47 replies to this topic

[rox5tar]

I know since it's inception, the pilot skill rating has been very... controversial. Recently though, I don't know if there have been changes or it's just been my bad luck but I've found that the "rating" is VERY inconsistent.

I've had matches where:

- I've solo capped the CTF point, (As in fully capped by myself) and while my damage was minimal I won us the game. Skill rating went down.

- Had a match with 542 Damage (Second most on my team), 1 Kill (KMDD) and 4 Assists. Match Score was 291 (Third highest on my team). My match score went down.

- Had a match where I dealt 691 Damage (second most on my team), 1 Kill and 5 Assists. Match score was 372 (Third best on my team). My match score stayed the same.

- (This one was a while ago so I don't remember all the exact details unlike above which I saved) Had a match as a backstabbing piranha. Got 3 kills (KMDD), but sub-200ish damage. Match score went down.

There have been matches where I've performed much worse than those matches, and yet still got a positive skill rating increase. In that case, what gives?! Why (according to match score) am I in the top 25% on my team but my skill goes down? What's the actual process for calculating pilot skill rating?

[Meep Meep]

Psr is relative to both teams combined performance. So even if you placed top 3 on your team if the other team had better overall stats you still wouldn't gain psr.

Damage done and other actions only matter in relation to everyone else. So you can get 600 damage and 3 kills one match and lose psr or get 150 damage and no kills or assists and get a green arrow the next.

[rox5tar]

Meep Meep, on 13 April 2024 - 06:34 PM, said:

Psr is relative to both teams combined performance. So even if you placed top 3 on your team if the other team had better overall stats you still wouldn't gain psr.

Damage done and other actions only matter in relation to everyone else. So you can get 600 damage and 3 kills one match and lose psr or get 150 damage and no kills or assists and get a green arrow the next.

That seems pretty silly. Not that I'm doubting you, but if the other team wins then clearly they are gonna do better as individuals. Even if you are the champion of your team, you will still perform poorly as an individual if your entire team isn't pulling their weight and can probably get weighed down because of it.

Edited by rox5tar, 13 April 2024 - 06:46 PM.

[Meep Meep]

rox5tar, on 13 April 2024 - 06:46 PM, said:

That seems pretty silly. Not that I'm doubting you, but if the other team wins then clearly they are gonna do better as individuals. Even if you are the champion of your team, you will still perform poorly as an individual if your entire team isn't pulling their weight and can probably get weighed down because of it.

You can get a green arrow on a loss. You just have to score higher than the aggregate of both teams. Just trying to say that psr is based off both teams performance so you are competing with everyone.

[LordNothing]

i think i get more green arrows on losses than wins.

[rox5tar]

Meep Meep, on 13 April 2024 - 07:00 PM, said:

You can get a green arrow on a loss. You just have to score higher than the aggregate of both teams. Just trying to say that psr is based off both teams performance so you are competing with everyone.

Just played a match, where I got a 217 Match Score, but got a positive skill rating increase. Based on the match score of the other team, I was in the bottom half of everyone's match score (Meaning I got a lower score than 12 of the players in the match). Was also a loss, if that matters.

[ambosen]

The entire PSR system is heavily skewed in favor of newer accounts with less overall games. Contrary to what some have claimed it has precisely **** all to do with overall team performance or averaging thereof. Numerous streamers over the years have literally demonstrated this on video how PSR is heavily skewed in favor of newer accounts and literally doesn't track specific things that could heavily influence matches at all or doesn't weigh ones that should have an effect enough (and that's even before you get into the very real difficulties of accurately tracking some of that stuff to begin with; for a pretty good example of that, a light on your team who spends the entire match essentially doing a combination of spotting, crowd control, and pulling enemy mechs into far more easily attacked posiitions; how Could you even accurately track that data to begin with?)

Inevitably the proposed solution to a lot of these issues a couple of very vocal people in this very forum come up with is inevitably some variation of "what can we do to make PSR even more of a joke by removing even more tracked variables and skewing the entire system even more in favor of low play total accounts?"

[martian]

ambosen, on 13 April 2024 - 09:52 PM, said:

The entire PSR system is heavily skewed in favor of newer accounts with less overall games. Contrary to what some have claimed it has precisely **** all to do with overall team performance or averaging thereof. Numerous streamers over the years have literally demonstrated this on video how PSR is heavily skewed in favor of newer accounts and literally doesn't track specific things that could heavily influence matches at all or doesn't weigh ones that should have an effect enough (and that's even before you get into the very real difficulties of accurately tracking some of that stuff to begin with; for a pretty good example of that, a light on your team who spends the entire match essentially doing a combination of spotting, crowd control, and pulling enemy mechs into far more easily attacked posiitions; how Could you even accurately track that data to begin with?)

Inevitably the proposed solution to a lot of these issues a couple of very vocal people in this very forum come up with is inevitably some variation of "what can we do to make PSR even more of a joke by removing even more tracked variables and skewing the entire system even more in favor of low play total accounts?"

Stop spamming your nonsense.

In this thread you wrote the same crap. You even boasted that it is "simple mathematical function" and that you can show "a clear statistical bias" etc.

When I challenged you to post your verifiable sources, numbers, formulae or your tables ... you disappeared from the thread.

So I am still waiting for your proofs to be posted.

[martian]

rox5tar, on 13 April 2024 - 06:25 PM, said:

I know since it's inception, the pilot skill rating has been very... controversial. Recently though, I don't know if there have been changes or it's just been my bad luck but I've found that the "rating" is VERY inconsistent.

Greetings!

There have been no significant changes in how PSR works for years. The last big change happened in mid-2020 when the entire PSR was reworked. Read this official postthis official link to understand how PSR works.

You can watch this fan video:

Now some explanation, if you do not mind:

rox5tar, on 13 April 2024 - 06:25 PM, said:

I've had matches where:

- I've solo capped the CTF point, (As in fully capped by myself) and while my damage was minimal I won us the game. Skill rating went down.

You mean you capped a base in Conquest?

rox5tar, on 13 April 2024 - 06:25 PM, said:

- Had a match with 542 Damage (Second most on my team), 1 Kill (KMDD) and 4 Assists. Match Score was 291 (Third highest on my team). My match score went down.

- Had a match where I dealt 691 Damage (second most on my team), 1 Kill and 5 Assists. Match score was 372 (Third best on my team). My match score stayed the same.

- (This one was a while ago so I don't remember all the exact details unlike above which I saved) Had a match as a backstabbing piranha. Got 3 kills (KMDD), but sub-200ish damage. Match score went down.

Okay.

So the next time, when you will feel that you have lost your PSR "unfairly", post both the Mission Summary Team Table and the Mission Summary Player Table, please.

The basic idea behind PSR is that your performace is compared with the performance of other (both friendly and enemy) players. Without those tables, there is not much to discuss.

But as a rule of thumb, usually only the best two players of the defeated team get the PSR rise. Being the third best player of the defeated team is often not enough.

rox5tar, on 13 April 2024 - 06:25 PM, said:

There have been matches where I've performed much worse than those matches, and yet still got a positive skill rating increase. In that case, what gives?! Why (according to match score) am I in the top 25% on my team but my skill goes down? What's the actual process for calculating pilot skill rating?

Being the third best player of the defeated team or being in the top 25% of the defeated team is often not enough. I am sorry.

Edited by martian, 14 April 2024 - 06:00 AM.

[martian]

rox5tar, on 13 April 2024 - 06:46 PM, said:

That seems pretty silly. Not that I'm doubting you, but if the other team wins then clearly they are gonna do better as individuals. Even if you are the champion of your team, you will still perform poorly as an individual if your entire team isn't pulling their weight and can probably get weighed down because of it.

Even if you are on the defeated team, you can still get the green arrow and earn the PSR rise.

Check this pair of screenshots, please:





When I saw that the team was falling apart, I did not resign and decided to try to do something. Thus, I got the green arrow.

rox5tar, on 13 April 2024 - 09:45 PM, said:

Just played a match, where I got a 217 Match Score, but got a positive skill rating increase. Based on the match score of the other team, I was in the bottom half of everyone's match score (Meaning I got a lower score than 12 of the players in the match). Was also a loss, if that matters.

Without seeing the Mission Summary Team Table and the Mission Summary Player Table, it is really hard to say.

Imagine it as a list of 24 players, sorted according their Match Score (the players of the victorious team get a small bonus) after the game. The upper 12 'Mechs get the green arrow, the lower half of the list gets the red arrow. The higher you are among the "green" 'Mechs, the higher is your PSR raise. The same is true for the "red" Mechs - in the negative sense, of course.

So it is not a pure list of players and their Match Scores. The players of the victorious team have a small bonus that pushes their MS higher in the list. I guess that this is what confused you.

Thus, in one game MS 217 can be good enough to place you in the upper half of the list, and in the next game the same Match Score could place you in the lower half of the list.

[Inatu Elimor]

These type of discussions go on and on and repeat themselves ad nauseam. An argument that I never see is one from a psychological standpoint of view. Every psychologist can tell you that you motivate people by giving positive feedback. Negative feedback will only frustrate people and hamper the learning process. So if you do very well in relation to your average skills and get a red arrow then from a psychological view, not a mathematical or statiscal, this is indeed silly.

[martian]

Inatu Elimor, on 13 April 2024 - 11:57 PM, said:

These type of discussions go on and on and repeat themselves ad nauseam. An argument that I never see is one from a psychological standpoint of view. Every psychologist can tell you that you motivate people by giving positive feedback. Negative feedback will only frustrate people and hamper the learning process. So if you do very well in relation to your average skills and get a red arrow then from a psychological view, not a mathematical or statiscal, this is indeed silly.

Negative feedback is also important in the process of learning. Knowing that you did something wrong is crucial for not repeating your mistakes.

Play as you like, use 'Mechs and loadouts that you like. Let PSR do its job. Eventually, PSR will move you to the most appropriate Tier, i.e. Tier where you will face other players of your skill level and running similar 'Mechs.

After some time, you reach your plateau. The number of green and red arrows will be in balance. In one game you will move up slightly, in the next game you will go down. Considering some longer time period, the number of green arrows and red arrows will be the same.

Edited by martian, 14 April 2024 - 01:09 AM.

[pattonesque]

rox5tar, on 13 April 2024 - 06:25 PM, said:

- I've solo capped the CTF point, (As in fully capped by myself) and while my damage was minimal I won us the game. Skill rating went down.

capping a base in assault might technically win the game but it's not really a measure of skill. Maybe you read the tactical situation perfectly and capped the base at the perfect time, but the matchmaker cannot differentiate between that and someone who mindlessly W-keyed into the base while the enemy team got slaughtered.

also to be frank, capping the base as a first resort and not a way to end the game when some Shadow Cat is fruitlessly stringing out an 11-1 stomp is kind of lame. I'm playing this game to shoot robots, and when someone Stands In Square it kinda makes one wonder why they're even playing Mech Warrior On Line.

[Gasboy]

pattonesque, on 14 April 2024 - 05:39 AM, said:

also to be frank, capping the base as a first resort and not a way to end the game when some Shadow Cat is fruitlessly stringing out an 11-1 stomp is kind of lame. I'm playing this game to shoot robots, and when someone Stands In Square it kinda makes one wonder why they're even playing Mech Warrior On Line.

I'd rather someone play the objective early and win than they not play the objective and lose. Nothing like running a heavy or assault and lose a domination match five steps away from the circle because the lights and mediums wouldn't step into the circle.

[pattonesque]

Gasboy, on 14 April 2024 - 09:43 AM, said:

I'd rather someone play the objective early and win than they not play the objective and lose. Nothing like running a heavy or assault and lose a domination match five steps away from the circle because the lights and mediums wouldn't step into the circle.

I'd rather play a game which ended up being a fun fight that was a loss than have some four-man of capping spiders end the match before anyone fires a shot. Like the whole point is that you load up in your big stompy robot and fire big guns at other stompy robots, not "stand in square lol"

[martian]

Gasboy, on 14 April 2024 - 09:43 AM, said:

I'd rather someone play the objective early and win than they not play the objective and lose. Nothing like running a heavy or assault and lose a domination match five steps away from the circle because the lights and mediums wouldn't step into the circle.

You are free to step in the circle in your heavy or assault 'Mech any time you wish.

[Gasboy]

martian, on 14 April 2024 - 11:27 AM, said:

You are free to step in the circle in your heavy or assault 'Mech any time you wish.

You miss my point.

I always head for the circle, no matter what class I am in. In lights or mediums, it's to try and get a few quick seconds off our timer. In heavies or assaults, I go because I know there will be a fight there.

But if I, in an assault, is the one who has to step into the circle to stop our team from losing, when there are lights and mediums still alive? They're doing it wrong.

pattonesque, on 14 April 2024 - 10:00 AM, said:

I'd rather play a game which ended up being a fun fight that was a loss than have some four-man of capping spiders end the match before anyone fires a shot. Like the whole point is that you load up in your big stompy robot and fire big guns at other stompy robots, not "stand in square lol"

Like the whole point of a game is to try to win, is it not? Whether it be 'stand in a square' or 'shoot all the red bots'.

[pattonesque]

Gasboy, on 14 April 2024 - 01:26 PM, said:

Like the whole point of a game is to try to win, is it not? Whether it be 'stand in a square' or 'shoot all the red bots'.

You can win both ways. One way is an *incredible waste of everyone’s time*

[martian]

Gasboy, on 14 April 2024 - 01:26 PM, said:

You miss my point.

I always head for the circle, no matter what class I am in. In lights or mediums, it's to try and get a few quick seconds off our timer. In heavies or assaults, I go because I know there will be a fight there.

But if I, in an assault, is the one who has to step into the circle to stop our team from losing, when there are lights and mediums still alive? They're doing it wrong.

Why?

Is any reason why you can not step in your assault 'Mech in the circle?

[crazytimes]

rox5tar, on 13 April 2024 - 06:25 PM, said:

I know since it's inception, the pilot skill rating has been very... controversial. Recently though, I don't know if there have been changes or it's just been my bad luck but I've found that the "rating" is VERY inconsistent.

I've had matches where:

- I've solo capped the CTF point, (As in fully capped by myself) and while my damage was minimal I won us the game. Skill rating went down.

"I literally did nothing, moved nowhere, shot nothing, and am surprised I didn't get marked up for skill".

Quote

- Had a match with 542 Damage (Second most on my team), 1 Kill (KMDD) and 4 Assists. Match Score was 291 (Third highest on my team). My match score went down.

"I was the third best loser and didn't get marked up for skill"

Quote

- Had a match where I dealt 691 Damage (second most on my team), 1 Kill and 5 Assists. Match score was 372 (Third best on my team). My match score stayed the same.

"I was the third best loser and didn't get marked up for skill"

Quote

What's the actual process for calculating pilot skill rating?

Here's a video BrioS prepared 3 years ago that get's endlessly reposted. I know it's already in the thread as well as every other PSR thread- but people keep claiming their poor kill is somehow hidden behind a mystery secret process. It's not.

Edited by crazytimes, 14 April 2024 - 10:27 PM.