14 replies to this topic

[Erasmos]

I just had this round, in my Huginn...



Not a spectacular round "on paper," but when you consider the team I was working with...

It was one of those games where all three assault mechs are clustered behind the same building trying to hide, while a Cicada scratches at them with with 2ERLL. Despite numerous attempts to warn the team of incoming enemies, (as a good Huginn should,) and despite numerous warnings by other teammates that we needed help from the fatties (who were still hiding,) we were predictably rolled and destroyed.

No biggie. The other team needs to have some fun too.

But here's my issue: I busted my ass for the team, and ended up contributing more than most of them combined. My final damage was even higher than all but 1 on my team, and with the game stats shown above, it's pretty obvious that I was contributing a lot more than most on the team.

But simply because it was a "loss," my pilot rating went down.

If anything, when one or two mechs have stats that TOWER above the rest of their PUG team, I think the rating should go up even more. There have been many losing armies who's heros gained fame and respect due to THEIR conduct, (despite the loss.)

Just sayin...


Anyway, I love the rest of the patch, thanks!

-CT

[Binary4]

Until now I only got one loss with a neutral change (with about 290 match score). I am also wondering why won matches with less than 150 ms give a positive change. IIRC PGI announced different ranges for the change, not in numbers but the current ones are not what I expected when i read them.

[ShinVector]

PSR is highly tied to match score... How did you do compared to the rest of the team ?

Possibly you lost a match you were suppose to win ?

Edited by ShinVector, 24 September 2015 - 08:23 AM.

[Nik Reaper]

I think it was tested that you need to be in the 600-ds damage to make sure you don't lose rating even if your team gets rolled, also it probably has to do with your personal average , so is that match score a lot lower than your average?

[Erasmos]

ShinVector, on 24 September 2015 - 08:23 AM, said:

PSR is highly tied to match score... How did you do compared to the rest of the team ?

Possibly you lost a match you were suppose to win ?

That's just it: my match score was best on team, and dameage was 2nd best. (Higher than our assault lance combined, actually.)

Haha, it was one of THOSE games.

[ScorpionNinja]

PSR or ELO, either way everthing is tied to WINS & LOSSES!!!

Too bad pilot "skill" wasnt the ONLY factor, to separate the CADETs, GREEN/REGULAR/VETERAN/ELITE Pilots... like including to that rating a pilot's "TO HIT % with weapons", how many MATCHES you have completed, etc is by far more "Skill" related than dmg/assists/wins/losses and the crap formula PGI uses!

[ShinVector]

Colonel Tonberry, on 24 September 2015 - 07:38 PM, said:

Haha, it was one of THOSE games.

Where as ideally PSR up/down rating should be dynamic.
I get the feeling it is fixed values.. Maybe something like
300-450 = No change on lost.
>450 = increase a little bit on lost.

Need more data from people... Prof. Dakka's latest post has given some insight into the matter.

Edited by ShinVector, 24 September 2015 - 09:49 PM.

[rolly]

I'm scratching my head at this too. I know I'm not a great player by any stretch, but I'm not horrible either and do my part AND even go so far as to help/cover other random PUGs in a fight.

Yet one of my clanmates who plays less than I do, dies more often, and generally only runs lights is a HIGHER tier than I. (Tier 4 [3/4th of bar])
On average he has less than 100 damage per match with the occasional kill, with rare moments breaking 200. With a total of 150 matches on his account. He almost exclusively runs a Panthers and some light LRM boats.
I average 1-2 kills, 4-6 assists with over 500 games under my belt as opposed to his. If I don't get kills I average 6-10 assists.
Average damage is 180-250. (Tier 5 [14/15th of the bar]) I run all weight classes, generally as a Assault guard, flanker, striker, indirect/direct fire support mechs.

What bothers me is that this score does not take into account that I spend entire matches voluntarily covering assault lances with ECM and triple AMS (Kit fox-C) TAG all match, cover stray/slow assaults joining the herd at the start, risk my neck to pop a UAV, or spend time being the only team mate typing out enemy movements and ID. But apparently I'm not being enough of a "team player" according to the PSR definition.

To me this PSR thing is really just another distraction. I already knew if my game play was improving or worsening, now its just arbitrarily rating it again. Its unnecessary and seems to serve those people who need a karate belt colour to know how good they are.

[yeah slightly miffed, but would love constructive feedback to explain how the above example works]

Edited by rolly, 24 September 2015 - 10:13 PM.

[no one]

Colonel Tonberry, on 23 September 2015 - 09:14 PM, said:

Not a spectacular round "on paper," but when you consider the team I was working with...

It was one of those games where all three assault mechs are clustered behind the same building trying to hide, while a Cicada scratches at them with with 2ERLL.

. . .

But here's my issue: I busted my ass for the team, and ended up contributing more than most of them combined. My final damage was even higher than all but 1 on my team, and with the game stats shown above, it's pretty obvious that I was contributing a lot more than most on the team.

There's your issue, really. People can play like dumb cowards and do better than a good, coordinated aggressor* on an individual level. Assaults that do things like hide behind a building all match? They frequently manage to take down two or three people after their team's been exhausted. That's not because they're a fantastic player, but because they're fresh, and facing 5 damaged 'Mechs, not the concentrated 12 man firing lines brawls start with.

There are a lot of cases like that where a player does well for themselves despite leaving their team to hang. You have to be a decisively winning influence to overcome such things, or just kind of hope the magic of averages eventually figures out that you're in fewer losing games than captain shywhale.

*I may be too aggressive, but I do try to support the engaged portion of my team.

[DeathlyEyes]

The bigger issue is, we're pre psr scores augmented. If you were stuck in an Elo range where when you played solo matches and the games were basically unwinnable, (your team is low Elo with you boosting the average against an above average elo team) do those games affect your psr differently?

[Commissar Aku]

It could be worse:

It could be worse, I got a match score of 5 on a win and my PSR went up.

Edited by Commissar Aku, 25 September 2015 - 05:52 AM.

[amenophis]

A team win or a team loss should not have any effect on a pilot rating. Pilot rating is an individual score where a team win or team loss is a multimember team score. One should not have anything to do with the other. If it is grouped together(which it seems to be) this makes the pilot tier rating a load of CRAP.
What this seems to mean is a lot of very good pilots will never advance because there win loss record is not very good. They get dropped with more losing groups than winning groups through no fault of their own.

[Random Wanderer]

To put this simply: I have gotten damage ratings of over 600 and match scores in the 400-500 range, and had my PSR go down on a match where my team lost; and conversely had a damage rating of 100, a match score of hell-if-I-know-but-it-can't-have-been-very-good, and my team has won, and my PSR has gone up.

That sort of thing makes it very obvious that my PSR is not rating me based on my individual abilities, but instead on my team's success or failure. Given that, as I understood it anyway, the entire purpose of PSR was to rate your individual abilities rather than the ELO which dragged you down due to things outside of your control, this seems like a failure of the system.

[Elbrun]

A purely skill based system would never take W/L into account. It would take into account damage taken (possibly, as a more skilled pilot should take less damage overall), K/D ratio, and weapon hit percentages (projectiles fired/projectiles hit) (which would be a bit random to LRM/SRM users, but that's kind of a missile system balance for all missile systems). As well ECM/TAG/AMS usage should somehow be added in to the determination of Pilot Skill Rating, they are key items in this game which can be a determining value of a win or loss (just try to get hits with LRMs in game against a team with 4-5 ecm loaded mechs...). Inclusion of W/L into the mix at all unfairly removes skill in the determination of PSR as it unfairly weighs against actual individual player skill.

[Anachronda]

Yes. At first I thought and some others thought that the numbers for poor/ok/well/extremely well were somehow dynamic based on the scores of the rest of the team. However that seems not to be the case. Apparently even when you score in the top 3 you will lose PSR if your team lost. What seems to be happening is that the numbers required are much higher on a loss than on a win. For instance even when I am scrub bottom with 100 points, I gain PSR when we win. I lose PSR when we lose even on 600+ damage games.

For more details and analysis see this thread and this one. There are, BTW, way too many threads on this issue. I blame the poor search tools in the forum. But it would really help if we tried to stick to one thread for this.

Edited by Anachronda, 27 September 2015 - 03:46 PM.