46 replies to this topic

[Almond Brown]

Kira_Onime, on 23 September 2015 - 09:25 AM, said:

It's an equal sign.... you know, this -> =
Means your PSR stayed equal to what it was previously.

FTFY

This performance on a loss made the PSR go up.

Out of curiosity, do you remember the final Team kill total? Was it a 12-10(11) Loss? I suspThat was a fine Match though...

Edited by Almond Brown, 23 September 2015 - 09:34 AM.

[nehebkau]

RePlayBoy101, on 23 September 2015 - 09:05 AM, said:

how much till i hit t1?

El Bandito, on 23 September 2015 - 09:06 AM, said:

How many matches did you play in total starting from January?

RePlayBoy101, on 23 September 2015 - 09:08 AM, said:

400-500

Liar. Paul, himself, said that in the super-duper best optimal you never lost a game and got deathstar 100 times you could possibly expect tier 2 in 1000 games. More likely, if you were a decent w/l expect 2500 games -- NOT 500

From here

<begin e-peen shrink>

Edited by nehebkau, 23 September 2015 - 09:37 AM.

[RePlayBoy101]

Bilbo, on 23 September 2015 - 09:28 AM, said:

People keep saying this, and I'm not entirely sure it's true. W/L 3,814 / 3,486 and I'm at the high end of tier 4. Of course I have no idea how much of that is since January.

its not just the win to lose ratio its the match score and the dmg and idk what else ...not just the W/L

and your w/l is awerage really ... try harder and ull get there

[Kira Onime]

Almond Brown, on 23 September 2015 - 09:33 AM, said:

Out of curiosity, do you remember the final Team kill total? Was it a 12-10(11) Loss? I suspThat was a fine Match though...

I have the screenshot recorded. Final result was 5-12

[El Bandito]

Bilbo, on 23 September 2015 - 09:28 AM, said:

People keep saying this, and I'm not entirely sure it's true. W/L 3,814 / 3,486 and I'm at the high end of tier 4. Of course I have no idea how much of that is since January.

What is your average C-Bill and XP per match? Should be on your stats page.

[WildeKarde]

Well I've only 1300 matches in total (and likely less than a quarter of that this year) with a positive W/L but I showed up lower quarter of T4. Seems to be number of matches is also a factor in your tier not simply a good score, seems most of the higher tiers have a lot more games played this year along with the scores.

[Bilbo]

El Bandito, on 23 September 2015 - 09:40 AM, said:

What is your average C-Bill and XP per match? Should be on your stats page.

Accumulative C-Bills Per Match 117,417.91
Avg. XP Per Match 950.18

[Sarlic]

nehebkau, on 23 September 2015 - 09:36 AM, said:

Liar. Paul, himself, said that in the super-duper best optimal you never lost a game and got deathstar 100 times you could possibly expect tier 2 in 1000 games. More likely, if you were a decent w/l expect 2500 games -- NOT 500

From here

<begin e-peen shrink>

I like the comment of primalchrome:

Quote

2500 games to get an accurate ranking? Are you kidding me? You should have an initial seed after 10-20 and be firmly ranked Tier 3-5 after 50 with a steady climb/fall thereafter. A player of "Tier 1" skill should have achieved that tier in under 300 matches. To expect an account to put in 400+ HOURS of playtime in your game to establish themselves in a sub top tier is asinine and myopic in the extreme. Does PGI honestly think that their game is the only one on a player's hard drive?

[Lugh]

Sarlic, on 23 September 2015 - 09:43 AM, said:

I like the comment of primalchrome:

That's a totally crazy amount of time. I now understand why I only have tier 3 though still about 800 matches shy of the magic 2500. Crazy. just wow...

[Ghogiel]

1 night of farming mine moved this much

[nehebkau]

Sarlic, on 23 September 2015 - 09:43 AM, said:

I like the comment of primalchrome:

I kind of agree with his sentiment.

[WarHippy]

I don't think my bar moves, or if it does it is so small that it is almost completely unnoticeable. What I am curious of is the exact amount of points I am going up or down compared to my performance since a colored arrow is just about as meaningless as you can get without being completely irrelevant. Without knowing the exact amount of movement that is occurring during gains or losses of rating it makes winning or losing the match feel a great deal more important than it is supposed to be when measuring your individual skill as this system is supposed to be doing.

[sycocys]

To me it seems like its tied to your teammates performance more than the W/L as weird as that sounds. I've been leveling 3 identical Crows from scratch and have quite consistent scores with them.

The losses where my team was terrible - I lose PSR
The losses where my team was ok - I equal
team was good - equal or plus (still similar consistent scores)

Wins tend to go up more than down, but I see many equal ones as well to my eyes it just seems like the rest of the team affects the chances of seeing an up rather than an equal. Too many low performers and it doesn't go up, everyone pitches in and I get an increase.

Just don't know how to read into this system yet, it doesn't feel like it has any reasoning behind it when the same performances result in every possible result and sometimes better/worse performances result in the opposite result you'd expect.

[El Bandito]

Bilbo, on 23 September 2015 - 09:41 AM, said:

Accumulative C-Bills Per Match 117,417.91
Avg. XP Per Match 950.18

117,417 C-Bills is OK, but 950 XP is not. Do you often use Hero mechs? To climb the tiers, you need to contribute in the match noticeably--which is reflected in the XP earning. Higher tier players tend to have 1100 XP per average match.

Edited by El Bandito, 23 September 2015 - 10:10 AM.

[Bilbo]

El Bandito, on 23 September 2015 - 10:08 AM, said:

117,417 C-Bills is OK, but 950 XP is not. Do you often use Hero mechs? To climb the tiers, you need to contribute in the match noticeably--which is reflected in the XP earning.

No, I rarely ever use heroes or champions.

[Davers]

El Bandito, on 23 September 2015 - 10:08 AM, said:

117,417 C-Bills is OK, but 950 XP is not. Do you often use Hero mechs? To climb the tiers, you need to contribute in the match noticeably--which is reflected in the XP earning. Higher tier players tend to have 1100 XP per average match.

I didn't think your stats would include Cbills or XP from Hero/Champion mechs?

[WarHippy]

El Bandito, on 23 September 2015 - 10:08 AM, said:

117,417 C-Bills is OK, but 950 XP is not. Do you often use Hero mechs? To climb the tiers, you need to contribute in the match noticeably--which is reflected in the XP earning. Higher tier players tend to have 1100 XP per average match.

I average 1603.3 xp per match and I am a lower tier than you so I don't think that plays all that much of a role if the higher tier players tend to average 1100xp. It really seems to boil down to number of games played and w/l.

[sycocys]

Average xp per match 1373.9 -- still in T4, so that's obviously not the greatest gauge.

If anything it's tied to w/l - which I don't know how to break out of without
A- running only top tier mechs and getting lucky with MM
B- only dropping with my group again combined with A

The solo queue is still such a roll of the dice on which team wins, and it really doesn't seem like individual pilot performance is really rated very well even in this system.

Edited by sycocys, 23 September 2015 - 10:24 AM.

[Chuck Jager]

Nightmare1, on 23 September 2015 - 09:21 AM, said:

I'm disappointed. I thought PSR was supposed to move up if you did well on a loss, and move down if you did poorly on a win. Instead, it seems little different from ELO since it's tied largely to your W/L.

It does do that. You just do not see it on the bar.
A middle of the road math guess would be that if your game score is 500 (a great game) and you get 10 pts to psr for every 25 above 100 that would be +160 and they give you +25 for the win, that looks like a good +185 to PSR.

Well this is kinda flawed because anybody who can breathe would keep advancing slowly and surely. PGI needs your average performance so they now average this amongst all your games in a period (approx 2000) so +185 becomes 185/2000=.0925pts added to your PSR. even more assume psr has 100pts per tier (could be 1000 but will still be similar result).

Determine modifier to PSR from game score:
X = ((game score-100)/25)*10)
Determine how modifier affects your game average and advancement to next tier"
(X/number of games)/points per tier = .000925% advancement to next tier

(((500-100)/25)+25)/2000)/100 = .000925%

Note: even if my guesses are off as to how the modifiers are factored. I am still demonstrating why moving from t2 to t1 in a string of great winning games will advance you less than .5% of the way

couple of things to learn

• The fewer games you have played the quicker you can move up or down
  
• PSR is set up to determine your ability for MM and it needs to look at your average play. So somebody moving into T1 from T2 has to have played more games so the progression will become slower as you move up
  
• Many people love sports but very few can play cut it at the college level and even fewer can do it pro. This is what PSR is designed to separate those levels of skills for the MM. It is not a title to be earned by learning a tutorial off of Utube.

[Bilbo]

RePlayBoy101, on 23 September 2015 - 09:37 AM, said:

its not just the win to lose ratio its the match score and the dmg and idk what else ...not just the W/L

and your w/l is awerage really ... try harder and ull get there

I'm not actually concerned with what my tier is. I'm just looking to find out how the system actually works.