Jump to content


Are our Skilled Players Pulling their Weight?

stats

  • Please log in to reply
81 replies to this topic

GeorgePreddy #21 Posted Mar 02 2019 - 17:47

    Major

  • Players
  • 14680 battles
  • 13,519
  • Member since:
    04-11-2013

View PostFestung, on Mar 02 2019 - 13:27, said:

These statistics prove that MM is structured wrong. Mixing Purples and reds frustrates both. MM should pit only red&orange together, yellow& green and blue&purple. This would enable the lesser players to gain some experience before facing the better players. I read here in the forum that "a tank is only as good as the player", so the type of tank shouldn't be the primary condition for MM but the skill of the player. As a player advances in skill, they would move up in color and subsequently into a higher match pool. Just a way to make the teams more evenly matched IMHO.

 

SBMM is a terrible thing for Random Battles...

 

that's why WG has never even seriously considered it for Random Battles.

 

 

 

 

 

 



Mfezi #22 Posted Mar 02 2019 - 17:48

    Captain

  • -Players-
  • 28524 battles
  • 1,697
  • [SADD] SADD
  • Member since:
    04-05-2016

View Postdunniteowl, on Mar 02 2019 - 17:19, said:

They are not only "pulling" their weight, they are also pulling the weight of up to four others on the same team that are not only NOT pulling their weight, but they are waiting to be pulled.

 


I think your chart shows that pretty obviously.

 

 

GL, HF & HSYBF!

OvO

 

The answer to this would be to segregate players by skill - is this what you are hinting at?.

Alternatively, we could put all the good players on one team and all the bad in the other - sorry!, what am I talking about - how would MM be able to do that?.... :)

 



GeorgePreddy #23 Posted Mar 02 2019 - 17:49

    Major

  • Players
  • 14680 battles
  • 13,519
  • Member since:
    04-11-2013

View PostMfezi, on Mar 02 2019 - 13:48, said:

 

The answer to this would be to segregate players by skill - is this what you are hinting at?.

Alternatively, we could put all the good players on one team and all the bad in the other - sorry!, what am I talking about - how would MM be able to do that?.... :)

 

 

SBMM is a terrible thing for Random Battles...

 

that's why WG has never even seriously considered it for Random Battles.

 

Your weak, repetitive whines will not likely change that fact.

 

 

 

 

 

 

 



Mfezi #24 Posted Mar 02 2019 - 17:53

    Captain

  • -Players-
  • 28524 battles
  • 1,697
  • [SADD] SADD
  • Member since:
    04-05-2016

View PostGeorgePreddy, on Mar 02 2019 - 18:45, said:

 

fixed

 

 

 

Sorry George, had to throw your stats out, they were just too unrealistic to be included!..... :)

 



3bagsfull #25 Posted Mar 02 2019 - 17:56

    First lieutenant

  • Players
  • 2713 battles
  • 942
  • Member since:
    02-16-2014

View PostDeviouslyCursed, on Mar 02 2019 - 16:46, said:

 

This is ridiculously stupid. You get better by playing against better players. Put all the fails together and all they will do is think they are good because all they ever meet are other fails, and the same fail tactics will work over and over because no one else in their matches is doing anything better.

 

This should be painfully obvious to anyone with half a functioning brain.

 

Now, if you argued that the reds and oranges won't or can't learn anything (and that is why they are red and orange anyway) and they'd have more fun playing only against other bads, you may have a point. But to try to claim they will learn more from bad players instead of facing good players is just dumb.

 

Only 2 things are painfully obvious:

1)  Not every player no matter how much or who they play with can reach the same levels in the game

2)  It's not that the difference is in one shot per game, it's in what that player has to do to make sure he gets that one shot.   

 

When you watch a good player get absolutely punished (as he should be) by making a completely boneheaded mistake, lose 90% of his HP's in the opening 2 minutes and turn that game around into a 6,000+ damage carry - it's only then that you realize that when people quote numbers they really don't understand how those numbers are created/maintained.


Edited by 3bagsfull, Mar 02 2019 - 17:56.


SnakePliskan #26 Posted Mar 02 2019 - 18:06

    First lieutenant

  • -Players-
  • 27 battles
  • 912
  • Member since:
    07-09-2016

View PostMfezi, on Mar 02 2019 - 17:38, said:

 

Not meant to prove anything -  as I said above.... just a simple set of figures which quantify the difference in skill levels in terms of damage and kills.

 

Very difficult to post any stats on here even for simple analysis, or just informative. There are the butt hurt that try to protect their own position and achievements/lack of either way. Rather than just looking at the posted numbers in a personal respect they need to lash out or demean.

Example abound in this single thread.



Buttknuckle #27 Posted Mar 02 2019 - 18:20

    Major

  • Players
  • 51813 battles
  • 3,076
  • [GOONZ] GOONZ
  • Member since:
    03-19-2013

View PostFestung, on Mar 02 2019 - 16:27, said:

These statistics prove that MM is structured wrong. Mixing Purples and reds frustrates both. MM should pit only red&orange together, yellow& green and blue&purple. This would enable the lesser players to gain some experience before facing the better players. I read here in the forum that "a tank is only as good as the player", so the type of tank shouldn't be the primary condition for MM but the skill of the player. As a player advances in skill, they would move up in color and subsequently into a higher match pool. Just a way to make the teams more evenly matched IMHO.

 

You are not thinking though what these numbers mean with respect to your proposal. If we take a group of players that are all doing on average more than 100% (the yellow and green players) and have them only play together, it is mathematically impossible for them all to maintain the greater than 100% average. What you propose would artificially deflate the +100% groups stats, while artificially inflating the -100% groups stats, with respect to the entire game population. At which point the stats you used to separate them become meaningless.

SnakePliskan #28 Posted Mar 02 2019 - 18:25

    First lieutenant

  • -Players-
  • 27 battles
  • 912
  • Member since:
    07-09-2016

View PostMfezi, on Mar 02 2019 - 17:38, said:

 

Not meant to prove anything -  as I said above.... just a simple set of figures which quantify the difference in skill levels in terms of damage and kills.

 

Reading is difficult for those with closed eyes.

Garandster #29 Posted Mar 02 2019 - 18:56

    Captain

  • -Players-
  • 6134 battles
  • 1,253
  • Member since:
    04-12-2018

View PostMfezi, on Mar 02 2019 - 11:38, said:

 

Not meant to prove anything -  as I said above.... just a simple set of figures which quantify the difference in skill levels in terms of damage and kills.

 

It leaves me wondering what the point is of posting this, but re-reading it I think I understand the intention. However you're data is flawed intentionally or not, you're data is not usable to create any kind of actual conclusion, I don't understand the reasoning of publishing flawed data unless you have a point that you are trying to make with it.

Cruiser133 #30 Posted Mar 02 2019 - 19:06

    First lieutenant

  • Players
  • 34998 battles
  • 553
  • Member since:
    05-28-2013

View PostFestung, on Mar 02 2019 - 11:27, said:

These statistics prove that MM is structured wrong. Mixing Purples and reds frustrates both. MM should pit only red&orange together, yellow& green and blue&purple. This would enable the lesser players to gain some experience before facing the better players. I read here in the forum that "a tank is only as good as the player", so the type of tank shouldn't be the primary condition for MM but the skill of the player. As a player advances in skill, they would move up in color and subsequently into a higher match pool. Just a way to make the teams more evenly matched IMHO.

An easy solution to this using the above data is to simply stick to lower tier play until you get gud. Look at the ratio of blues and purple's at the lower tiers compared to the high number of bads. If you want a better chance of matching up with other bads then stick to the beginning tiers like the devs intended.



dunniteowl #31 Posted Mar 02 2019 - 19:09

    Major

  • -Players-
  • 31387 battles
  • 8,080
  • Member since:
    09-01-2014

View PostMfezi, on Mar 02 2019 - 10:48, said:

 

The answer to this would be to segregate players by skill - is this what you are hinting at?.

Alternatively, we could put all the good players on one team and all the bad in the other - sorry!, what am I talking about - how would MM be able to do that?.... :)

 

 

NO.   No.  no.  no   no    no

 

SBMM is NOT MEANT for Random Public Matches.  That would be like demanding that all kids on the corner lot playing baseball have to have leagues and ordered teams.

 

Skill Based Match Maker is NOT an answer for what can objectively be called a Non-Competitive Environment.  

 

What does it mean to compete in a 'competitive' environment?

 

Picture it this way....

 

      You're at the  bowling alley and you rent shoes, grab a rental ball and go out and slide some gutterballs with your friends.  This is a non-competitive environment.  Four lanes over the League Lanes exist and it's League Night, so they get lanes 10 to 40, three fourths of the alley.  You're not playing over there.

 

In the League Lanes there are:

Teams of players who play together regularly

Practiced groups and averaged scores, inclusive of handicapping for players to 'even the field' of new player versus skilled players

League Brackets so that Well Established Teams do not play scrubs and vice versa and you have to play a level of ability to change brackets

 

THAT is a competitive environment.

 

In the second environment a Skill Based Matchup System *(such as Round Robin, Elimination Ladder, Scheduled Play Aggregate) makes sense.  In the first environment, there is no real competition, there are no real teams of regularly practicing players and there are no restrictions on units, classes, tiers or skill levels in them -- or as a whole of their ability over time.

 

You cannot use a Skill Based System in a non-competitive environment and make anyone but the least able players happy -- and that's a very liberal estimate.  I would be willing to wager that a true SBMM introduced into public randoms would also require a single tier MM to be 'fair' in this sort of system.

 

Additionally, in Random Public Matches, the idea is to get a game as quickly as possible based on which unit and its MM "weight" along with its tier and class can be tossed together without wasting too much time for the waiting players.

 

In this sort of system, players can be placed together in one of two ways that make sense:

A)  All players within a bracket of 'skill' (however that level of Skill is defined, which it would have to be well defined in order to code) are placed together and if you win too often you will be automatically 'bumped' into the next bracket higher.  Each time you manage this trick of winning more, you will be 'bumped' into another bracket. 

 

     By the same token, all the players losing too much will be 'adjusted' down a bracket automatically.  In this case, the players may find themselves now playing others who are having the same level of clueless luck as they are.

 

B) All players will be pooled into the MM and then the MM will have to "Pick Teams" in a "balanced" manner wherein it picks the best two players based on this definition of "skill," then pick the next two and the next two and so forth such that fifteen players on each team are 'paired' with others of their overall (or within a bracket of 'skill' used to allow MM a bit of 'fudge factor' to complete its task) level of ability.

 

     In this scenario, NOTHING, not a damn thing, not one iota, not a molecule would be different than the current MM we now have overall.  You will still be pitted against players who stand a good chance of being much better than you or much worse than you in a match.  You cannot guaranty where on the map the best players are going to go or what tanks, tactics and tiers they will be in before they show up in the team compositions.  And, without XVM, you will be completely ignorant of how 'stacked' such a team may or may not be relative to the potential for you to meet the best player on the other team versus the worst player on the other team, combined with where you stand in that mix.

 

In case A, above, you could be having a bad streak and get bumped down a bracket, then do better, go up a bracket, slump, go down a bracket and so on and so forth, making you switch brackets every so many games.  Wouldn't this sort of thing be something you'd want to be notified of?  I know I would think it would be important enough to send an informational were it me that designed this system.  Wouldn't you want to know you went up or down a bracket?

 

Ultimately, though, how do you justify such an expenditure of effort in coding to get all this to work in what amounts to a Non-Competitive Environment.  The Argument: Playing for Fun is a very justifiably reasonable explanation for a NO to SBMM.  This is not meant to be that competitive.  If you want more competitive, then join a CLAN and play Strong Holds.  If you want more than that, Practice for League Play.

 

If you don't think it's fair to 'constantly' be pitted against more experienced players in higher tier vehicles, then this game is not for you.  This game is as fair as it gets in the way it treats the players.  There is RNG at +/- 25% to prevent the Better Players from Pawning the Bad Players all the time.  That RNG allows the 'watch is broken twice a day' level players to overcome a situation that, were it all Skill Based, would support that better player -- every time except twice a day.

 

With a wide and constantly tilting level of skill going between teams in play, the opportunity to rise and dominate exists much more readily for sub-par players than a skill based system could ever provide.  In a SBMM, you would ALWAYS be held to a potential that sees a pretty much fifty fifty outcome all the time.

 

And EVEN  in this case, fifty fifty does not mean it will work out that way.

 

People are not consistent from day to day, hour to hour or even meal to meal.  This game requires one's attention to a very high degree.  This is a "High Clutter" environment.  This means there is a wealth of data and a dearth of time in which to apply it.  Good players are going to be able to leverage this situation the majority of the time and -- because this game is a Single Death Per Match Zero Sum Game with Finite Resources -- the person who overcomes is the person who wins that engagement.  The moment you lose a tank the system is now UNBALANCED and thus, according to the dictates of a SBMM, someone should be substituted into the match of similar skill/ability -- except you can't.  A player out is a player out and that's the way it goes.

 

All that this means is that, even in a SBMM, the guy who is just ON FIRE or the player who seems to be ALL THUMBS that day cannot be adjudged except by that 'skill' definition that the SBMM will be using will not act in accordance with that defined level.  And now you have a bad match, because the guy with the rain cloud over his head is on your team and his skill level is BELOW his defined level, because he's pissed about traffic, that guy who cut him off, just had an argument with his wife/girlfriend, got fired/demoted/punished at work, etc.

 

None of these can ever be accounted for in a Random System, but it evens out over time.  In a SBMM System, this cannot really be factored out and smoothed over with anything else other than doing it over time.  The blowouts will still occur.  The snowballing loss is still going to occur.  The unexpected win/loss will still occur.

 

No matter what you end up using to define this 'skill' it will require more than one or two metric numbers taken as an aggregate and no single number will adequately cover the definition of what 'skill' even is in this game.  Due to people being somewhat fickle as a rule, the system cannot cover the variance that each person can display from day to day or even hour to hour.

 

In short, all that extra time spent coding could have been better spent for the game developers.  For the Players, all those tears, explanations, pounding of podiums/chests or exhortations of how this or that could be Fixed, could have been better spent by focusing on what they can do to better learn how the game mechanics work, how to apply those ideas to the techniques folks teach and to focus their mental efforts on asking instead, "What can I do to play better, do better, win more, hit more, etc.?"

 

The Random System, over time, promotes those who insist on playing their best and winning as much as they can and still allows those who are an overall net negative on their team's potential to win to be 'right twice a day' in the right circumstances.  RNG provides the 'fudge factor' that, as often as not punishes the good player more than it profits the poor player.  RNG allows the tomato to not be pasted in situations where no RNG would provide them a slicer and a masher.

 

No RNG benefits those with better hand/eye coordination.  No RNG benefits those who can read the battle flow on the Mini-map better than others.  No RNG provides benefits from those who can better anticipate their opponents on a consistent basis.  If you manage to fall into all three of these categories, no RNG would make you supreme amongst your peers.

 

SBMM ultimately will not change the outcomes of most matches.  It will not significantly increase playing time or make for more 'close' games.  It will not level the playing field in the sense that the competition level would be more even.  People are people and their quirks, natures and idiosyncrasies in a system that does not place people in the same place with the same people all the time cannot be fixed in this manner.  Even a SBMM cannot cover the wide variety that people can exhibit when they are still 'randomly' placed on teams in the sense that it's very likely you will:

    A) Not recognize a single player on either team from previous play

    B) Be placed with some folks who cannot even speak your language (but they're at the same overall skill level, so it's fine, right)

    C) See one, possibly two names you recognize in every ten or so matches

 

In short, your skill may be 'consistent' over time, but you are not over each day as a general rule in terms of winning or losing.  You cannot control how well you will perform each day, each time, but we all know that some days we're just 'not that into it' at that moment and our performances suffer.  Because you are not practicing regularly as part of a team that can be aggregated as an overall metric relative to other such teams, the SBMM is still just a crap shoot sorting out players as well as it can according to the principles and instructions it is given along with the players it has to sort from in the first place.

 

SBMM will not fix this situation and more time spent on it than already is a waste of time UNLESS you are interested in making WoT a MUCH MORE COMPETITIVE Environment than it now is.

 

GL, HF & HSYBF!

OvO



SnakePliskan #32 Posted Mar 02 2019 - 19:18

    First lieutenant

  • -Players-
  • 27 battles
  • 912
  • Member since:
    07-09-2016

View PostGarandster, on Mar 02 2019 - 18:56, said:

 

It leaves me wondering what the point is of posting this, but re-reading it I think I understand the intention. However you're data is flawed intentionally or not, you're data is not usable to create any kind of actual conclusion, I don't understand the reasoning of publishing flawed data unless you have a point that you are trying to make with it.

 

There is NOTHING flawed and the purpose ( if your piss poor attitude can allow your mind to work) is a simple chart of what can easily help take a red player to a higher level ( color) performance. What does it take in terms of damage to move up? TAH DAH percentage of damage in relation to your HP. SIMPLE EASY not complicated.

 

WOW you must be a peach to be around. Here is a peanut butter sandwich. THAT'S not a peanut butter sandwich your information is WRONG FLAWED. There is no butter in it. I don't want it.



Blazzy #33 Posted Mar 02 2019 - 19:35

    Captain

  • Players
  • 34961 battles
  • 1,000
  • Member since:
    03-05-2012

Define "pulling their weight"

 

Based on how this is structured you're looking for them to carry you every game, without actually saying it.  No 1 player should be required to pick up the slack of others.

 

You should concern yourself with personal gain before you worry about if a player is carrying you or not.


Edited by Blazzy, Mar 02 2019 - 19:36.


Mfezi #34 Posted Mar 02 2019 - 19:44

    Captain

  • -Players-
  • 28524 battles
  • 1,697
  • [SADD] SADD
  • Member since:
    04-05-2016

View Postdunniteowl, on Mar 02 2019 - 20:09, said:

 

NO.   No.  no.  no   no    no

 

SBMM is NOT MEANT for Random Public Matches.  That would be like demanding that all kids on the corner lot playing baseball have to have leagues and ordered teams.

 

Skill Based Match Maker is NOT an answer for what can objectively be called a Non-Competitive Environment.  

 

What does it mean to compete in a 'competitive' environment?

 

Picture it this way....

 

      You're at the  bowling alley and you rent shoes, grab a rental ball and go out and slide some gutterballs with your friends.  This is a non-competitive environment.  Four lanes over the League Lanes exist and it's League Night, so they get lanes 10 to 40, three fourths of the alley.  You're not playing over there.

 

In the League Lanes there are:

Teams of players who play together regularly

Practiced groups and averaged scores, inclusive of handicapping for players to 'even the field' of new player versus skilled players

League Brackets so that Well Established Teams do not play scrubs and vice versa and you have to play a level of ability to change brackets

 

THAT is a competitive environment.

 

In the second environment a Skill Based Matchup System *(such as Round Robin, Elimination Ladder, Scheduled Play Aggregate) makes sense.  In the first environment, there is no real competition, there are no real teams of regularly practicing players and there are no restrictions on units, classes, tiers or skill levels in them -- or as a whole of their ability over time.

 

You cannot use a Skill Based System in a non-competitive environment and make anyone but the least able players happy -- and that's a very liberal estimate.  I would be willing to wager that a true SBMM introduced into public randoms would also require a single tier MM to be 'fair' in this sort of system.

 

Additionally, in Random Public Matches, the idea is to get a game as quickly as possible based on which unit and its MM "weight" along with its tier and class can be tossed together without wasting too much time for the waiting players.

 

In this sort of system, players can be placed together in one of two ways that make sense:

A)  All players within a bracket of 'skill' (however that level of Skill is defined, which it would have to be well defined in order to code) are placed together and if you win too often you will be automatically 'bumped' into the next bracket higher.  Each time you manage this trick of winning more, you will be 'bumped' into another bracket. 

 

     By the same token, all the players losing too much will be 'adjusted' down a bracket automatically.  In this case, the players may find themselves now playing others who are having the same level of clueless luck as they are.

 

B) All players will be pooled into the MM and then the MM will have to "Pick Teams" in a "balanced" manner wherein it picks the best two players based on this definition of "skill," then pick the next two and the next two and so forth such that fifteen players on each team are 'paired' with others of their overall (or within a bracket of 'skill' used to allow MM a bit of 'fudge factor' to complete its task) level of ability.

 

     In this scenario, NOTHING, not a damn thing, not one iota, not a molecule would be different than the current MM we now have overall.  You will still be pitted against players who stand a good chance of being much better than you or much worse than you in a match.  You cannot guaranty where on the map the best players are going to go or what tanks, tactics and tiers they will be in before they show up in the team compositions.  And, without XVM, you will be completely ignorant of how 'stacked' such a team may or may not be relative to the potential for you to meet the best player on the other team versus the worst player on the other team, combined with where you stand in that mix.

 

In case A, above, you could be having a bad streak and get bumped down a bracket, then do better, go up a bracket, slump, go down a bracket and so on and so forth, making you switch brackets every so many games.  Wouldn't this sort of thing be something you'd want to be notified of?  I know I would think it would be important enough to send an informational were it me that designed this system.  Wouldn't you want to know you went up or down a bracket?

 

Ultimately, though, how do you justify such an expenditure of effort in coding to get all this to work in what amounts to a Non-Competitive Environment.  The Argument: Playing for Fun is a very justifiably reasonable explanation for a NO to SBMM.  This is not meant to be that competitive.  If you want more competitive, then join a CLAN and play Strong Holds.  If you want more than that, Practice for League Play.

 

If you don't think it's fair to 'constantly' be pitted against more experienced players in higher tier vehicles, then this game is not for you.  This game is as fair as it gets in the way it treats the players.  There is RNG at +/- 25% to prevent the Better Players from Pawning the Bad Players all the time.  That RNG allows the 'watch is broken twice a day' level players to overcome a situation that, were it all Skill Based, would support that better player -- every time except twice a day.

 

With a wide and constantly tilting level of skill going between teams in play, the opportunity to rise and dominate exists much more readily for sub-par players than a skill based system could ever provide.  In a SBMM, you would ALWAYS be held to a potential that sees a pretty much fifty fifty outcome all the time.

 

And EVEN  in this case, fifty fifty does not mean it will work out that way.

 

People are not consistent from day to day, hour to hour or even meal to meal.  This game requires one's attention to a very high degree.  This is a "High Clutter" environment.  This means there is a wealth of data and a dearth of time in which to apply it.  Good players are going to be able to leverage this situation the majority of the time and -- because this game is a Single Death Per Match Zero Sum Game with Finite Resources -- the person who overcomes is the person who wins that engagement.  The moment you lose a tank the system is now UNBALANCED and thus, according to the dictates of a SBMM, someone should be substituted into the match of similar skill/ability -- except you can't.  A player out is a player out and that's the way it goes.

 

All that this means is that, even in a SBMM, the guy who is just ON FIRE or the player who seems to be ALL THUMBS that day cannot be adjudged except by that 'skill' definition that the SBMM will be using will not act in accordance with that defined level.  And now you have a bad match, because the guy with the rain cloud over his head is on your team and his skill level is BELOW his defined level, because he's pissed about traffic, that guy who cut him off, just had an argument with his wife/girlfriend, got fired/demoted/punished at work, etc.

 

None of these can ever be accounted for in a Random System, but it evens out over time.  In a SBMM System, this cannot really be factored out and smoothed over with anything else other than doing it over time.  The blowouts will still occur.  The snowballing loss is still going to occur.  The unexpected win/loss will still occur.

 

No matter what you end up using to define this 'skill' it will require more than one or two metric numbers taken as an aggregate and no single number will adequately cover the definition of what 'skill' even is in this game.  Due to people being somewhat fickle as a rule, the system cannot cover the variance that each person can display from day to day or even hour to hour.

 

In short, all that extra time spent coding could have been better spent for the game developers.  For the Players, all those tears, explanations, pounding of podiums/chests or exhortations of how this or that could be Fixed, could have been better spent by focusing on what they can do to better learn how the game mechanics work, how to apply those ideas to the techniques folks teach and to focus their mental efforts on asking instead, "What can I do to play better, do better, win more, hit more, etc.?"

 

The Random System, over time, promotes those who insist on playing their best and winning as much as they can and still allows those who are an overall net negative on their team's potential to win to be 'right twice a day' in the right circumstances.  RNG provides the 'fudge factor' that, as often as not punishes the good player more than it profits the poor player.  RNG allows the tomato to not be pasted in situations where no RNG would provide them a slicer and a masher.

 

No RNG benefits those with better hand/eye coordination.  No RNG benefits those who can read the battle flow on the Mini-map better than others.  No RNG provides benefits from those who can better anticipate their opponents on a consistent basis.  If you manage to fall into all three of these categories, no RNG would make you supreme amongst your peers.

 

SBMM ultimately will not change the outcomes of most matches.  It will not significantly increase playing time or make for more 'close' games.  It will not level the playing field in the sense that the competition level would be more even.  People are people and their quirks, natures and idiosyncrasies in a system that does not place people in the same place with the same people all the time cannot be fixed in this manner.  Even a SBMM cannot cover the wide variety that people can exhibit when they are still 'randomly' placed on teams in the sense that it's very likely you will:

    A) Not recognize a single player on either team from previous play

    B) Be placed with some folks who cannot even speak your language (but they're at the same overall skill level, so it's fine, right)

    C) See one, possibly two names you recognize in every ten or so matches

 

In short, your skill may be 'consistent' over time, but you are not over each day as a general rule in terms of winning or losing.  You cannot control how well you will perform each day, each time, but we all know that some days we're just 'not that into it' at that moment and our performances suffer.  Because you are not practicing regularly as part of a team that can be aggregated as an overall metric relative to other such teams, the SBMM is still just a crap shoot sorting out players as well as it can according to the principles and instructions it is given along with the players it has to sort from in the first place.

 

SBMM will not fix this situation and more time spent on it than already is a waste of time UNLESS you are interested in making WoT a MUCH MORE COMPETITIVE Environment than it now is.

 

GL, HF & HSYBF!

OvO

 

I can see that you have given this a lot of thought, much of which I agree with. This certainly wasn't meant to be a contentious post and I have no wish to be drawn into a SBMM debate this time, I will keep digging and maybe come up with something we can debate in earnest!.... :)

 



Mfezi #35 Posted Mar 02 2019 - 20:08

    Captain

  • -Players-
  • 28524 battles
  • 1,697
  • [SADD] SADD
  • Member since:
    04-05-2016

View PostBlazzy, on Mar 02 2019 - 20:35, said:

Define "pulling their weight"

 

Based on how this is structured you're looking for them to carry you every game, without actually saying it.  No 1 player should be required to pick up the slack of others.

 

You should concern yourself with personal gain before you worry about if a player is carrying you or not.

 

You are 100% correct except that I am only a very small fraction of the players represented by these stats and I don't expect anything from any player,

However, if players continue to do what the stats suggest, then the natural order of things will dictate that the skills of better players will make up for those of the less skilled players.

If you want to get all egotistical about it then let me be the first to grovel at your feet and say that we owe you a huge debt of gratitude!.... :)

 



_Tsavo_ #36 Posted Mar 02 2019 - 20:17

    Major

  • Beta Testers
  • 46222 battles
  • 19,710
  • [BRVE] BRVE
  • Member since:
    02-16-2011

View PostThePigSheFlies, on Mar 02 2019 - 10:58, said:

there are no skilled players.  only padders, lucky #$^'s and hackers.

 

Kappa

 

Oh, carp, he's on to me D:

Blazzy #37 Posted Mar 02 2019 - 20:21

    Captain

  • Players
  • 34961 battles
  • 1,000
  • Member since:
    03-05-2012

View PostMfezi, on Mar 02 2019 - 19:08, said:

 

You are 100% correct except that I am only a very small fraction of the players represented by these stats and I don't expect anything from any player,

However, if players continue to do what the stats suggest, then the natural order of things will dictate that the skills of better players will make up for those of the less skilled players.

If you want to get all egotistical about it then let me be the first to grovel at your feet and say that we owe you a huge debt of gratitude!.... :)

 

 

You have a weird way of bringing ego into things, especially since what I wrote had nothing to do with ego more-so to do with players seeing a purple on their team and thinking they'll get a win.

 

It works the same in reverse, they'll get paired up against too many good players and think its a for sure loss.

 

All I'm saying is stop worrying about stats and focus on personal gain.  Nobody started out a purple, but there are several records of players working towards it.  So making a post like this just shows you have no regard for personal improvement rather than ensuring you'll get carried if XYZ player is on your team.

 


Edited by Blazzy, Mar 02 2019 - 20:24.


Mfezi #38 Posted Mar 02 2019 - 20:32

    Captain

  • -Players-
  • 28524 battles
  • 1,697
  • [SADD] SADD
  • Member since:
    04-05-2016

View PostBlazzy, on Mar 02 2019 - 21:21, said:

 

You have a weird way of bringing ego into things, especially since what I wrote had nothing to do with ego more-so to do with players seeing a purple on their team and thinking they'll get a win.

 

It works the same in reverse, they'll get paired up against too many good players and think its a for sure loss.

 

All I'm saying is stop worrying about stats and focus on personal gain.  Nobody started out a purple, but there are several records of players working towards it.

 

 

You have a weird way of looking a a simple set of stats without any hypothesis or conclusion and then assuming that OP is asking to be carried. I can assure you that there is no ulterior motive nor hidden agenda in this post except that which you invent in your own mind!....

 



Butcher_Bay #39 Posted Mar 02 2019 - 22:30

    Sergeant

  • -Players-
  • 9824 battles
  • 209
  • Member since:
    01-09-2016
I really like this guy's posts.:)

Garandster #40 Posted Mar 02 2019 - 23:03

    Captain

  • -Players-
  • 6134 battles
  • 1,253
  • Member since:
    04-12-2018

View PostSnakePliskan, on Mar 02 2019 - 13:18, said:

 

There is NOTHING flawed and the purpose ( if your piss poor attitude can allow your mind to work) is a simple chart of what can easily help take a red player to a higher level ( color) performance. What does it take in terms of damage to move up? TAH DAH percentage of damage in relation to your HP. SIMPLE EASY not complicated.

 

WOW you must be a peach to be around. Here is a peanut butter sandwich. THAT'S not a peanut butter sandwich your information is WRONG FLAWED. There is no butter in it. I don't want it.

 

1. His data is bad, there are so many variable on what tank is used and what values are expected to actually say you are performing well. An SPG or TD needs a lot more damage in proportion to HP than other tank classes to get a decent WN8 (a standard measure of performance on an individual tank basis). Doing 800 damage in a tier 8 LT will get a much higher WN8 (color) than doing 800 in a tier 8 TD.

2.The sample size is WAY to small to draw any conclusions in any way.

3. Notice any inconsistancies with the data? You can't take the average of an average to get an answer. You can't show average damage and have a reliable number when certain groups are mostly only high tier games (Purple) versus yellow or red that have more games at lower tier where there is less HP resulting in lower average values.

4. Look at the player base he selected, it doesn't look to represent a cross section example of what the player base is comprised of currently if this is what the data is supposed to represent.

 

This is not the first time I have seen him present edited "data" on the forum to make a point.

 

 







Also tagged with stats

1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users