member

  • PipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 862 posts

Posted 13 June 2020 - 06:01 PM

View PostOZHomerOZ, on 13 June 2020 - 04:52 PM, said:

I'm no maths whizz but here is one I stole earlier.

Posted Image
Simple, minimal effort that should work with yearly resets, no worries mate hey Fahad Posted Image
The gradual change point can be adjusted of course, like to the average matchscore perhaps.

Pretty sure it's going to have the same divergence problem as the other one that I showed in my simulation. After a given period of time players are going to bounce off the bottom/top.

View PostOZHomerOZ, on 13 June 2020 - 05:36 PM, said:

Its called Pilot Skill Rating PSR.
Not Pilots Team Skill Rating
So yes pilots must compete with each other for PSR ranking

And it's used for matchmaking. The goal is to win. A pilots skill is how good they are at wining. Everything else is secondary. You compete against your team being bad to be skilled enough to carry them to a win.

#530 OneTeamPlayer

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • 399 posts

Posted 13 June 2020 - 06:02 PM

View PostOZHomerOZ, on 13 June 2020 - 05:36 PM, said:

Its called Pilot Skill Rating PSR.
Not Pilots Team Skill Rating
So yes pilots must compete with each other for PSR ranking


I get that that's your opinion. My point still stands as written.

#531 Jochi Kondur

    Member

  • PipPipPip
  • Shredder
  • 66 posts

Posted 13 June 2020 - 06:25 PM

View PostXiphias, on 13 June 2020 - 03:26 PM, said:

I appreciate your contributions and not trying to attack you/your approach in particular. Just missed the issues on my first pass (been awhile since I've had to deal with any significant statistics and I'm a bit rusty) and wanted to make it more clear for others who might not understand the issues that Nighbird has been bringing up (took me a minute to realize the problem and I have some background in stats).

I appreciate how much you've done with how little you've been given. That said, if PGI is serious about fixing the matchmaker I think they need to go back and redraw the scope, because otherwise they aren't going to fix the problem.


Am I right in your assessment that in terms of the benefit to the player experience, Nightbird's solution >= Gagis' solution > Jay Z's solution. However, only Jay Z's solution fits within the scope of changes and therefore has the best chance of being implemented?

Edited by Jochi Kondur, 13 June 2020 - 06:50 PM.


#532 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 13 June 2020 - 06:48 PM

View PostJochi Kondur, on 13 June 2020 - 06:25 PM, said:

Am I right in your assessment that in terms of the benefit to the player experience, Nightbird's solution > Gagis' solution > Jay Z's solution. However, only Jay Z's solution fits within the scope of changes and therefore has the best chance of being implemented?


I do not have an accurate assessment of Gagis' solution. What I said is about his system is formed from my pre-conceptions, which is why I also keep saying that simulations are needed to prove them. It very well may be that his solution works better than mine, I just don't know.

#533 General Solo

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • 3,625 posts

Posted 13 June 2020 - 08:10 PM

View PostXiphias, on 13 June 2020 - 06:01 PM, said:

Pretty sure it's going to have the same divergence problem as the other one that I showed in my simulation. After a given period of time players are going to bounce off the bottom/top.

And it's used for matchmaking. The goal is to win. A pilots skill is how good they are at wining. Everything else is secondary. You compete against your team being bad to be skilled enough to carry them to a win.


I read that other contenders who have remained within the scope/limits have the same problem of divergence which imo is easily fixed with regular reset's like some other games do, apparently, I read that somewhere on the forums.

Correct me if I am wrong.

View PostXiphias, on 13 June 2020 - 06:01 PM, said:


And it's used for matchmaking. The goal is to win. A pilots skill is how good they are at wining. Everything else is secondary. You compete against your team being bad to be skilled enough to carry them to a win.



What or who's goal is to win, match maker, solo players or groups?

A solo pilot does not have the same chance at winning as a group pilot.

If you make winning a large part of the formula it will favour grouped pilots and disadvantage solo pilots in terms of PSR movement. Is this really measuring a pilots skill is what I am saying.
Edit: Even Handsome Proton mentioned that maintaining a positive win loss as a solo was hard work or some thing close to it. Correct me if I'm wrong beautiful? Posted Image

Groups don't need more advantages.


Match making is used to match pilots and groups with similarish skill to avoid hugh skillgaps between two teams to my knowledge.

Again correct me if I am wrong. amirite?

Edited by OZHomerOZ, 14 June 2020 - 06:00 AM.


#534 General Solo

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • 3,625 posts

Posted 13 June 2020 - 08:16 PM

View PostOneTeamPlayer, on 13 June 2020 - 06:02 PM, said:

I get that that's your opinion. My point still stands as written.


I know it is your point, however I disagree with it.

Thats why I posted my opinion.

Posted Image

#535 Gagis

    Member

  • PipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 1,731 posts

Posted 13 June 2020 - 09:21 PM

View PostJochi Kondur, on 13 June 2020 - 06:25 PM, said:

However, only Jay Z's solution fits within the scope of changes and therefore has the best chance of being implemented?

The original version of my proposal was definitely out of scope. Then I used lube and a mallet to kinda hopefully maybe make it fit. Only PGI knows if its too tight or not.

#536 Jochi Kondur

    Member

  • PipPipPip
  • Shredder
  • 66 posts

Posted 13 June 2020 - 10:04 PM

View PostGagis, on 13 June 2020 - 09:21 PM, said:

The original version of my proposal was definitely out of scope. Then I used lube and a mallet to kinda hopefully maybe make it fit. Only PGI knows if its too tight or not.


Well, Nightbird presented the solution on how to do a rolling average without a database query to Russ and Paul via a tweet. In that updated context, is your solution still out of scope?

#537 Dauntless Blint

    Member

  • PipPipPipPipPipPip
  • Fire
  • Fire
  • 408 posts
  • LocationPlaying other games.

Posted 13 June 2020 - 10:55 PM

View PostDar1ngOne, on 12 June 2020 - 06:53 PM, said:

This caught my eye, and made me log in to comment.

The issue is not the formula or the player base. It's in the balance of weapons and tech, and experience.

For example - If the game was locked to 1 mech, with the same tech base, same weapons, across the 24 players, then you would get a real sense of who knows what to do and the skill level involved.

So for instance with have 24 Commando's on the field, all the same, and I would expect the account named Mycrus to dominate most of the matches.

Then what happens is people remember that name as being a threat, so the other players start to focus Mycrus. Now Mycrus may make an alt account, or change his account name to keep the focus of him.

So then PGI introduces mech X and weapon X, and then suddenly, the meta changes, and a great player in a Commando is very average in a Hunchback.

Your punishing a players account based on the fact that he hasn't got the right mech, the right weapons, despite having the skill in another chassis?

If it were me, and I have said this to you in the past PGI, you would have the Tiers renamed as Elite, Veteran, Rookie, and Green.

You would have a mode where the mechs have to be the same builds, and then sort the best from the worst in that mode.

Currently, your trying to assess a player's worth on shifting sands, and if your building a system on that, it will fail.

Good luck with that. Thanks again for MW5.


This is why is why we need players to be bench marked against chassis global averages.
Because some mechs are more equal then others.

Player(W/L x MS x K/D) = PSR (how you go up and down tiers)
+
Global chassis average(W/L x MS x K/D) =Chassis PSR (chassis modifier)
=
Match effective PSR. to feed Match Maker.

Zero sum to infinity.

The scale should be to infinity and percentile based too I guess, the caps break the dispersion.
(Experience should be removed from the equation.

Sorry if we're trying to get you to write a red line with a blue pen.

Edited by Dauntless Blint, 13 June 2020 - 11:55 PM.


#538 Horseman

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Mercenary
  • The Mercenary
  • 4,738 posts
  • LocationPoland

Posted 14 June 2020 - 01:26 AM

View PostNightbird, on 13 June 2020 - 03:23 PM, said:

I ask you, professional to professional, do you believe your cure to the PSR and MM will benefit the playerbase? Or is it, as you say, just to deliver a project?
I would call it an improvement rather than a cure. Anything better than what we have now will benefit the playerbase, the only question is by how much.

View PostOZHomerOZ, on 13 June 2020 - 08:10 PM, said:

I read that other contenders who have remained within the scope/limits have the same problem of divergence which imo is easily fixed with regular reset's like some other games do, apparently, I read that somewhere on the forums.
Or, instead of repeatedly throwing matchmaking into disarray, decay the PSR ratings by one point towards global average each day.

Edited by Horseman, 14 June 2020 - 01:28 AM.


#539 General Solo

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • 3,625 posts

Posted 14 June 2020 - 02:03 AM

View PostHorseman, on 14 June 2020 - 01:26 AM, said:

I would call it an improvement rather than a cure. Anything better than what we have now will benefit the playerbase, the only question is by how much.
Or, instead of repeatedly throwing matchmaking into disarray, decay the PSR ratings by one point towards global average each day.


If its regular then its expected and thus less disarray happens imo.
Plus if people are having more enjoyable games maybe the resets will upset less.

Also the I lost moi Tier crowd seems to be less than I want better more meaning full (I had an effect on the outcome) games, which a good matchmaker that tried to contain manage better the skill gaps would provide.

imo

Edited by OZHomerOZ, 14 June 2020 - 02:45 AM.


#540 General Solo

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • 3,625 posts

Posted 14 June 2020 - 03:04 AM

My background is from a unit with a large skill gap profile, CWI, Heard comments from all skill levels in dat unit
Every one wants to be a star, regard less of actual skill, most imo dont care for or respect skill just the winning part.
Match maker magic may not make that happen but at least if MM is workin good they wont give up as soon as they see the lobby and the names of their opponents.
Posted Image Rhats my hope, yeah sure shank redeemtion code aka I will believe it when I see it, yes bit of a doubting Thomas, Yet I still have faith, kinda!

Edited by OZHomerOZ, 14 June 2020 - 03:12 AM.






4 user(s) are reading this topic

0 members, 4 guests, 0 anonymous users