Jump to content


Bad Match Making, Too Many Blowout Games, Statistical Anomalies, And Quitting WoT


  • Please log in to reply
107 replies to this topic

Dragon6actual #61 Posted Feb 11 2019 - 02:50

    Corporal

  • Players
  • 15229 battles
  • 48
  • [PRIC] PRIC
  • Member since:
    03-01-2011
I forgot to mention... Webium used to provide (maybe they still, I don't know) at percentage on the initial team load page of the percent chance of winning and through 1869 games, those percentages were almost right on.  Within 1-3 percentage points, if I would have tracked more game I'm convinced the percentages would have only proven to be closer.

Edited by Dragon6actual, Feb 11 2019 - 02:53.


jdub11 #62 Posted Mar 11 2019 - 04:02

    Sergeant

  • Players
  • 23422 battles
  • 104
  • [-ZOO-] -ZOO-
  • Member since:
    05-06-2013

WG has a patent on their software which includes the design overview of the MM. If you seak it out, you can enlighten yourselves. In short, it is designed to take in account your recent battle history on several factors and it will create matches based on your previous games. There are other factors in WG MM build, but you can look it up yourselves as i cannot summarize the several pages of theory behind thier build in this quick discussion. 

They said recently that some of their parameters are outdated due to changes in average tiers played which is negitively effecting today's games. It has seamed to have gotten a bit better since February  in my eyes. Like maybe they quickly visited and made some simple adjustments. Lets hope they modernize the MM and balance the game soon. I won't hold my breath though. Have not played too much since January due to lack of WG's overhauling and... well, interest.



ixovera #63 Posted Mar 11 2019 - 13:39

    Sergeant

  • Players
  • 42781 battles
  • 138
  • [-_-] -_-
  • Member since:
    11-05-2011
I saw you do statistically better in T4/5/6 games than T7/8/9/10 games.

My suggestion is to find out what makes you good in the lower tier games, then find (and play) higher tier tanks that fit that profile. For me, I do better in highly mobile tanks (LTs, some MTs, and some HTs). I don't do well in most tanks that are slow. Another responder mentioned that if half your team drops the ball and dies early, it's pretty difficult to recover--and that's just the way it is (queue song). Perhaps that is why I do better in (and prefer) highly mobile tanks. They can switch flanks, retreat, recover, and take advantage of opportunities much more easily than slower tanks. That's what works for me. It may also work for you if you are finding the JPEG hard to win in. (The Type 5 is an exception because everyone is scared of getting derped in the face with a 50/50 chance of penning their shot.)

So, again, if you really enjoy playing (and are good in) tanks like the Cromwell or Hellcat or M41 LT, well those are highly mobile tanks and perhaps that is where you shine. If you are killer in something like the T-150 or the O-I, then hmm, work on angling more? A lot of times I experiment in games to get a good feel for a tank's strengths and weaknesses. Like, I intentionally take a hit at certain angles to see if I get penned or not so that I know how to angle that tank in future battles.
 
Anyway, just some thoughts. I hope it helps.

Bugs911 #64 Posted Mar 11 2019 - 18:00

    Corporal

  • Beta Testers
  • 28281 battles
  • 12
  • [ROBOT] ROBOT
  • Member since:
    09-28-2010

I 100% agree with this thread. Even to the point of considering quitting the game after 8 years. After these 8 to 15 games in a row, blowouts (to me) the factors are obvious. Overall team avg's on win% and avg dmg is drastic. Somehow these factors have to go into matchmaking if you want players to feel they are getting a "fair" game most of the time.

 

Its not the losing. It is the losing 12 in a row and the the other team has 10 or more tanks left. The blowouts are the issue. I can guess right most of the time without even using a mod. This is simply by the clan stacked teams that show up these days. Unfair from the start. 



PhilthyOne1 #65 Posted Mar 11 2019 - 22:16

    Corporal

  • Players
  • 34214 battles
  • 28
  • Member since:
    04-13-2013

View PostPipinghot, on Jan 07 2019 - 04:33, said:

No, tinfoil hatted conspiracy theorists don't understand randomness and game play. Everyone who correctly understands how randomness works, and how win/loss strings work, also understand that the game is working just as it should. Players can either choose to take responsibility for their game play, or they can blame the game because their egos are too weak.

 

 

I have a very clear understanding of randomness.  You are ignoring quite a bit of my posting.  So once again, while it is possible to flip a coin and have strings of heads and tails, such as HHHHHHTTTTTTHHHHHHHHTTTTTTTT, this "string" is not likely.  Yes, I could flip a coin 10 times in a row and each time it comes up heads.  The odds of it being heads again on the 11th flip is still 50/50.  But the odds of 11 heads in a row are 1 in 2,048.  So "strings" are not normal.  Possible, yes, but not likely.  And I do take responsibility for my game play.  I don't blame the team because I've seen too often complete blowouts and during blowouts accuracy and pen changes greatly.  Either I hit and pen everything with snap shots, or nothing I do seems to matter.  I rarely take snap shots - very rarely for that matter.  I almost always dial into 100% before shooting, so errant shots are VERY noticeable.

I'm not wearing a tinfoil hat...  I'm making an observation here that I believe is backed by my analysis of tracking multiple games and my calculations demonstrate a highly unlikely series of events.  Yes, possible, but highly unlikely.



BayouPhantom #66 Posted Mar 12 2019 - 06:46

    Corporal

  • -Players-
  • 21 battles
  • 56
  • Member since:
    02-25-2019

View PostPhilthyOne1, on Jan 06 2019 - 08:03, said:

 

I have a 54% win rate in most tanks and in the last year I've have these frequent runs of 7-8 losses in a row, one win, then another 5-6 losses in a row.  Same for wins, several in a row, then a loss, then several more wins in a row. 

 

I was right there with you until you said you get several wins in a row, then a loss, then several more wins in row.  I would love to experience that on occasion.  The strings of losses?  Sure, a lot.  But strings of wins?  I think 4 is tops in a long time, and then it is followed by a freight train of losses.  When it comes to math and statistics, I am NOT smarter than a 5th grader, so I don't know if it is "statistically" wrong or not, but I do know that it is no fun to just get loss after loss on a regular basis.  It isn't ALL me either, there's an entire team losing and the player's change every game so there has to be something in the selection of the players and vehicles that MM makes.  I don't know.  Just lost one match 15-0.  First of those in a while, but certainly not the first lopsided loss.  I took almost a year off and played Warships.  I'll go back to it soon and visit tanks to see if anything has gotten better.  I guess not enough people play Warships for WG to sink too much into messing it up.

Jer1413 #67 Posted Mar 12 2019 - 15:04

    Captain

  • Players
  • 47072 battles
  • 1,442
  • [RR13] RR13
  • Member since:
    02-24-2013

View PostPhilthyOne1, on Mar 11 2019 - 21:16, said:

 

I have a very clear understanding of randomness.  You are ignoring quite a bit of my posting.  So once again, while it is possible to flip a coin and have strings of heads and tails, such as HHHHHHTTTTTTHHHHHHHHTTTTTTTT, this "string" is not likely. It's actually just as likely as any other particular combination.

 

 Yes, I could flip a coin 10 times in a row and each time it comes up heads.  The odds of it being heads again on the 11th flip is still 50/50.  But the odds of 11 heads in a row are 1 in 2,048.  So "strings" are not normal.  Possible, yes, but not likely. You have actually played over 31,500 "strings" of 11 games, so it is quite likely. By your math, you should have expected about 15 eleven game streaks so far in your WOT career.

 

And I do take responsibility for my game play.  I don't blame the team because I've seen too often complete blowouts and during blowouts accuracy and pen changes greatly.  Either I hit and pen everything with snap shots, or nothing I do seems to matter.  I rarely take snap shots - very rarely for that matter.  I almost always dial into 100% before shooting, so errant shots are VERY noticeable.

I'm not wearing a tinfoil hat...  I'm making an observation here that I believe is backed by my analysis of tracking multiple games and my calculations demonstrate a highly unlikely series of events.  Yes, possible, but highly unlikely. Again, all very likely given the number of games you have played.

 

 

 



Copacetic #68 Posted Mar 12 2019 - 17:00

    Captain

  • -Players-
  • 45980 battles
  • 1,229
  • [ZEUS] ZEUS
  • Member since:
    02-04-2014

View PostPhilthyOne1, on Mar 11 2019 - 22:16, said:

 

I have a very clear understanding of randomness.  You are ignoring quite a bit of my posting.  So once again, while it is possible to flip a coin and have strings of heads and tails, such as HHHHHHTTTTTTHHHHHHHHTTTTTTTT, this "string" is not likely.  Yes, I could flip a coin 10 times in a row and each time it comes up heads.  The odds of it being heads again on the 11th flip is still 50/50.  But the odds of 11 heads in a row are 1 in 2,048.  So "strings" are not normal.  Possible, yes, but not likely.  And I do take responsibility for my game play.  I don't blame the team because I've seen too often complete blowouts and during blowouts accuracy and pen changes greatly.  Either I hit and pen everything with snap shots, or nothing I do seems to matter.  I rarely take snap shots - very rarely for that matter.  I almost always dial into 100% before shooting, so errant shots are VERY noticeable.

I'm not wearing a tinfoil hat...  I'm making an observation here that I believe is backed by my analysis of tracking multiple games and my calculations demonstrate a highly unlikely series of events.  Yes, possible, but highly unlikely.

 

can we see your data sir

Nunya_000 #69 Posted Mar 12 2019 - 17:24

    Major

  • Players
  • 21138 battles
  • 13,771
  • [PACNW] PACNW
  • Member since:
    09-20-2013

View PostPhilthyOne1, on Mar 11 2019 - 13:16, said:

 

I have a very clear understanding of randomness.  You are ignoring quite a bit of my posting.  So once again, while it is possible to flip a coin and have strings of heads and tails, such as HHHHHHTTTTTTHHHHHHHHTTTTTTTT, this "string" is not likely.  Yes, I could flip a coin 10 times in a row and each time it comes up heads.  The odds of it being heads again on the 11th flip is still 50/50.  But the odds of 11 heads in a row are 1 in 2,048.  So "strings" are not normal.  Possible, yes, but not likely.  And I do take responsibility for my game play.  I don't blame the team because I've seen too often complete blowouts and during blowouts accuracy and pen changes greatly.  Either I hit and pen everything with snap shots, or nothing I do seems to matter.  I rarely take snap shots - very rarely for that matter.  I almost always dial into 100% before shooting, so errant shots are VERY noticeable.

I'm not wearing a tinfoil hat...  I'm making an observation here that I believe is backed by my analysis of tracking multiple games and my calculations demonstrate a highly unlikely series of events.  Yes, possible, but highly unlikely.

 

Comparing the results of coin flipping to the win/loss results of battles is stupid and is like comparing apples to oranges.  While the make up of teams is random, the results of the battles are not random, as the performance of the players on each team influences the outcome.

 

The major random (variable) components in a coin flip are the placement of the coin, the height difference between the flipping mechanism and the landing area, and the force used in the flip.  If these are exactly equal in every single flip, the results will also likely be equal.  Meaning that it is possible to get 100 heads in 100 flips.

 

In WOT, there are many different variables in just player's performance, reactions, and skills that will influence battles...just to name a few.  There are also 30 different players involved that apply these variables.  Take 2 teams with the same exact players on both teams and play 100 battles on the same map.  It is very likely that the battles will unfold in 100 different ways.



Dragon6actual #70 Posted Mar 12 2019 - 17:50

    Corporal

  • Players
  • 15229 battles
  • 48
  • [PRIC] PRIC
  • Member since:
    03-01-2011

View PostPhilthyOne1, on Mar 11 2019 - 16:16, said:

 

I have a very clear understanding of randomness.  You are ignoring quite a bit of my posting.  So once again, while it is possible to flip a coin and have strings of heads and tails, such as HHHHHHTTTTTTHHHHHHHHTTTTTTTT, this "string" is not likely.  Yes, I could flip a coin 10 times in a row and each time it comes up heads.  The odds of it being heads again on the 11th flip is still 50/50.  But the odds of 11 heads in a row are 1 in 2,048.  So "strings" are not normal.  Possible, yes, but not likely.  And I do take responsibility for my game play.  I don't blame the team because I've seen too often complete blowouts and during blowouts accuracy and pen changes greatly.  Either I hit and pen everything with snap shots, or nothing I do seems to matter.  I rarely take snap shots - very rarely for that matter.  I almost always dial into 100% before shooting, so errant shots are VERY noticeable.

I'm not wearing a tinfoil hat...  I'm making an observation here that I believe is backed by my analysis of tracking multiple games and my calculations demonstrate a highly unlikely series of events.  Yes, possible, but highly unlikely.

 

I think the person I quoted above did an excellent job of explaining their point.  I do not believe he is saying the game results are decided by a coin toss but the game results should look more like a series of coin tosses... more random, less long streaks. This is plain and simple common knowledge.  Ignoring the logic is wrong.  The win/lose cycle is there.  When I did my study on 1869 games it was very obvious the losing streaks and winning streaks are part of the game.  In fact proven over and over during my study the outcome of game IS ALWAYS decided by the quality of the teams.  Players don't play good for 10-12 games and then just suddenly forget how to play.  I think the desired goal is not to win all the time but 50/50 is okay just more random 2 & 2, 3 & 3 not winning 12 out of 15 only to lose 12 out of 15 (or something along those lines).   I use 12 out of 15 because that seemed to be the statistical norm throughout the 1869 game study.  the longest win streak I had was 14 games using the tier V Churchill III, the longest losing streak was 12 games in the same tank, several hundred games apart.  Of course your performance has an influence on the outcome of every game but I've had games (during the study) where I killed 5 tanks and lost.  Of course I've played numerous games where I didn't kill anybody and won.  I've seen games where players killed 14 tanks and I'm not talking about those great players but the average players... your team makeup is the greatest variable in the game... PERIOD!   And what controls the team makeups?   The Match-maker.

 

I have the data on an XP version of EXCEL and I would love to share it but the file is HUGE!!!!!!



Dragon6actual #71 Posted Mar 12 2019 - 18:25

    Corporal

  • Players
  • 15229 battles
  • 48
  • [PRIC] PRIC
  • Member since:
    03-01-2011

View PostNunya_000, on Mar 12 2019 - 11:24, said:

 

Comparing the results of coin flipping to the win/loss results of battles is stupid and is like comparing apples to oranges.  While the make up of teams is random, the results of the battles are not random, as the performance of the players on each team influences the outcome.

 

The major random (variable) components in a coin flip are the placement of the coin, the height difference between the flipping mechanism and the landing area, and the force used in the flip.  If these are exactly equal in every single flip, the results will also likely be equal.  Meaning that it is possible to get 100 heads in 100 flips.

 

In WOT, there are many different variables in just player's performance, reactions, and skills that will influence battles...just to name a few.  There are also 30 different players involved that apply these variables.  Take 2 teams with the same exact players on both teams and play 100 battles on the same map.  It is very likely that the battles will unfold in 100 different ways.

 

Your very last sentence... "It is very likely that the battles will unfold in 100 different ways".   I couldn't agree with you more BUT while the games will play out 100 different ways, only THREE possible outcomes can occur... Team A wins, Team B wins or there is a draw.  What none of us can predict is what percentage of games Team A will win.  I might be speaking for the original author but what's at issue is not how many times Team A wins but how many times that team wins in a streak.  On the surface having the same (evenly balanced) teams play 100 different games, one would expect small streaks (3 to 5) to occur, maybe even slightly longer streaks (6 to 8) but if the teams were evenly balanced longer streaks would be VERY unlikely!  What others have pointed out is not that streaks occur but winning streaks seem to be followed immediately buy losing streaks or vise versa.  Of course that can happen but it shouldn't happen with any regularity.  Which is does... apparently by a significant amount of OBSERVANT players.  Play 10-20 battles a day and you probably won't see the big picture but start keeping track of wins/losses, tank type, top, middle, bottom tier etc... and you start to see patterns.  I'm assuming what everyone would like to see is more evenly balanced teams and let the chips fly!  My guess is streaks will still happen but more small streaks! IMHO!!!!!!!!!

Nunya_000 #72 Posted Mar 12 2019 - 18:28

    Major

  • Players
  • 21138 battles
  • 13,771
  • [PACNW] PACNW
  • Member since:
    09-20-2013

View PostDragon6actual, on Mar 12 2019 - 08:50, said:

 

I think the person I quoted above did an excellent job of explaining their point.  I do not believe he is saying the game results are decided by a coin toss but the game results should look more like a series of coin tosses... more random, less long streaks. This is plain and simple common knowledge.  Ignoring the logic is wrong.  

 

And right there is your fallacy.  No, the results of an individual's win/losses in this game SHOULD NOT look like a series of coin tosses.  If you feel this is "common knowledge" or logical", then you do not fully understand probability and statistics (no offense intended).  It is also not true that most players play the same in one series of battle as they do in another series of battles.....heck, most players do not play consistently from one single battle to the next.  We all take chances while playing this game.  Sometimes those chances pay off....sometimes they do not.

 

While, I do agree that if teams are unbalanced skillwise, it will often "pre-determine" what the outcome MIGHT be, it does not determine what the outcome is GUARANTEED to be.  You claim that MM "controls" the team make-up (I assume you mean in regard to skill level), but that is also a fallacy.  Skill is not considered by MM, so there is NO control.  The lack of consideration of player skill does not correlate to it being "controlled".

 

Players like to claim that "streaks" are proof that wins and loses are "rigged", or "controlled", or "manipulated".  In fact, it is just the opposite.  The lack of streaks would be an indication of these things.  If the developers are good enough to "pre-determine" the wins and loses of 30 player per battle, then one would think they were good enough to manipulate the game so that any one player would have very few streaks.  This, of course, would be a much better system as it would alienate fewer players.  Win a battle....lose the next one.....etc.  Less frustration would cause player to be more satisfied while playing.



dunniteowl #73 Posted Mar 12 2019 - 18:46

    Major

  • -Players-
  • 30185 battles
  • 7,542
  • Member since:
    09-01-2014

Coin Toss Variables:

 

1) Only Two Sides to the coin are usable

 

2) If a Person Does It:

     a) Placement of Coin On Toss (potential of both a) and b) to have a greater or lesser influence on toss)

     b) Force of Toss (height of coin toss)

     c) Angular Momentum Applied (speed of coin flip)

     d) Landing Zone

 

3) If Done By Machine:

     a) Placement of Coin on Toss (potential of both a) and b) to have a greater or lesser influence on toss)

     b) Force of Toss (is machine consistent or variable in amount of force applied)

     c) Angular Momentum Applied (is machine consistent or variable in angular momentum applied)

     d) Landing Zone

 

In 2) Above, the forces of position, force and angular momentum will be less stable than if a machine did it, however, in most cases, the ability to purposely apply as much NON RANDOM effort as possible can result in repeated coin landings that belie statistical 'random' behavior in streaks.  That said, a person cannot be 100% consistent in those a, b, c and d variables all the time and, thus, we see the occasional string of tails, over heads or a topsy turvey sort of 'apparent' randomness (which it actually is, however, we're not going there for the moment) that appears Random is something that doesn't have a lot of long streaks to it.

 

In 3) Above, it is quite possible that the only Real Variable could be the Landing Zone.  If you place a machine into the position of tossing the coin, you're going to want a little holder for the coin, so it doesn't fall off (rigging).  You are also going to want the machine to flip with the same amount of force each time (rigging).  Lastly, due to the positioning and force being consistent, the angular moment will now be set (rigging).  The only variable you cannot change is the Landing Zone and how the coin responds.  Even so, you're going to be able to heavily influence the coin toss in such a way (called a Control in science) that eliminates much of the random elements of said coin toss and will likely see LONG and Streaky behavior overall, because a machine can flip such a coin in this instance without very much that is random at all.

 

 

Now.  Let's talk about a 15 v 15 person team's variables

 

1) Skill Level of Each Player (30 separate variables ALL BY ITSELF)

     a) is the player consistently performing (x30)

     b) is the player a high skill or low skill player (x30)

     c) is the player Hot or Cold right now (x30)

 

2) Map the game is being played (x35)

 

3) Spawn Point for Players (x6, due to Encounter, Assault and Standard Spawn differences, doubled by side of map spawned)

 

4) Player Comfort Level Per Map (x60, due to players being more able from one side of a map than another, x6 for the above spawn points as well)

 

5) Team Composition (x30 as a MINIMUM, though with over 400 different units, well you can see that the math will be hard pressed to stably account for all that and this is just as much 'determined' by the players choosing units and then MM placing them into a match in their specific queues)

 

 

From the above it is easy to see that comparing this game to a coin toss is LUDICROUS to the point of being laughable.

 

In the actual game system we also have:

Single Death Per Battle

 

Zero Sum Game (both sides simply cannot achieve Victory, either one side or the other is a Victor or Both Sides Lose in a Draw)

 

Technical Issues (some players are using different computers with different connections and different system requirements being met, resulting in a plethora of variables impossible to ascertain to any reliable degree)

 

Player Variables:

1) Distraction Level

2) Attention Ability (tied to distraction level, though not necessarily by outside forces to the player)

3) Health

4) Overall Preparedness to Play (again, tied to distraction and health, though not necessarily tied only to external forces, one could be hungry, tired, pissed off, stressed out, etc)

 

 

You add all that up and you have RANDOM in Spades.

 

In that Random in Spades you are going to occasionally see Large Streaks.  Depending on one's personal skill level and other factors, those streaks will be shorter or longer than 'statistically' likely, especially as compared to a simple and low random variable situation as a Coin Toss.  You will see the occasional 16 game Winning Streak (my top so far) and the occasional 13 game Losing Streak (also my top so far).  And there will be plenty of times you win one, lose three, win two, lose one, win four, lose five, win one, lose three, win four, lose two and so forth mixed into that.

 

 

What Random is NOT:

Something you can expect a 50% chance of losing or winning at any given moment in time.

Something you can reliably predict the outcomes to, no matter how smart or experienced you are, sometimes you'll be wrong.

Something you can say, "Statistically Impossible" to in terms of outcomes.

 

In Random, it is statistically UNLIKELY that certain events will occur, however, those events CAN STILL OCCUR, even if you don't like the results.

 

This is what the OP is dealing with and for all those who "agree 100% with OP" in his feels.

 

It is STILL RANDOM and STILL POSSIBLE to OCCUR.

 

 

My work here is dunn.

 

 

GL, HF & HSYBF!

OvO

 


Edited by dunniteowl, Mar 12 2019 - 18:49.


Nunya_000 #74 Posted Mar 12 2019 - 18:48

    Major

  • Players
  • 21138 battles
  • 13,771
  • [PACNW] PACNW
  • Member since:
    09-20-2013

View PostDragon6actual, on Mar 12 2019 - 09:25, said:

 

Your very last sentence... "It is very likely that the battles will unfold in 100 different ways".   I couldn't agree with you more BUT while the games will play out 100 different ways, only THREE possible outcomes can occur... Team A wins, Team B wins or there is a draw.  What none of us can predict is what percentage of games Team A will win.  I might be speaking for the original author but what's at issue is not how many times Team A wins but how many times that team wins in a streak.  On the surface having the same (evenly balanced) teams play 100 different games, one would expect small streaks (3 to 5) to occur, maybe even slightly longer streaks (6 to 8) but if the teams were evenly balanced longer streaks would be VERY unlikely!  

 

If the tanks of all 30 players were computer controlled and act predictably, then I would agree with your thesis.  However, that is not the case.  They are human controlled and humans tend to act unpredictably.  Add in the fact that there are 15 human controlled tanks per team....all acting unpredictably....and not in conjunction with one another, the results of each battle CAN and WOULD cause streaks....and many times create instances of losing streaks followed by winning streaks....and vise verse.

 

The issue is that it is human nature to notice these streaks (or patterns) much more than the days where they have few streaks.  It is called Apophenia.  Player see these "patterns" and believe they are very common....and ignore the number of times where they have had alternating sessions where there were only a few "streaks" above 2-3 wins or losses.....or they will discount the break in the streaks (i.e. the win after 3 losses, followed by 3 more losses and make them believe they had a 6 game losing streak).

 

Block Quote

What others have pointed out is not that streaks occur but winning streaks seem to be followed immediately buy losing streaks or vise versa.  Of course that can happen but it shouldn't happen with any regularity.  Which is does... apparently by a significant amount of OBSERVANT players.  Play 10-20 battles a day and you probably won't see the big picture but start keeping track of wins/losses, tank type, top, middle, bottom tier etc... and you start to see patterns.  I'm assuming what everyone would like to see is more evenly balanced teams and let the chips fly!  My guess is streaks will still happen but more small streaks! IMHO!!!!!!!!!

 

Poorer players are more likely to have more losing streaks than better player will.  That is because they tend to handicap their team.  Better players will likely have more winning streaks than poorer players.  That is because they tend to help their team more.  However, while individual performance is important, the performance of the team AS A WHOLE will ultimately decide the outcome of a battle.  A player can have a great game, but if there is too much "dead weight" on their team (i.e. too many players not doing their HP in damage), that great game will not make up for the players that perform badly. 

 

This game is also a game of attrition.  The team that gains an advantage in the number of guns still in the battle will often maintain that advantage to the end. 

Question:  You are fighting a battle and you see the score is 6-9 (6 red tanks killed vs 9 green tanks killed)....what goes through your mind?  Do you think your team still has a good chance of winning?  While there is still a chance, more often than naught, it will not end well for the green team.

 

 



Dragon6actual #75 Posted Mar 12 2019 - 18:58

    Corporal

  • Players
  • 15229 battles
  • 48
  • [PRIC] PRIC
  • Member since:
    03-01-2011

View PostNunya_000, on Mar 12 2019 - 12:48, said:

 

If the tanks of all 30 players were computer controlled and act predictably, then I would agree with your thesis.  However, that is not the case.  They are human controlled and humans tend to act unpredictably.  Add in the fact that there are 15 human controlled tanks per team....all acting unpredictably....and not in conjunction with one another, the results of each battle CAN and WOULD cause streaks....and many times crate instances of losing streaks followed by winning streaks....and vise verse.

 

The issue is that it is human nature to notice these streaks (or patterns) much more than the days where they have few streaks.  It is called Apophenia.  Player see these "patterns" and believe they are very common....and ignore the number of times where they have had alternating sessions where there were only a few "streaks" above 2-3 wins or losses.....or they will discount the break in the streaks (i.e. the win after 3 losses, followed by 3 more losses and make them believe they had a 6 game losing streak).

 

 

Poorer players are more likely to have more losing streaks than better player will.  That is because they tend to handicap their team.  Better players will likely have more winning streaks than poorer players.  That is because they tend to help their team more.  However, while individual performance is important, the performance of the team AS A WHOLE will ultimately decide the outcome of a battle.  A player can have a great game, but if there is too much "dead weight" on their team (i.e. too many players not doing their HP in damage), that great game will not make up for the players that perform badly. 

 

This game is also a game of attrition.  The team that gains an advantage in the number of guns still in the battle will often maintain that advantage to the end. 

Question:  You are fighting a battle and you see the score is 6-9 (6 red tanks killed vs 9 green tanks killed)....what goes through your mind?  Do you think your team still has a good chance of winning?  While there is still a chance, more often than naught, it will not end well for the green team.

 

 

 

I'm not sure how to reply!!!!  You are saying what I'm saying but in a different way but saying I'm incorrect.  You know what, just forget everything I said.... I'm wrong and you're right.  Feel better!  :-)

Nunya_000 #76 Posted Mar 12 2019 - 19:22

    Major

  • Players
  • 21138 battles
  • 13,771
  • [PACNW] PACNW
  • Member since:
    09-20-2013

View PostDragon6actual, on Mar 12 2019 - 09:58, said:

 

I'm not sure how to reply!!!!  You are saying what I'm saying but in a different way but saying I'm incorrect.  You know what, just forget everything I said.... I'm wrong and you're right.  Feel better!  :-)

 

Except we are not saying the same thing.

 

You: "There should not be streaks in this game....or at least as regularly as we see them".

 

Me: "I would expect streaks because of the nature of the game and the variables that 30 human players per battle introduce to the mechanics of the game".



PhilthyOne1 #77 Posted Mar 12 2019 - 22:45

    Corporal

  • Players
  • 34214 battles
  • 28
  • Member since:
    04-13-2013

View Postdunniteowl, on Mar 12 2019 - 17:46, said:

Coin Toss Variables:

 

1) Only Two Sides to the coin are usable

 

2) If a Person Does It:

     a) Placement of Coin On Toss (potential of both a) and b) to have a greater or lesser influence on toss)

     b) Force of Toss (height of coin toss)

     c) Angular Momentum Applied (speed of coin flip)

     d) Landing Zone

 

3) If Done By Machine:

     a) Placement of Coin on Toss (potential of both a) and b) to have a greater or lesser influence on toss)

     b) Force of Toss (is machine consistent or variable in amount of force applied)

     c) Angular Momentum Applied (is machine consistent or variable in angular momentum applied)

     d) Landing Zone

.......................  This goes on and on...............................  Not quoting the whole thing...

 

 

I applaud your effort.  You are missing many, many variables though.  For the coin, air temperature, humidity, temperature of the coin, temperature of the ground, barometric pressure, local gravity, how many atoms fly off the coin after you flip it as each flip it will lose some material due to abrasion.  Then there are the players, type of PC, mouse, internet connection, monitor size, mouse surface, keyboard reaction time, and another million variables.  All of these will affect the outcome - agreed. 

 

I was not trying to compare the game and coin flipping.  My point in comparing the two is in only stating that the goal is to have 50/50 outcomes, outcomes that are evenly balanced.  If you played 30,000 games and have a 50% win rate, that means, on average, you win 50% of your games.  If you flip a coin 30,000 times and it lands on heads 15,000 times, it's a reliable coin.  Based on the number of games played, 50% is a pretty reliable percentage for outcomes.  So if you win 10 games in a row, that's as if you flipped a coin 10 times and it landed on heads each time.  The odds of that happening are 2^10 or 1:1024.  Not saying it can't happen, just that it's not as likely as a string of 2-3 wins in a row followed by a few losses in a row.



PhilthyOne1 #78 Posted Mar 12 2019 - 23:04

    Corporal

  • Players
  • 34214 battles
  • 28
  • Member since:
    04-13-2013

You guys are all crazy talking about all these goofy variables that account for more fair play.  You're missing the point.  WoT should try to achieve a more balanced play.  If I played the lottery and win it five times in a row, while that is possible to happen, it is so unlikely that I'd probably be brought up under a very serious investigation.  Many of you guys seem to have a "meh" opinion, would shrug your shoulders and say all the losers are just crying about their loss.  I'd say it should be investigated and questioned.

I have a 54% win rate.  I don't clan and never team up, so I'm solo.  I know I'm a good player and my stats prove that out.  I win more games than I lose.  The problem I'm perceptive to is that there are strings of wins and losses (and these are blowout losses and wins) combined with the fact that my accuracy seems to increase/decrease respective to the wins and losses, leads me to believe WoT is playing with RNG.  I don't run mods, but after a blowout I'll look at player win%'s for each team.  I'll see an obvious imbalance in win percentages.  One team will have many players with many games in the 47% range, and the other has many players with many games played in the 53% range.

 

In a well balanced match, a blowout win/loss should be the exception, not the norm.  WoT should be working to create more close games as those are the best win or lose.  Blowout win/loss is no fun.



Nunya_000 #79 Posted Mar 12 2019 - 23:06

    Major

  • Players
  • 21138 battles
  • 13,771
  • [PACNW] PACNW
  • Member since:
    09-20-2013

View PostPhilthyOne1, on Mar 12 2019 - 13:45, said:

For the coin, air temperature, humidity, temperature of the coin, temperature of the ground, barometric pressure, local gravity, how many atoms fly off the coin after you flip it as each flip it will lose some material due to abrasion.  

 

All of that is only relevant if it impacts the rotation and changes the number of rotations of the coin during the experiment.  all of those things are very minor.  It might impact an experiment over a number of days (or months)...while the whole time the coin is being flipped, but for an experiment of 100 flips, they will not impact the experiment.

 

This is actually a physic experiment that I participated in during college.  We were assigned the task of building a machine to flip a coin where we would get the same result in 50 flips.  We hit 49 out of 50 as heads.  The 1 tails was a result of the table the machine was sitting on having a failure.

 

Block Quote

I was not trying to compare the game and coin flipping.  My point in comparing the two is in only stating that the goal is to have 50/50 outcomes, outcomes that are evenly balanced.  If you played 30,000 games and have a 50% win rate, that means, on average, you win 50% of your games.  If you flip a coin 30,000 times and it lands on heads 15,000 times, it's a reliable coin.  Based on the number of games played, 50% is a pretty reliable percentage for outcomes.  So if you win 10 games in a row, that's as if you flipped a coin 10 times and it landed on heads each time.  The odds of that happening are 2^10 or 1:1024.  Not saying it can't happen, just that it's not as likely as a string of 2-3 wins in a row followed by a few losses in a row.

 

You were doing so well until you posted the bolded part.  NO!!!!  winning 10 battles in a row does not correlate to flipping a coin 10 times and landing on heads each time.....in no way, shape, or form.  Comparing the odds of such to wins/losses in WOT is a complete and total fail.

 



PhilthyOne1 #80 Posted Mar 12 2019 - 23:23

    Corporal

  • Players
  • 34214 battles
  • 28
  • Member since:
    04-13-2013

View PostNunya_000, on Mar 12 2019 - 22:06, said:

 

All of that is only relevant if it impacts the rotation and changes the number of rotations of the coin during the experiment.  all of those things are very minor.  It might impact an experiment over a number of days (or months)...while the whole time the coin is being flipped, but for an experiment of 100 flips, they will not impact the experiment.

 

This is actually a physic experiment that I participated in during college.  We were assigned the task of building a machine to flip a coin where we would get the same result in 50 flips.  We hit 49 out of 50 as heads.  The 1 tails was a result of the table the machine was sitting on having a failure.

 

 

You were doing so well until you posted the bolded part.  NO!!!!  winning 10 battles in a row does not correlate to flipping a coin 10 times and landing on heads each time.....in no way, shape, or form.  Comparing the odds of such to wins/losses in WOT is a complete and total fail.

 

Actually, you are quite wrong.  If both events have a 50% chance of happening over a large sample set (30,000 samples in this case), then they are equally likely to occur.  If you played 30,000 games and had a 50% win rate, would you say your odds of winning a single game are 50/50.  Let's just agree on that - can we?  Now, we can go into all kinds of variables such as the other team getting stacked with a clan of experts, and that will affect that single game.  The same odds exist that in the next game that same clan of experts will be on the other team.  It balances out over time.  But as an average, before the game starts and teams are drawn, you should have a 50% chance of winning that game.  Again, BEFORE the game starts.  That's the disconnect.  You guys are computing in variables about game play that will affect a single game, but has no effect on averages as all the variables you bring into play have an equal chance for either team over time.

If you lose 10 games in a row, then perhaps WoT stacked the teams against you for those 10 games or played around with RNG for those 10 games.  You might accept that as just chance luck, but if it was happening consistently, then you must accept the fact that it is an unlikely outcome.  Still possible, but unlikely.  And the more unlikely it becomes, the more obvious it becomes that something is not quite right and the thought of rigging comes to mind.






2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users