Jump to content
You need to play a total of 5 battles to post in this section.
Ombromondo

PR vs WR

167 comments in this topic

Recommended Posts

84
[RM-I]
Members
345 posts
21,599 battles

Hi
I always return to the topic because I find that the PR rating is totally wrong, which takes into account the number of battles won.
I repeat it for my precious time, PR is, in fact, personal. It has nothing to do with the battles won.
If, I repeat "if", I'm pretty good but I often find myself with really bad team, I can not win alone. So since the formula is also inserted, the WR is eventually penalized (of course I mean all of us).
It would be more appropriate to be evaluated not so much by the number of battles won (which has its own logic, but separate) but how good it is with the ships.
This is my thought !!

 

 

- translated with google translate -

Share this post


Link to post
Share on other sites
Members
2,614 posts
3,583 battles

Here is my unpopular opinion. Personal rating, WTR, whatever you want to call it. It is not as far off as most claim it is when indicating player skill.

 

The reason? In order to effectively win matches in this game, you need to be able to carry your own weight (server averages) multiplied two or three times.

 

That means, in my opinion, if the server average for the ship you are playing is 1 kill and 50,000 damage per match. Your personal average needs to be 2-3 kills, and 100-150k damage to see a large amount of control for your win rate (in order to stay above 59%).

  • Cool 2

Share this post


Link to post
Share on other sites
84
[RM-I]
Members
345 posts
21,599 battles

Then? I do not understand if you are comfortable with how PR is calculated.
For example, I can almost, almost = not always or often, in every battle sink from 1/2 to 4 ships and I get from 90 to 120k. Yet I've been blocked for a while at a low level in my opinion. But I do not want to talk about me. I mean the PR should be your level, which has nothing to do with WRs.
Because the WRs also depend on the team and, unfortunately, the teams are so casual that you can find players who have just begun, perhaps buying high-quality premium vessels but they may not use it.
So I repeat the PR is, in my opinion, and must be free from other considerations.

Share this post


Link to post
Share on other sites
488
[O7]
[O7]
Beta Testers
1,575 posts
9,511 battles

WR is still a decent determination of skill, especially over the course of thousands of battles. I myself an approaching 66% WR and I know many others have WRs even higher than that. We are all fairly skilled players and can often out preform many average players. This skill reflects in most cases higher WR in any particular ship compared with the average. 

Share this post


Link to post
Share on other sites
723
[NEUTR]
Members
2,207 posts
8,127 battles
48 minutes ago, twitch133 said:

Here is my unpopular opinion. Personal rating, WTR, whatever you want to call it. It is not as far off as most claim it is when indicating player skill.

 

The reason? In order to effectively win matches in this game, you need to be able to carry your own weight (server averages) multiplied two or three times.

 

That means, in my opinion, if the server average for the ship you are playing is 1 kill and 50,000 damage per match. Your personal average needs to be 2-3 kills, and 100-150k damage to see a large amount of control for your win rate (in order to stay above 59%).

Bull crap. WTR is a horrible rating. If there's to be any rating it should be based solely on Win/loss.

 

For a statistical "raiting" to be of any use there are two factors reliability and validity. Reliable as in it is consistent for every player. Validity means the measurement is what fits the game in question. WTR is definitely reliable. But its validity is up to debate. 

There's a reason in tournament matches and competitive matches the behavior of players differs from random battle, even though they both chase the same ultimate goal - win the match. The meta in competitive plays is "optimal" because it maximizes the utility and damage output of all players involved., But by doing so damage done is spread fairly evenly across all members, and by WTR's rating, all players involved are 1000 rating plebs. Because they all do the same damage, the average.

 

WTR does not promote "optimal" meta. It is heavily based on raw damage numbers, to the point one can get into top 100 rating with sub 50% win rate. Raw damage number is not how the game reward players, the game uses % damage done to target. The game also rewards players for "win". If WTR changes their raw damage number to % dmg to target, then it would be a more "valid" rating.


Good players have better WTR ratings, but good players will have good any ratings. Just because good players have high WTR ratings does not make WTR a "good" rating. Competitive rating is its own unique field of research in math and statistics. For a rating system to be valid, since it is a mathematical operation, there needs to be "proof". WTR has none of that. It is a arbitrary standard.

Edited by NeutralState
  • Cool 2

Share this post


Link to post
Share on other sites
Members
2,614 posts
3,583 battles
22 minutes ago, Ombromondo said:

Then? I do not understand if you are comfortable with how PR is calculated.
For example, I can almost, almost = not always or often, in every battle sink from 1/2 to 4 ships and I get from 90 to 120k. Yet I've been blocked for a while at a low level in my opinion. But I do not want to talk about me. I mean the PR should be your level, which has nothing to do with WRs.
Because the WRs also depend on the team and, unfortunately, the teams are so casual that you can find players who have just begun, perhaps buying high-quality premium vessels but they may not use it.
So I repeat the PR is, in my opinion, and must be free from other considerations.

 

I am addressing the bold section.

 

WR has a huge amount to do with personal performance. Sure, over extremely small sample sizes (under 50 battles) teams can play a big part in it. You can get carried to victories you did not deserve, and you can lose matches with 200k+ damage with 5 or more kills. This is true in small samples.

 

But, once you start talking hundreds or thousands of battles, the ONLY defining factor in your win rate is YOU.

 

Do you think that the poster above me, @Kevik70 got to 66% by shear luck? By getting the luck RNG roll on team line-up for 4,300 battles? Believe me, nobody is that lucky, or unlucky. He got that win rate by being good, by doing lots of damage to the right targets at the right place, at the right time.

  • Cool 1

Share this post


Link to post
Share on other sites
Members
2,614 posts
3,583 battles
3 minutes ago, NeutralState said:

Bull crap. WTR is a horrible rating. If there's to be any rating it should be based solely on Win/loss.

 

For a statistical "raiting" to be of any use there are two factors reliability and validity. Reliable as in i

 

So you are telling me that somebody with 40% less than server average damage in a ship, and less than 0.5 frags per battle could be capable of maintaining a >60% win rate in that ship, solo, over >100 battles?

 

(you know, all of the other stuff that goes into WTR? Do you think it is a coincidence that the best players are purple in win rate AND WTR?)

Edited by twitch133
  • Cool 1

Share this post


Link to post
Share on other sites
488
[O7]
[O7]
Beta Testers
1,575 posts
9,511 battles
6 minutes ago, twitch133 said:

 

So you are telling me that somebody with 40% less than server average damage in a ship, and less than 0.5 frags per battle could be capable of maintaining a >60% win rate in that ship, solo, over >100 battles?

 

(you know, all of the other stuff that goes into WTR? Do you think it is a coincidence that the best players are purple in win rate AND WTR?)

 

Indeed, typically people with purple WR have purple WTR/PR, blue WRs have Blue WTR/WR and so on and so forth. Now some individual ships many vary like my Minotaur and Mahan stats which are very solid green (purple and blue in WTR) with sub 50% WRs, but for the most part I get purple WRs and Purple/ high blue PR in most ships.

  • Cool 1

Share this post


Link to post
Share on other sites
723
[NEUTR]
Members
2,207 posts
8,127 battles
12 minutes ago, twitch133 said:

 

I am addressing the bold section.

 

WR has a huge amount to do with personal performance. Sure, over extremely small sample sizes (under 50 battles) teams can play a big part in it. You can get carried to victories you did not deserve, and you can lose matches with 200k+ damage with 5 or more kills. This is true in small samples.

 

But, once you start talking hundreds or thousands of battles, the ONLY defining factor in your win rate is YOU.

 

Do you think that the poster above me, @Kevik70 got to 66% by shear luck? By getting the luck RNG roll on team line-up for 4,300 battles? Believe me, nobody is that lucky, or unlucky. He got that win rate by being good, by doing lots of damage to the right targets at the right place, at the right time.

 

It doesn't matter. Using Chess's Elo rating as an example, it takes # of matches for a player to "converge" to their "true rating". If player numbers increase and game modes/maps increase, the convergence time simply increases (quite dramatically). So if in a 1 v 1 game, a few games is enough to find a player's true rating. A 5 vs 5 takes about 1000, but around 100 it's pretty much in the ball park. For a game of 12 vs 12 with dozens of maps, it takes quite a bit of time. 

Share this post


Link to post
Share on other sites
Members
2,614 posts
3,583 battles
2 minutes ago, Kevik70 said:

 

Indeed, typically people with purple WR have purple WTR/PR, blue WRs have Blue WTR/WR and so on and so forth. Now some individual ships many vary like my Minotaur and Mahan stats which are very solid green (purple and blue in WTR) with sub 50% WRs, but for the most part I get purple WRs and Purple/ high blue PR in most ships.

 

I have seen some exceptions. Like DD mains that mostly play in division. When they are better at supporting division mates than they are dealing damage, but still contributing a huge amount to the win by smoking and spotting for their div mates.

 

But, its not like they are going to have a purple win rate, with red or yellow PR doing that. If they are good enough to support their div mates like that, they are likely still going to be dark blue on PR.

Share this post


Link to post
Share on other sites
351
[GRFOX]
Members
2,242 posts
5,214 battles
32 minutes ago, Ombromondo said:

I do not understand the way

 

I.e. All the losses you accrue if you don't free xp a bad ship. Great example of a ship which is on purpose bad IJN BB tier 9 for a good reason Yamato is tier 10

Edited by GreyFox78659
  • Cool 1

Share this post


Link to post
Share on other sites
Members
2,614 posts
3,583 battles
3 minutes ago, NeutralState said:

 

It doesn't matter. Using Chess's Elo rating as an example, it takes # of matches for a player to "converge" to their "true rating". If player numbers increase and game modes/maps increase, the convergence time simply increases (quite dramatically). So if in a 1 v 1 game, a few games is enough to find a player's true rating. A 5 vs 5 takes about 1000, but around 100 it's pretty much in the ball park. For a game of 12 vs 12 with dozens of maps, it takes quite a bit of time. 

 

I will ask you to stop beating around your point and just come out and say it.

 

You believe that these players that are >65% win rate over 4-8 thousand battles, got lucky with their teams that many times???? And they have not had enough time for their rating to converge yet?

 

Are you telling me my personal skill level has absolutely nothing to do with my above average win rate, because I do not have enough battles yet? That my dark blue / light purple that I have been maintaining is all luck of RNG?

Edited by twitch133

Share this post


Link to post
Share on other sites
351
[GRFOX]
Members
2,242 posts
5,214 battles
5 minutes ago, twitch133 said:

 

I will ask you to stop beating around your point and just come out and say it.

 

You believe that these players that are >65% win rate of 4-8 thousand battles, got lucky with their teams that many times???? And they have not had enough time for their rating to converge yet?

No, they typically division a lot and like I said free xp past bad ships and also buy OP premiums. I lone wolf which is why my rating sucks I never division and play ships that aren't good but have something cool about them. Stat padding is nothing new to video games there are always ways to cheat the system without actually cheating at the game. In the old people used to quit before a loss was recorded now they play OP stuff.

Edited by GreyFox78659

Share this post


Link to post
Share on other sites
Members
2,614 posts
3,583 battles
3 minutes ago, GreyFox78659 said:

No, they typically division a lot and like a I said free xp past bad ships and also buy OP premiums.

 

So, you are telling me that somebody capable of maintaining 65-80% win rate in a division, is all of a sudden going to be a <50% player if you put them in matches solo?

 

And you are telling me you can take three 45% players, stick them in a division and they will be able to pull down a 65% win rate?

Share this post


Link to post
Share on other sites
351
[GRFOX]
Members
2,242 posts
5,214 battles
1 minute ago, twitch133 said:

 

So, you are telling me that somebody capable of maintaining 65-80% win rate in a division, is all of a sudden going to be a <50% player if you put them in matches solo?

 

And you are telling me you can take three 45% players, stick them in a division and they will be able to pull down a 65% win rate?

Yep, you didn't  say who would be in the Yamato and who would be in the Bismarck.

Share this post


Link to post
Share on other sites
351
[GRFOX]
Members
2,242 posts
5,214 battles

I have seen high wtr players in low tiers and killed them with ease because the MM gods said to. So ratings means nothing to me. If I cared trust me my WTR would be much higher. I just don't believe in stat padding.

Edited by GreyFox78659
  • Cool 1

Share this post


Link to post
Share on other sites
Members
2,614 posts
3,583 battles
1 minute ago, GreyFox78659 said:

Yep, you didn't  say who would be in the Yamato and who would be in the Bismarck.

 

It does not matter if they are in Yamato, Bismarck, Nikolai or Krispy Kreme.

 

A good player knows where he needs to be with his ship, and is going to be able to do damage to the right targets when he needs to. Regardless of whether the ship he is driving is considered OP or UP by the masses. Even in underwhelming ships like New Orleans. The top 5% of players on the server can win 70% of their games in it.

Share this post


Link to post
Share on other sites
Members
2,614 posts
3,583 battles
2 minutes ago, GreyFox78659 said:

I have seen high wtr players in low tiers and killed them with ease because the MM gods said to. So ratings means nothing to me. If I cared trust me my WTR would be much higher. I just don't believe in stat padding.

 

So, explain to me how I am padding my stats?

Share this post


Link to post
Share on other sites
84
[RM-I]
Members
345 posts
21,599 battles
6 minutes ago, twitch133 said:

 

It does not matter if they are in Yamato, Bismarck, Nikolai or Krispy Kreme.

 

A good player knows where he needs to be with his ship, and is going to be able to do damage to the right targets when he needs to. Regardless of whether the ship he is driving is considered OP or UP by the masses. Even in underwhelming ships like New Orleans. The top 5% of players on the server can win 70% of their games in it.

but come on, what do you say! you can know how to do your job well, but if you have a bad team you still died before you started.

especially from Friday to Sunday

Edited by Ombromondo

Share this post


Link to post
Share on other sites
Members
2,614 posts
3,583 battles
3 minutes ago, Ombromondo said:

but come on, what do you say! you can know how to do your job well, but if you have a bad team you still died before you started.

 

I won a game against fratstar yesterday, one of the best players on both NA and EU servers, along with an OPG clan mate that was with him. Now, could I repeat that on a regular basis? Probably not.

 

But... My team was by every means, the underdog. And with some skillful, and tactful playing on my part, with the help of a few similarly skilled clan mates, and a few mistakes by the flank of frat's team that he was not on, we managed to squeak out a win.

 

Not by luck, but by skillful play. (Even more so considering that Frat and his clan mate had 6 kills between the two of them.)

Edited by twitch133

Share this post


Link to post
Share on other sites
84
[RM-I]
Members
345 posts
21,599 battles
2 minutes ago, twitch133 said:

 

I won a game against fratstar yesterday, one of the best players on both NA and EU servers, along with an OPG clan mate that was with him. Now, could I repeat that on a regular basis? Probably not.

 

But... My team was by every means, the underdog. And with some skillful, and tactful playing on my part, with the help of a few similarly skilled clan mates, and a few mistakes by the flank of frat's team that he was not on, we managed to squeak out a win.

 

Not by luck, but by skillful play.

this confirms what I say! most likely the adversary team had 2/3 very good and 9 pitiful. That is why I say that the WR should not be considered in the PR account

Share this post


Link to post
Share on other sites
84
[RM-I]
Members
345 posts
21,599 battles

Anyway, I'm talking about the fact that the WR should not be considered in the WR count, which is another thing.
PR is the clean result of your actions, not the victories.

Edited by Ombromondo

Share this post


Link to post
Share on other sites
Members
2,614 posts
3,583 battles

Why am I even arguing.

 

Those who want to believe that their performance is predetermined by RNG are going to believe it no matter what proof there is otherwise. They are going to refuse to admit that it could possibly be their fault, and they are going to stagnate at their current performance.

 

Where as those of us that choose the red pill, will recognize that performance comes from within, we will learn, we will make our strengths stronger, and try to minimize our weaknesses. And improve.

  • Cool 2

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×