Average Team Ranking vs Winrate

[Hawg]

Does WG share enough information so that sites like wow-numbers could provide average team ranking in addition to win rate?   I am currently enjoying one of those losing streaks and it got me thinking that average team ranking would be a more reliable indicator of "skill" than win rate.   Because even if you lose a lot, if you are at the top of your team you would still do well.  And you have a lot more control over your own performance than you have influence on the win loss (ofc you still effect both).

 

Its probably not ideal, but in many ways I think it would be a more effective way to rate people than the current practice of of judging people by their win rates.   What do you think?

[Tyrendian89]

that could indeed be a rather interesting stat, except for CV players who suffer from crazy bad economy (as in, you could only compare that to other CV players, not other classes). Too bad average XP is also a 105% worthless stat as per WG API...

[EmilyFara]

I don't think that would be a good indication. Some ships have really good xp multipliers and end up higher on the list. And other players might do nothing but hang back,survive and farm damage while contributing nothing, but ending on top due to them surviving the longest while sacrificing allies.

[waxx25]

So OP wants yet another win/loss predictor so people can rage quit before battle starts like in WOT? Yes thats a great idea.......what can go wrong.

Can you imagine if xvm was common in this game??? Then all the people with those fancy purple stats (like me ) would know what it feels like to be Flamu and the stupid-focus he gets every match.

[Tyrendian89]

So OP wants yet another win/loss predictor so people can rage quit before battle starts like in WOT? Yes thats a great idea.......what can go wrong.

 

that's what I thought this was about when reading just the title - but actually reading his post helps...

[Active0815]

I do wish that there would ba a "Best player of the losserteam" badge so often...

I guess the stats are playerbased and wow-numbers can't get the nessecary informations. Some players also hide their stats...

 

[waxx25]

 

that's what I thought this was about when reading just the title - but actually reading his post helps...

 

I did read, I also looked ahead as to what that kind of info would enable.

[Hawg]

 

I did read, I also looked ahead as to what that kind of info would enable.

 

I dont think what Im asking about would make the problem you are complaining about any better or worse.    If someone is the type of idiot who quits games immediately upon researching his teammates' win rankings, I think he would do the same with team rankings.   And yes this woukd not be perfect, but I think it could be useful additional information.     But Im not sure if it is technically feasible.   Im not sure WG makes that data avaialble?

[ColonelPete]

Yeah, teamranking says a lot when you get 8th place with 1000 points and the first place has 1055 points....

 

WR is a better indicator in the long run.

[waxx25]

 

I dont think what Im asking about would make the problem you are complaining about any better or worse.    If someone is the type of idiot who quits games immediately upon researching his teammates' win rankings, I think he would do the same with team rankings.   And yes this woukd not be perfect, but I think it could be useful additional information.     But Im not sure if it is technically feasible.   Im not sure WG makes that data avaialble?

 

I used to think that xvm in WOT was "usefull" information until it became clear it was doing more harm than good.

[Muttley44]

I think win rate is totaly random for most players.

 

Do not care what anyone else says as I am always right.

 

WG could you please let me win more and give more points for camping, also would prefer a Caravan

[eliastion]

I think win rate is totaly random for most players.

 

Do not care what anyone else says as I am always right.

 

Well, you COULD be right. Assuming that most players have about 10 battles played, that is 

[Hawg]

Yeah, teamranking says a lot when you get 8th place with 1000 points and the first place has 1055 points....

 

WR is a better indicator in the long run.

 

In the long run both team ranking and win ranking would reach a more accurate "answer" in which outlier games like the one you describe are averaged out by their true performance.   Examples like the one you give happen in both ranking methods (for example, the game where you get killed by a devastating strike and are out of the game before you do a single damage, yet your team still wins, counts just as much as the game where you do 200k damage).    

 

Ignoring the technical feasibility, why would you think that someone with 1500 games and a team ranking average of 2.68 (meaning on average they are ranked 2 between second and third in any given battle, counting both victories and defeats), is a less accurate indicator  of their "skill" than knowing that same person has a win ranking of 54%?   (Or pick any other numbers.)

 

In some senses team-ranking may actually be beneficial, because people would still try when the battle is a lost cause because they want to at least maximize their team ranking, whereas now there are people who bail for the edges once victory becomes "all but impossible" because a loss is a loss.

[Gnirf]

 

In the long run both team ranking and win ranking would reach a more accurate "answer" in which outlier games like the one you describe are averaged out by their true performance.   Examples like the one you give happen in both ranking methods (for example, the game where you get killed by a devastating strike and are out of the game before you do a single damage, yet your team still wins, counts just as much as the game where you do 200k damage).    

 

Ignoring the technical feasibility, why would you think that someone with 1500 games and a team ranking average of 2.68 (meaning on average they are ranked 2 between second and third in any given battle, counting both victories and defeats), is a less accurate indicator  of their "skill" than knowing that same person has a win ranking of 54%?   (Or pick any other numbers.)

 

In some senses team-ranking may actually be beneficial, because people would still try when the battle is a lost cause because they want to at least maximize their team ranking, whereas now there are people who bail for the edges once victory becomes "all but impossible" because a loss is a loss.

 

You also may want to do some damage as this a major source to credits regardless of winning/losing

[eliastion]

In some senses team-ranking may actually be beneficial, because people would still try when the battle is a lost cause because they want to at least maximize their team ranking, whereas now there are people who bail for the edges once victory becomes "all but impossible" because a loss is a loss.

 

It might also be harmful since people would be even less inclined to play for the victory rather than individual rewards (especially compared to the team) - have you never seen matches that could be won if someone just capped a nearby point instead of chasing a kill? Or ones that WERE won because someone sacrificed a bit of damage to either cap a point and ensure point victory, or went back to defend (and often in a suicidal manner too) when the rest of the team was too occupied slugging it out with some remnants defending enemy base to notice the enemy force threatening own base?

And don't get me started on CVs that have the choice of retrieving their planes after strike or keeping an enemy DD or CV spotted, getting some spot damage that's hardly even rewarded while sacrificing own damage potential and boosting that of their allies' - basically working actively to lower their position in the team (while trying their best to have the team be victorious).

 

The optimal actions to be taken to ensure victory are pretty often outright harmful to your final standing in the team because when you help your teammates, you enchance their relative performance while expending time and resources (planes, smokes etc.) you could use to farm some more damage. Frankly, I'm pretty happy that we don't have such obviously selfish statistic where you directly benefit from your teammates being unsuccessful.

[daki]

Does WG share enough information so that sites like wow-numbers could provide average team ranking in addition to win rate?   I am currently enjoying one of those losing streaks and it got me thinking that average team ranking would be a more reliable indicator of "skill" than win rate.   Because even if you lose a lot, if you are at the top of your team you would still do well.  And you have a lot more control over your own performance than you have influence on the win loss (ofc you still effect both).

 

Its probably not ideal, but in many ways I think it would be a more effective way to rate people than the current practice of of judging people by their win rates.   What do you think?

 

Before any other comments, let me state my firm belief: There are no "good" single indicators in WoWs. Every single one of them has major issues and is unreliable to a greater or lesser extent.

Only a combination of indicators can give some more accurate assessment of skill, but even then there are major issues since all of them focus on "quantity" and none is reliable in measuring "quality".

 

Now to your idea, I understood it as follows:

- It is XP based hence includes quantifiable actions which are currently not available on external sites such as contribution to cap/def...

- To address the issue of what is "good" or "bad" amount of XP which otherwise impossible to accurately assess, you propose to use it as a relative indicator, i.e. relative to the performance of the rest of teammates

All in all, I would like to see such an indicator since it is quite an interesting one, and it has some merits.

 

And now the "however" part :

Despite it being interesting and indicative of contribution to a certain extent, it still cannot beat winrate - if we only look at single indicators.

Why?

1. Camping playstyle is for me a "low skill" playstyle, overly rewarded currently by XP

2. When you look at XP in a losing team, there are way too many weak players in the top half

3. In games like WoWs where "winning" is the ultimate/main match objective what does "skill" matter if the match is lost?

4. XP from time to time poorly measures contribution to win. I have often seen players with 1000 base XP contribute more to the win than those in the 1500-2000 range (if they got most of that XP after the battle has been already decided)

 

All in all, despite all the issues with winrate as indicator - needing large series both overall and per ship to be relevant, prone to dependence on teammates in case of shorter series... - it is still by far the best single stat available.

If nothing else, than at least because it captures (on average) your crucial contributions to the win better than any other available stat.

Just a small example: You did below average damage, and the XP isn't great either, but you sinking that pesky enemy DD and preventing cap was the main reason for the team's win. Or you positioned well to scare away those BBs pushing a flank (not that it is too difficult to scare them away ) and hence indirectly prevented the enemy gathering a large number of points or flanking your teammates which would lead to them getting obliterated and losing the game.

 

If I could chose teammates, I would always pick a high winrate player with lower damage, XP (or any other stat) instead of the low WR player but who has higher damage...

[loppantorkel]

I suggested something similar a good while ago - it could be a colour indicator showing an approximate how good/bad the teams were as a whole - post battle. It might be of some comfort when in losing streaks.

[Hawg]

 

Before any other comments, let me state my firm belief: There are no "good" single indicators in WoWs. Every single one of them has major issues and is unreliable to a greater or lesser extent.

Only a combination of indicators can give some more accurate assessment of skill, but even then there are major issues since all of them focus on "quantity" and none is reliable in measuring "quality".

 

Now to your idea, I understood it as follows:

- It is XP based hence includes quantifiable actions which are currently not available on external sites such as contribution to cap/def...

- To address the issue of what is "good" or "bad" amount of XP which otherwise impossible to accurately assess, you propose to use it as a relative indicator, i.e. relative to the performance of the rest of teammates

All in all, I would like to see such an indicator since it is quite an interesting one, and it has some merits.

 

And now the "however" part :

Despite it being interesting and indicative of contribution to a certain extent, it still cannot beat winrate - if we only look at single indicators.

Why?

1. Camping playstyle is for me a "low skill" playstyle, overly rewarded currently by XP

2. When you look at XP in a losing team, there are way too many weak players in the top half

3. In games like WoWs where "winning" is the ultimate/main match objective what does "skill" matter if the match is lost?

4. XP from time to time poorly measures contribution to win. I have often seen players with 1000 base XP contribute more to the win than those in the 1500-2000 range (if they got most of that XP after the battle has been already decided)

 

All in all, despite all the issues with winrate as indicator - needing large series both overall and per ship to be relevant, prone to dependence on teammates in case of shorter series... - it is still by far the best single stat available.

If nothing else, than at least because it captures (on average) your crucial contributions to the win better than any other available stat.

Just a small example: You did below average damage, and the XP isn't great either, but you sinking that pesky enemy DD and preventing cap was the main reason for the team's win. Or you positioned well to scare away those BBs pushing a flank (not that it is too difficult to scare them away ) and hence indirectly prevented the enemy gathering a large number of points or flanking your teammates which would lead to them getting obliterated and losing the game.

 

If I could chose teammates, I would always pick a high winrate player with lower damage, XP (or any other stat) instead of the low WR player but who has higher damage...

 

Yes you understood my idea correctly and actually stated it better than I did   On your "however" point, I agree with what you are saying, but keep in mind I was saying to add this as an additional measure, not a replacement statistic.  (The title I chose for the thread may not make that apparent.)  

 

In the long run, I suspect that for most players there would be a high level of correlation between the two measures.   But, to some extent it would provide a bit of "bad luck protection" for people so they can still improve on one stat even when the other stat is taking a hit (i.e. still get credit for the win rate when they do something to help the team win that is not rewarded in exp, or when they gain lots of experience but their team loses).  

 

People who tried to "game the system" by playing in a way that disproportionately benefits the team ranking stat would suffer in the win rate stat (unless it also happened to help their teams win more of the time).   So I wouldn't think that we'd suddenly see every battle full of 12 Sims for example.   

 

And then for people who have both high win rates, and high team rankings, I think it would be about as good of an indicator as we can get to identify who the "great" players are (it certainly wouldn't be me).   And when someone has one stat that is great while the other is weak, it would be a good indicator that they're doing something odd.

 

And just to be clear, I may sound like I'm strongly advocating it, but I'm really just trying to have the discussion, a bit of theorycrafting if you will.  I don't know the right answer.

[daki]

 

Yes you understood my idea correctly and actually stated it better than I did   On your "however" point, I agree with what you are saying, but keep in mind I was saying to add this as an additional measure, not a replacement statistic.  (The title I chose for the thread may not make that apparent.)  

 

In the long run, I suspect that for most players there would be a high level of correlation between the two measures.   But, to some extent it would provide a bit of "bad luck protection" for people so they can still improve on one stat even when the other stat is taking a hit (i.e. still get credit for the win rate when they do something to help the team win that is not rewarded in exp, or when they gain lots of experience but their team loses).  

 

People who tried to "game the system" by playing in a way that disproportionately benefits the team ranking stat would suffer in the win rate stat (unless it also happened to help their teams win more of the time).   So I wouldn't think that we'd suddenly see every battle full of 12 Sims for example.   

 

And then for people who have both high win rates, and high team rankings, I think it would be about as good of an indicator as we can get to identify who the "great" players are (it certainly wouldn't be me).   And when someone has one stat that is great while the other is weak, it would be a good indicator that they're doing something odd.

 

And just to be clear, I may sound like I'm strongly advocating it, but I'm really just trying to have the discussion, a bit of theorycrafting if you will.  I don't know the right answer.

 

Pretty much agree with everything there.

To note, I was commenting the following sentence from your original post:

"Its probably not ideal, but in many ways I think it would be a more effective way to rate people than the current practice of of judging people by their win rates."

I just tried to explain why IMHO nothing beats the good old winrate

Of course, when examined together with winrate, it truly would be a good indicator

 

[Admiral_noodle]

 

 

I would like to see this stat included *aswell* as the others. Especially average position on a win vs average position on a loss.

All three would be interesting statistics.

Personally I think I'd probably average 4th, with a lower average position on a win than on a loss. ---> suggests defensive player. But I'd like to see it.

 

I don't think any stat alone indicates "skill". 

In fact all stats do is tell you where you might be able to improve. 

 

If you want to get super nerdy (and I do) I'd like to see distribution curves by ship and overall of danage and xp. By wins and losses. 

[ColonelPete]

 

Ignoring the technical feasibility, why would you think that someone with 1500 games and a team ranking average of 2.68 (meaning on average they are ranked 2 between second and third in any given battle, counting both victories and defeats), is a less accurate indicator  of their "skill" than knowing that same person has a win ranking of 54%?   (Or pick any other numbers.)

 

You can get a high ranking just by farming damage and not play for the win.

[SinkTheOthersNotMe]

I find hard to believe someone can consistently get high ranking (in randoms) while having a bad WR.

For example, never met a unicum who had a below 50% WR....

 

Yes it is possible to get a high ranking for a few battles here and there but getting consistent results while playing like a potato and camping at the back seems rather impossible to me.

 

This said, with the current little app that gives you, for the current battle, the players stats for the ship they are currently using is quite useful...

Trusting a yamato with 20k avg dmg or seeing your CV deals on average 25k dmg with a 40% WR -> compared to the enemy CV who deals 190k average dmg with a 60% WR can sometimes help.

 

This said I am not in favor of XVM.....

[eliastion]

Ignoring the technical feasibility, why would you think that someone with 1500 games and a team ranking average of 2.68 (meaning on average they are ranked 2 between second and third in any given battle, counting both victories and defeats), is a less accurate indicator  of their "skill" than knowing that same person has a win ranking of 54%?   (Or pick any other numbers.)

 

That's easy: if someone has a win ranking of 54% and, sat, team ranking of 8,13 - it means that he helps his team win matches. Hard to say HOW he possibly does this without getting his own XP up; maybe he's a US DD player who mostly spots and smokes teammates while dealing very little damage himself. But whatever he's doing: it works.

 

On the other hand, if we have a player with winrate of 45% and average team ranking of 2,68 - it means that he's a leech that scores decent XP at the expense of his team. Once again, it's hard to say how such huge disparity is possible, since getting to top 3 requires quite some damage and it's hard to consistently score good damage while consistently dragging your team down (especially since most matches don't end with winners chasing down the last survivor - the points usually decide things before that happens). Still, the stats show that such player is, more often than not, a liability to his team rather than an asset, despite often getting good position in the team ranking.

[ColonelPete]

I find hard to believe someone can consistently get high ranking (in randoms) while having a bad WR.

For example, never met a unicum who had a below 50% WR....

 

Yes it is possible to get a high ranking for a few battles here and there but getting consistent results while playing like a potato and camping at the back seems rather impossible to me.

 

This said, with the current little app that gives you, for the current battle, the players stats for the ship they are currently using is quite useful...

Trusting a yamato with 20k avg dmg or seeing your CV deals on average 25k dmg with a 40% WR -> compared to the enemy CV who deals 190k average dmg with a 60% WR can sometimes help.

 

This said I am not in favor of XVM.....

 

At least a higher ranking than his WR would suggest.

There are some ways to game the system (playing for XP) and not for the win.

[22cm]

There are some ways to game the system (playing for XP) and not for the win.

 

I hate that kind of play.