Jump to content

- - - - -

The Player Skill Rating (Psr) System Explained... (As Best I Can)

Guide

149 replies to this topic

#61 L Y N X

    Member

  • PipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 629 posts
  • LocationStrana Mechty

Posted 12 October 2015 - 12:16 PM

View PostVoid Angel, on 11 October 2015 - 11:24 PM, said:

You mean except for the contrary data that you are hand-waving away? It's not up to others to prove you wrong - the burden of proof is on you, and you're claiming more than your data supports. Sorry about your loss - but "conviction" in the face of evidence is called "stubbornness." You've got three separate people on this page alone telling you that they've had results that don't fit within your hypothetical framework - and you've basically just told them, "that's nice, but until I see a screen shot, I'm going to assume my data set is inclusive." That's not reasonable; particularly when the part of your hypothesis being questioned is demonstrably wrong.

You might have identified average breakpoints based on typical behaviors in the observed tiers, and that's great - I applaud your effort, and your intent. However, your intransigent insistence on things you haven't proven embarrasses you, and promotes disinformation in a forum dedicated to helping new players.



When I was in school, my english teacher would chide us for starting a sentence with "I think..." She said it is obvious this position is your thought and opinion, you wrote it, you do not need to add the extra words "I think..."

There are contrary opinions and those are fine Void, you have them as well. But there has yet to be posted hard evidence of contrary data that would prove me wrong. I am not speaking about a border value of a point of match score but of anything you have linked, Everything that PGI has said about PSR is consistent with what they have said. Some people are of the opinion that PGI has contradicted themselves. I'm sorry, but those folks are reading contradictions where none exist.

Again, I have shared repeatable results that have been verified as solid by many in my unit and several others in this thread. It is within your right to disagree, but until you can post data, it is not a fact. My facts have been repeatedly and independently verified. I do not have a burden to prove anything to any one. I simply shared my findings.

I am not embarrassed despite your own assertions and name-calling pettiness.

Edited by 7ynx, 12 October 2015 - 12:18 PM.


#62 Not A Real RAbbi

    Member

  • PipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 1,688 posts
  • LocationDeath to Aladeen Cafe

Posted 12 October 2015 - 12:26 PM

Question: Is it possible that solo and group queue scores are weighted differently?

For instance, my observations on break points...

Loss: < 250, PSR v. 250 to < 400, PSR =. > 400, PSR ^.
Win: < 100, PSR =. >= 100, PSR ^.

No idea how low a score one must have in a win to go down, and no idea what the breaks are between small and large movements in PSR.

That said, that's ALL been in solo queue. (And yes, I DO have craploads of screen shots to support the above.) I wouldn't have any idea if the same would hold true for GROUP queue match scores.

So again, is there any chance that the weights are different for the two queues, and that THAT might account for the discrepancies?

#63 L Y N X

    Member

  • PipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 629 posts
  • LocationStrana Mechty

Posted 12 October 2015 - 02:57 PM

Rabbi,

Yes same holds for Group Queue. Thank you your findings support my own.

#64 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 7,063 posts
  • LocationParanoiaville

Posted 13 October 2015 - 10:21 AM

You seem to be under the impression that your anecdotal opinion has been verified by some kind of science. You are wrong in this, as you are wrong in your dogmatic insistence that your non-representative data set has given you authoritative results. Allow me to be clear: in order to have a "verified" result, you cannot skip a step in the scientific process - the first step, as it turns out. Since you have literally not done the first thing you need to do in order to scientifically verify your findings, you can't reasonably say, "well, if you think I'm wrong, the burden of proof is on you." I've given you the proof you keep on "asking for," and you ignored it. This should not surprise me - you quoted the link in the same post in which you called me "petty" while accusing me of name-calling in the same breath.

Your hypothesis - that there are static break points in observed match score that will result in predictable changes in PSR based on the match result - is wrong. We know it is wrong because it makes incorrect assumptions which I've already pointed out to you - and unlike yours, my position is based on the proper research. It doesn't matter that you see correlation in your personal data; it doesn't matter how many of your unit buddies agree with you; it doesn't matter whether a couple of other folks in this thread agree with you - there is no way to reconcile your claims with the known mechanisms of PSR.

You say PGI hasn't contradicted itself, but that would be the only way your hypothesis could hold up: PGI says that the individual components of your observed match score are weighted by multipliers. This would mean that even if you had a statistically significant data set, you might just be observing the results of common player behaviors in match - unless they got their wires crossed and posted misinformation.

If you were saying, "these are the break points where you'll see your PSR change as a general rule," I wouldn't have a problem. That conclusion fits all the data. But you're not saying that - your dismissal of anomalous data makes that clear. What you're saying is that your opinion has been reliably proven by (your) data, and that any pesky bits of reality that don't concur had better take screen shots!Precisely how you can demand a level of verification that you haven't presented yourself is an interesting question, but one that does not need an answer for the reasons I described above. The bottom line is that you don't have the proof you think you have - and that even if you did, it's not the proof you'd need to prove your claim.

Edited by Void Angel, 13 October 2015 - 10:24 AM.


#65 Not A Real RAbbi

    Member

  • PipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 1,688 posts
  • LocationDeath to Aladeen Cafe

Posted 13 October 2015 - 10:43 AM

Let me just be clear, dear Hatfields and McCoys of the MWO forums, that I am NOT taking a side here. I don't want to get caught up in the whole "No, YOU'RE a big doodoo head!" thing.

I was not specifically aware of any statement by any PGI personnel of some multipliers for various match score components, that would affect how that match score weighs in PSR movement. I'm sure there's a link in one of the previous LENGTHY (and let me just come out and say it, BO-RING) posts in this discussion-turned-pissing-contest. Sorry. Whole thread is TL;DR.

Here's an idea. Let's all go observe again. Let's all commit to screen capping at least 100 matches' summaries, and all in either solo or group (but not both) queues. AND, the kicker, let's screencap NOT ONLY the match score summary that shows all 24 players, but also the next page (which shows a breakdown of the components of the match score).

THEN, let's bring those observations back here to discuss. LIKE CIVILIZED PEOPLE, mind you.

And for the love of Kerensky, can we STOP with the poo flinging? At least until then? PURTY PLEASE?!

#66 L Y N X

    Member

  • PipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 629 posts
  • LocationStrana Mechty

Posted 13 October 2015 - 12:33 PM

I agree Rabbi, and look forward to seeing what Void, yourself and others come up with...


The TL;DR version as I see it...
PGI basically said the scoring formula for PSR will not be released and kept hidden ( I think for good cause, as it allows them to tune it without being accused of lying or anything else nefarious)

Then Paul says the PSR changes are a result of match score. And as most us us know the new match score formula's are hidden and not released. Therefore it is not contradicting, unlike some who claim PGI is contradicting themselves...

#67 L Y N X

    Member

  • PipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 629 posts
  • LocationStrana Mechty

Posted 13 October 2015 - 12:39 PM

View PostVoid Angel, on 13 October 2015 - 10:21 AM, said:

Snip...

You say PGI hasn't contradicted itself, but that would be the only way your hypothesis could hold up: PGI says that the individual components of your observed match score are weighted by multipliers.

Snip...



Bingo!!! We have a winner folks... Void this is where our opinions overlap buddy! I agree with you on that statement! Thanks for that!

#68 Not A Real RAbbi

    Member

  • PipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 1,688 posts
  • LocationDeath to Aladeen Cafe

Posted 13 October 2015 - 12:45 PM

So, we're fairly certain then, that the score number we see in the match summary is NOT the number used to make that tier decision? No two 250 scores are alike?

Interesting...

#69 L Y N X

    Member

  • PipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 629 posts
  • LocationStrana Mechty

Posted 13 October 2015 - 01:00 PM

View PostTheRAbbi, on 13 October 2015 - 12:45 PM, said:

So, we're fairly certain then, that the score number we see in the match summary is NOT the number used to make that tier decision? No two 250 scores are alike?

Interesting...

I'm not sure that I understand your first question...

but the second... I wouldn't go so extreme as to suggest that all match scores are individually unique, but there is a very good chance that any two match scores of the same value, reached it with a varying mix of parameters... yes.

#70 Not A Real RAbbi

    Member

  • PipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 1,688 posts
  • LocationDeath to Aladeen Cafe

Posted 13 October 2015 - 01:36 PM

View Post7ynx, on 13 October 2015 - 01:00 PM, said:

I'm not sure that I understand your first question...

but the second... I wouldn't go so extreme as to suggest that all match scores are individually unique, but there is a very good chance that any two match scores of the same value, reached it with a varying mix of parameters... yes.


Right. And if the values of those parameters are weighted different for consideration in tier placement, than they are for the match score itself, then the first question is answered. That is, it's not THAT match score number itself that's being considered, but some other value based on the same parameters as the match score displayed to players in the match summary.

THUS, the players can see a breakdown of their match scores, as well as the cumulative score itself, but the match score that determines their tier movement/placement (whether you move up or down, a lot or a little, or not at all) is still kept secret. It's based on the same considerations/parameters, but those are weighted differently in the secret formula (11 delicious herbs and spices? Ancient Chinese secret?).

#71 L Y N X

    Member

  • PipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 629 posts
  • LocationStrana Mechty

Posted 13 October 2015 - 02:24 PM

View PostTheRAbbi, on 13 October 2015 - 01:36 PM, said:

Right. And if the values of those parameters are weighted different for consideration in tier placement, than they are for the match score itself, then the first question is answered. That is, it's not THAT match score number itself that's being considered, but some other value based on the same parameters as the match score displayed to players in the match summary.

THUS, the players can see a breakdown of their match scores, as well as the cumulative score itself, but the match score that determines their tier movement/placement (whether you move up or down, a lot or a little, or not at all) is still kept secret. It's based on the same considerations/parameters, but those are weighted differently in the secret formula (11 delicious herbs and spices? Ancient Chinese secret?).



It is possible, but ask yourself, why would PGI go through that extra effort? If they have one match score that the formula to get there is secret, why create a second formula that is secret? Why not just alter the magnitude of PSR increase/decrease to tune it? Anyhow, this is all speculation...on both parts... What do you see? when does your match score go up? down? sideways? This is what we can measure. I look forward to your results Rabbi!

Edited by 7ynx, 13 October 2015 - 02:36 PM.


#72 Not A Real RAbbi

    Member

  • PipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 1,688 posts
  • LocationDeath to Aladeen Cafe

Posted 13 October 2015 - 02:37 PM

To be fair, I've gotten MOST of those match scores by playing a certain role in a certain kind of mech. I'd say that 15% or less of the scores I based that on, have been in mechs other than the ACH or SHC, and in roles other than reconnaissance. So if there WERE some difference in how certain items were weighted, I might not have noticed for sake of not having enough diversity in play style to see it.

For instance, the "BRAWLING" points? I rarely see those. I couldn't tell you what one has to do in a match to gain points for "BRAWLING" per PGI. I'm getting a grip on "FLANKING" and "SCOUTING". There could be other points-worthy accomplishments out there, of which I am unaware due to a relatively low average score on heavier mechs, and the relatively little that I play them. (I'm working on THAT, too, lately.)

But anyhow, I'll get back to tracking match scores, and also tracking their breakdowns on the Match Summary page, and see what, if anything, that tells me. And of course, I'll share with y'all, because DAMN. I've run into the both of you in game, and IIRC, you were both pretty good at this, and you were both valuable to your teams, and mostly I'm a fan of folks getting along well with one another...

#73 L Y N X

    Member

  • PipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 629 posts
  • LocationStrana Mechty

Posted 13 October 2015 - 06:03 PM

PLease note: I do not think that the break down points add up to your match score... It seems to be about 50% of damage dealt plus modifiers times assists and again for kills and etc for all those other pie slices... Damage Dealt seems to be the largest pie slice to contribute to match score and as far as it seems to me Assists next... beyond that I don't have a feel for it. I play all weight classes and all roles. As far as match score goes, I seem to do the best in Heavies as they seem to be a nice blend of firepower, armor and maneuverability. When I play lights I tend to back stabbed enemies and those kills are low damage kills leading to lower match scores for me. I have seen others get 1K+ games in lights.

#74 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 7,063 posts
  • LocationParanoiaville

Posted 13 October 2015 - 06:43 PM

View Post7ynx, on 13 October 2015 - 02:24 PM, said:

It is possible, but ask yourself, why would PGI go through that extra effort? If they have one match score that the formula to get there is secret, why create a second formula that is secret? Why not just alter the magnitude of PSR increase/decrease to tune it? Anyhow, this is all speculation...on both parts... What do you see? when does your match score go up? down? sideways? This is what we can measure. I look forward to your results Rabbi!

We know PGI is weighting the match score elements we can see with a hidden formula because PGI has told us that they are weighting the match score elements we can see with a hidden formula. This is not something they could reasonably be mistaken about - so unless you can demonstrate the claim that PGI lied, all this "hey, ask yourself" conspiracy theorizing is nothing more than baloney.

PS: No, your (unreleased) data does not prove that point, for reasons already presented - and ignored by you - in this thread.

#75 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Marauder
  • The Marauder
  • 7,063 posts
  • LocationParanoiaville

Posted 13 October 2015 - 06:48 PM

View PostVoid Angel, on 13 October 2015 - 10:21 AM, said:

You say PGI hasn't contradicted itself, but that would be the only way your hypothesis could hold up: PGI says that the individual components of your observed match score are weighted by multipliers.

View Post7ynx, on 13 October 2015 - 12:39 PM, said:


Bingo!!! We have a winner folks... Void this is where our opinions overlap buddy! I agree with you on that statement! Thanks for that!

So you agree with me that your hypothesis is invalid? Great! I look forward to your corrected OP. Thank you.

#76 Not A Real RAbbi

    Member

  • PipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 1,688 posts
  • LocationDeath to Aladeen Cafe

Posted 13 October 2015 - 09:39 PM

Alright, you two! Get back in your respective corners, or get a room! <3

Seriously, though. I went ahead and started a VERY rudimentary spreadsheet with values for Score, K, A, DMG, all the score variables I've seen listed so far (through four Pug drops tonight in AHCs), a Sum for the previous, and some other interesting nonsense.

Why am I sharing this here? Look, I've come across the both of you on the virtual battlefield before, and you both impressed me there. FURTHER, though this discussion has kinda devolved, I feel that you BOTH want to offer better insights to the community at-large. FURTHER YET, both of you have been in here and the NPH forum helping newbies and half-asses (like me) to improve through solid advice. In short, I have an above-average respect for each of you.

So, hey, rather than tossing insults and non-friendly remarks at one another, how 'bout we work together on better understanding this PSR stuff?

If you want to do this thing, just message a brutha. Excel spreadsheet's on my Dropbox public folder (don't mind the old exams and lecture slides... used to teach). There's a master sheet, and I've already filled in tonight's four PUG drops on my sheet. Holla/PM, and I'll shoot you the link. OR, you could be a sweetie and host it on GoogleDocs. Whichever.

The 31st Century is a confusing, violent place. No need to make it any worse that it must be...

#77 L Y N X

    Member

  • PipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 629 posts
  • LocationStrana Mechty

Posted 13 October 2015 - 10:22 PM

Well Spoken Rabbi, I apologize for any non-friendly remarks. I have been careful to not hurl insults. Please do share what you find out, I'd be interested. As far as void is concerned, he is entitled to his opinions. The OP stands until shown not helpful or invalid. If he is not finding this info useful then don't use it. I know a bunch of folks who found this information very helpful.

#78 Yozzman

    Member

  • PipPipPipPipPipPip
  • The Ardent
  • The Ardent
  • 273 posts

Posted 13 October 2015 - 11:16 PM

If winning or losing has such an big inpact PSR, then it's not an SKILL rating. It's a winning rate...

IMHO, if winning is such an big factor it means the calculation PGI uses is just not elaborate enough to compensate.

#79 L Y N X

    Member

  • PipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 629 posts
  • LocationStrana Mechty

Posted 24 October 2015 - 08:31 PM

Any findings yet Rabbi? I have not found any match score that do not match up with my findings to date.

#80 Not A Real RAbbi

    Member

  • PipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 1,688 posts
  • LocationDeath to Aladeen Cafe

Posted 24 October 2015 - 08:43 PM

Nothing so far. No one's taken me up on the document, so I have only my own few observations to work with. Kind of a bummer, since I don't get a whole lot of playing time these days (working at a university this time of year, in transportation no less, means a LOT of overtime with all the football games, and then basketball soon).

Again, if you're interested in sharing some observations, let me know and I'll shoot you a link for the DropBox public folder document. Simple Excel spreadsheet; just fill in the appropriate numbers and we can start attempting to tease out the relevant data once there are sufficient observations.





6 user(s) are reading this topic

0 members, 6 guests, 0 anonymous users