Jump to content


Fast battles


  • Please log in to reply
22 replies to this topic

Geoffrey_Ironfist #1 Posted 11 August 2018 - 05:22 PM

    Lance-corporal

  • Player
  • 4623 battles
  • 58
  • Member since:
    06-23-2018
Some battles are very fast, with one side overwhelming the other in 3-4 minutes. Occasionally, however, I get quite a few of these in a row and start feeling paranoid when it is always the other team that is winning them. Are they after me?

Babbet_1 #2 Posted 11 August 2018 - 05:42 PM

    Warrant Officer

  • Player
  • 10572 battles
  • 722
  • [WJDE] WJDE
  • Member since:
    12-13-2015

View PostGeoffrey_Ironfist, on 11 August 2018 - 05:22 PM, said:

Some battles are very fast, with one side overwhelming the other in 3-4 minutes. Occasionally, however, I get quite a few of these in a row and start feeling paranoid when it is always the other team that is winning them. Are they after me?

 

Are who after you?  :sceptic:

wobblybob #3 Posted 11 August 2018 - 05:51 PM

    Staff Sergeant

  • Player
  • 64967 battles
  • 453
  • [TNAC] TNAC
  • Member since:
    07-23-2011
The game is notorious for winning and losing streaks. There are endless posts on the topic. Whether it is random chance or coded in to the game engine is a matter of opinion but you will certainly encounter them. There are mods that you can load that keep track of daily performance but rule of thumb is if you lose 5 in a row take a break.

Babbet_1 #4 Posted 11 August 2018 - 06:01 PM

    Warrant Officer

  • Player
  • 10572 battles
  • 722
  • [WJDE] WJDE
  • Member since:
    12-13-2015

View Postwobblybob, on 11 August 2018 - 05:51 PM, said:

 rule of thumb is if you lose 5 in a row take a break.

 

But then how can you ever hope to achieve a record run of losses? - Mine stands at 11 at the moment, but I expect to improve on that soon!  :rolleyes:

Geoffrey_Ironfist #5 Posted 11 August 2018 - 07:02 PM

    Lance-corporal

  • Player
  • 4623 battles
  • 58
  • Member since:
    06-23-2018

View PostBabbet_1, on 11 August 2018 - 06:01 PM, said:

 

But then how can you ever hope to achieve a record run of losses? - Mine stands at 11 at the moment, but I expect to improve on that soon!  :rolleyes:

 

Oh, that's close. I think I had a streak of 11 loses in my first 12 games two days ago.

 

EDIT: I just quit for today. How can half the enemy team be behind our lines less than 2 minutes into the game? The game before I was holding up a few enemy tanks on one side of the map with just two other guys, thinking, that will give time to the rest of the team to get behind them, and there were more enemy tanks coming in and more, some from the front, others from up on the bridge, from the hills, from behind, from the sides - and I realised the rest of our team had just died. Maybe I get teamed up with folks who have ordered a pizza?

 

EDIT 2: Just started playing again today and in a team with 3 SPGs, all the other tanks were dead in 3 minutes having killed between them just one enemy tank. Twelve tanks killed just one enemy tank, that's right.


Edited by Geoffrey_Ironfist, 12 August 2018 - 03:58 PM.


Baldrickk #6 Posted 12 August 2018 - 10:49 PM

    Field Marshal

  • Player
  • 29530 battles
  • 13,766
  • [-TAH-] -TAH-
  • Member since:
    03-03-2013

View PostGeoffrey_Ironfist, on 11 August 2018 - 07:02 PM, said:

 

Oh, that's close. I think I had a streak of 11 loses in my first 12 games two days ago.

 

EDIT: I just quit for today. How can half the enemy team be behind our lines less than 2 minutes into the game? The game before I was holding up a few enemy tanks on one side of the map with just two other guys, thinking, that will give time to the rest of the team to get behind them, and there were more enemy tanks coming in and more, some from the front, others from up on the bridge, from the hills, from behind, from the sides - and I realised the rest of our team had just died. Maybe I get teamed up with folks who have ordered a pizza?

 

EDIT 2: Just started playing again today and in a team with 3 SPGs, all the other tanks were dead in 3 minutes having killed between them just one enemy tank. Twelve tanks killed just one enemy tank, that's right.

If you want to see how your teams stack up against the enemy,  try this: http://forum.worldof...the-mm/#topmost

You do need a collection of replays to run it on though, so turn replay recording on of you haven't already. 


Edited by Baldrickk, 12 August 2018 - 10:49 PM.


Geoffrey_Ironfist #7 Posted 13 August 2018 - 02:59 AM

    Lance-corporal

  • Player
  • 4623 battles
  • 58
  • Member since:
    06-23-2018
Thanks for this. That will be interesting. My winrate seems to oscillate every few days with ups and downs. These oscillations were biggest the first couple of weeks after opening an account then calmed down. Maybe I am weird and I play badly for a bunch of games every four or five days but why is everyone else on the team doing the same?

Baldrickk #8 Posted 13 August 2018 - 07:21 AM

    Field Marshal

  • Player
  • 29530 battles
  • 13,766
  • [-TAH-] -TAH-
  • Member since:
    03-03-2013

View PostGeoffrey_Ironfist, on 13 August 2018 - 02:59 AM, said:

Thanks for this. That will be interesting. My winrate seems to oscillate every few days with ups and downs. These oscillations were biggest the first couple of weeks after opening an account then calmed down. Maybe I am weird and I play badly for a bunch of games every four or five days but why is everyone else on the team doing the same?

Player populations vary with time,  so time of day, day of week have an impact on the overall games you get.

Beyond that,  it's random:

It's random who ends up on which team

It's effectively random therefore how a team chooses to deploy, which is a bigger factor in the result of a battle than pure skill

It can be random if an influential action occurs in a game,  e.g. https://youtu.be/412aft3xD94 (sorry about the video quality,  it recorded 8n an aspect ratio YouTube doesn't like) The chances of hitting that shell were very low, but it led to the loss of a higher tier scout. On an open map.  My team dominated that flank with relative ease after that. 



SilentGaze #9 Posted 13 August 2018 - 08:33 AM

    Sergeant

  • Player
  • 771 battles
  • 260
  • Member since:
    03-22-2017

View PostBaldrickk, on 13 August 2018 - 07:21 AM, said:

Player populations vary with time,  so time of day, day of week have an impact on the overall games you get.

Beyond that,  it's random:

It's random who ends up on which team

It's effectively random therefore how a team chooses to deploy, which is a bigger factor in the result of a battle than pure skill

It can be random if an influential action occurs in a game,  e.g. https://youtu.be/412aft3xD94 (sorry about the video quality,  it recorded 8n an aspect ratio YouTube doesn't like) The chances of hitting that shell were very low, but it led to the loss of a higher tier scout. On an open map.  My team dominated that flank with relative ease after that. 

I've been thinking that wot games are chaotic systems. They are deterministic, non-linear, dynamic systems whose behavior can be predicted from the initial situation.  My experience is, that games Lyapunov time is about two minutes. That means that you can predict game results only 2 minutes ahead, which means, that after that time game changes to chaotic. After that change (bifurcation point), game behavior can be predicted from that new initial situation again 2 minutes.

 

I've had several games where there have been several bifurcation points, points where game changes totally. Team which at the start had 70% chance to win, suddenly has only 40% chance.  That kind of chances may happen several times during a single game. I think that that there are many reasons for those bifurgation points, one is RNG in game.



Baldrickk #10 Posted 13 August 2018 - 09:01 AM

    Field Marshal

  • Player
  • 29530 battles
  • 13,766
  • [-TAH-] -TAH-
  • Member since:
    03-03-2013

View PostSilentGaze, on 13 August 2018 - 08:33 AM, said:

I've been thinking that wot games are chaotic systems. They are deterministic, non-linear, dynamic systems whose behavior can be predicted from the initial situation.  My experience is, that games Lyapunov time is about two minutes. That means that you can predict game results only 2 minutes ahead, which means, that after that time game changes to chaotic. After that change (bifurcation point), game behavior can be predicted from that new initial situation again 2 minutes.

 

I've had several games where there have been several bifurcation points, points where game changes totally. Team which at the start had 70% chance to win, suddenly has only 40% chance.  That kind of chances may happen several times during a single game. I think that that there are many reasons for those bifurgation points, one is RNG in game.

Sounds like a good explanation,  only I'd posit that it is more chaotic at the beginning, where it can literally go either way,  and that as a general rule the outcome becomes more predictable over time



Tensim #11 Posted 13 August 2018 - 11:10 AM

    Lance-corporal

  • Player
  • 2278 battles
  • 72
  • Member since:
    05-05-2018

View PostSilentGaze, on 13 August 2018 - 08:33 AM, said:

I've been thinking that wot games are chaotic systems. They are deterministic, non-linear, dynamic systems whose behavior can be predicted from the initial situation.  My experience is, that games Lyapunov time is about two minutes. That means that you can predict game results only 2 minutes ahead, which means, that after that time game changes to chaotic. After that change (bifurcation point), game behavior can be predicted from that new initial situation again 2 minutes.

 

I've had several games where there have been several bifurcation points, points where game changes totally. Team which at the start had 70% chance to win, suddenly has only 40% chance.  That kind of chances may happen several times during a single game. I think that that there are many reasons for those bifurgation points, one is RNG in game.

 

I've never studied dynamical systems theory but you may be right over a number of games where recurrence may exist (many complain of losing streaks). However, a single game is a closed system and the parameter change of a player being killed results in a reduced set of bifurcation outcomes. So, as Baldrickk suggests, as the number of players in the game tends to zero, the outcome is likely to become more predictable.



Geoffrey_Ironfist #12 Posted 13 August 2018 - 11:26 AM

    Lance-corporal

  • Player
  • 4623 battles
  • 58
  • Member since:
    06-23-2018

I can see how killing a good scout could decide the course of a battle. I cannot see why 12 tanks can only kill one enemy tank or why all 10 turreted tanks will go down one side leaving five SPGs and TDs at the base either to trail behind them out of cover or die. Of course they will die. And this battle after battle. I sometimes change my type of vehicle when this happens but the trend hits back with the new vehicle in another way. Let's say if the turreted tanks kept leaving the arty undefended, I switch to a heavy tank, I win a battle, then next time I never even get to see an enemy tank before my entire team is gone and this happens again and again. I switch to a TD, I kill a couple of tanks the first battle, then every next battle for half a dozen battles I get swarmed by light tanks that pass through the frontlines. Once a light tank went through our entire team straight down the middle and killed my SPG, it was the first friendly tank to die. However, these bizarre sorts of things happen just on those streaks of bad luck or whatever.

 

Having battle disasters is one thing but having two dozen of them in a row is another. Over a thousand games it might still be just chance that this will happen - not that this will make it any less frustrating. On the other hand, looking at how my winrate oscillates and the fact that I get these runs of battle disasters just when I have reached a peak in the winrate oscillation makes me wonder if this is entirely random. I also rarely have such disastrous battles outside those runs, that last from around 15 to 30 battles. Even after taking a break and coming back, even the next day, this goes on until my winrate has dropped significantly. The first time it happened I had just reached a winrate of 47% in about 1000 games. There was quite a run of these disaster battles after that point. There were at least a couple dozen battles when my team did not kill more than a couple of enemy tanks. But my other battles are fairly normal. I don't think I often get outcomes with one or the other team losing just 2 or 3 tanks but during those runs almost every battle is like that. It might still be chance, I mean there was a very small chance that all the dinosaurs would go extinct and they did. Next time I have such a battle I will start saving battles in case I can catch another such run.

 

View PostSilentGaze, on 13 August 2018 - 08:33 AM, said:

I've been thinking that wot games are chaotic systems. They are deterministic, non-linear, dynamic systems whose behavior can be predicted from the initial situation.  My experience is, that games Lyapunov time is about two minutes. That means that you can predict game results only 2 minutes ahead, which means, that after that time game changes to chaotic. After that change (bifurcation point), game behavior can be predicted from that new initial situation again 2 minutes.​

 

In those losing streaks I had, all battles were practically over in 3 minutes, some lasted up to 4 minutes just because the enemy could not find the last surviving friendly tank or something and I could predict the outcome of some of them in the first minute seeing how our team was deploying, so from your point of analysis they were pre-determined practically from the start, though this is not what I was thinking. I am paranoically thinking that an algorithm perhaps incidentally keeps the running winrate progress of a player at around a certain rate by trying to match teams so that the chances of a win are even over a running average for each player of that ability. So If a player has been improving fast for a hundred games, he may be thrown into teams that have many weak players who have been recently lucky to keep the chance of win:lose at 50%. Now if this improving player is a new player with a thousand games or two thousand games and he is thrown in a team of other recently lucky newbies just so that the other guys in the other teams who may each have had some recent loses might now win for a change and this goes on for a while until everyone equalises, you may get the kind of losing streaks I have seen. Just maybe.


Edited by Geoffrey_Ironfist, 13 August 2018 - 11:59 AM.


SilentGaze #13 Posted 13 August 2018 - 11:55 AM

    Sergeant

  • Player
  • 771 battles
  • 260
  • Member since:
    03-22-2017

View PostTensim, on 13 August 2018 - 11:10 AM, said:

 

I've never studied dynamical systems theory but you may be right over a number of games where recurrence may exist (many complain of losing streaks). However, a single game is a closed system and the parameter change of a player being killed results in a reduced set of bifurcation outcomes. So, as Baldrickk suggests, as the number of players in the game tends to zero, the outcome is likely to become more predictable.

 

The less there are variables in system the easier it is to predict it's outcome. It doesn't mean that your prediction is more reliable.  Of course if the difference between team member skills or player numbers is big, system is then not anymore so sensitive to changes.  Still if there is only one player left in both teams, there is always RNG which may make the predict  unreliable. 

Baldrickk #14 Posted 13 August 2018 - 12:37 PM

    Field Marshal

  • Player
  • 29530 battles
  • 13,766
  • [-TAH-] -TAH-
  • Member since:
    03-03-2013

View PostGeoffrey_Ironfist, on 13 August 2018 - 11:26 AM, said:

I can see how killing a good scout could decide the course of a battle. I cannot see why 12 tanks can only kill one enemy tank or why all 10 turreted tanks will go down one side leaving five SPGs and TDs at the base either to trail behind them out of cover or die. Of course they will die. And this battle after battle. I sometimes change my type of vehicle when this happens but the trend hits back with the new vehicle in another way. Let's say if the turreted tanks kept leaving the arty undefended, I switch to a heavy tank, I win a battle, then next time I never even get to see an enemy tank before my entire team is gone and this happens again and again. I switch to a TD, I kill a couple of tanks the first battle, then every next battle for half a dozen battles I get swarmed by light tanks that pass through the frontlines. Once a light tank went through our entire team straight down the middle and killed my SPG, it was the first friendly tank to die. However, these bizarre sorts of things happen just on those streaks of bad luck or whatever.

 

Having battle disasters is one thing but having two dozen of them in a row is another. Over a thousand games it might still be just chance that this will happen - not that this will make it any less frustrating. On the other hand, looking at how my winrate oscillates and the fact that I get these runs of battle disasters just when I have reached a peak in the winrate oscillation makes me wonder if this is entirely random. I also rarely have such disastrous battles outside those runs, that last from around 15 to 30 battles. Even after taking a break and coming back, even the next day, this goes on until my winrate has dropped significantly. The first time it happened I had just reached a winrate of 47% in about 1000 games. There was quite a run of these disaster battles after that point. There were at least a couple dozen battles when my team did not kill more than a couple of enemy tanks. But my other battles are fairly normal. I don't think I often get outcomes with one or the other team losing just 2 or 3 tanks but during those runs almost every battle is like that. It might still be chance, I mean there was a very small chance that all the dinosaurs would go extinct and they did. Next time I have such a battle I will start saving battles in case I can catch another such run.

 

 

In those losing streaks I had, all battles were practically over in 3 minutes, some lasted up to 4 minutes just because the enemy could not find the last surviving friendly tank or something and I could predict the outcome of some of them in the first minute seeing how our team was deploying, so from your point of analysis they were pre-determined practically from the start, though this is not what I was thinking. I am paranoically thinking that an algorithm perhaps incidentally keeps the running winrate progress of a player at around a certain rate by trying to match teams so that the chances of a win are even over a running average for each player of that ability. So If a player has been improving fast for a hundred games, he may be thrown into teams that have many weak players who have been recently lucky to keep the chance of win:lose at 50%. Now if this improving player is a new player with a thousand games or two thousand games and he is thrown in a team of other recently lucky newbies just so that the other guys in the other teams who may each have had some recent loses might now win for a change and this goes on for a while until everyone equalises, you may get the kind of losing streaks I have seen. Just maybe.

Oh I miss the days when light tanks could do that with some reliability. 

 

Ever considered that your winrate going up is the suspicious streak, and that the loss streaks are your status quo?

 

The real answer is that randomness forms streaks both good and bad. It's just how it is.

When you toss a coin you don't expect HTHTHTHTHTHTHTHTHTHTHTHT do you? But it's still a 50% chance of either.



Hellraiser0201 #15 Posted 13 August 2018 - 12:40 PM

    Corporal

  • Player
  • 23411 battles
  • 164
  • Member since:
    02-21-2013

View Postwobblybob, on 11 August 2018 - 06:51 PM, said:

The game is notorious for winning and losing streaks. There are endless posts on the topic. Whether it is random chance or coded in to the game engine is a matter of opinion but you will certainly encounter them. There are mods that you can load that keep track of daily performance but rule of thumb is if you lose 5 in a row take a break.

 

:D I forgot to take a break and i lost 14 in a row once :trollface: 

And how about 25 winning streak (my longest one) ?

Pure form of random ;) :bajan:

 

 


Edited by Hellraiser0201, 13 August 2018 - 12:46 PM.


Geoffrey_Ironfist #16 Posted 13 August 2018 - 01:43 PM

    Lance-corporal

  • Player
  • 4623 battles
  • 58
  • Member since:
    06-23-2018

Baldrickk, on 13 August 2018 - 12:37 PM, said:

Oh I miss the days when light tanks could do that with some reliability. 

 

Ever considered that your winrate going up is the suspicious streak, and that the loss streaks are your status quo?

 

The real answer is that randomness forms streaks both good and bad. It's just how it is.

When you toss a coin you don't expect HTHTHTHTHTHTHTHTHTHTHTHT do you? But it's still a 50% chance of either.

 

I do not need irony, I am just wondering, maybe. I am not arguing against the overall average being about 50%, only about the pattern. Yes I also have winning streaks. However, I do not remember a couple dozen victories practically in a row. I have been checking my running winrate and on most days days it was improving. Then one day it was 42%, then went back to 47-50% for a few days. I am not sure this is random but it may be. I am not excluding it, as I said all dinosaurs did actually go extinct from a random event, every single one of them, all in a row and so I guess maybe I have now and then such a bad day by chance.

 

If you have a winrate of 42 % on a day and for part of the day it was worse, then you can imagine probably I was having a terrible time, so probably something was going wrong for me in that sense quite reliably.

 

I thought I had kept a screenshot for that day's winrate but cannot find it. Unfortunately I have also not kept the replays. So I cannot claim anything, just wondering. I attach a screenshot of the winrate oscillations from https://tanks.gg/eu/Geoffrey_Ironfist. If there is a way to remove the smoothing of the curve, then one might see if there are losing streaks against a more regular background, or if this is just my mistaken impression.

 

winrate.jpg

 

 


Edited by Geoffrey_Ironfist, 13 August 2018 - 03:08 PM.


Baldrickk #17 Posted 13 August 2018 - 03:15 PM

    Field Marshal

  • Player
  • 29530 battles
  • 13,766
  • [-TAH-] -TAH-
  • Member since:
    03-03-2013

View PostGeoffrey_Ironfist, on 13 August 2018 - 01:43 PM, said:

 

I do not need irony, I am just wondering, maybe. I am not arguing against the overall average being about 50%, only about the pattern. Yes I also have winning streaks. However, I do not remember a couple dozen victories practically in a row. I have been checking my running winrate and on most days days it was improving. Then one day it was 42%, then went back to 47-50% for a few days. I am not sure this is random but it may be. I am not excluding it, as I said all dinosaurs did actually go extinct from a random event, every single one of them, all in a row and so I guess maybe I have now and then such a bad day by chance.

If you have a winrate of 42 % on a day and for part of the day it was worse, then you can imagine probably I was having a terrible time, so probably something was going wrong for me quite reliably.

 

Unfortunately I have not kept the replays. So I cannot claim anything, just wondering. I have kept though a screenshot of a 42% winrate at the end of the day one day and at the start of that day it was less than even that. I was relieved when it went over 40%. Unfortunately, I also have not kept a record of the string of terrible scores in those defeats.

Not trying to be ironic, just to address the facts.

 

Even in a 40% session you've only lost one battle less than 50% for every 10 battles you play. 

Block Quote

 However, I do not remember a couple dozen victories practically in a row.

 Not surprising really.

You don't "feel" the win if you die, quit back to the garage and the team win without you. 

We as a species are wired to pay more attention to negative events. Those are the ones that are likely to be life threatening, over good times.

 

You might want to turn on replays, for self reflection if nothing else. 

 

 

Edit: that looks like a generally smooth WR curve.


Edited by Baldrickk, 13 August 2018 - 03:27 PM.


Geoffrey_Ironfist #18 Posted 14 August 2018 - 09:22 AM

    Lance-corporal

  • Player
  • 4623 battles
  • 58
  • Member since:
    06-23-2018

View PostBaldrickk, on 13 August 2018 - 03:15 PM, said:

Not trying to be ironic, just to address the facts.

 

Even in a 40% session you've only lost one battle less than 50% for every 10 battles you play. 

 Not surprising really.

You don't "feel" the win if you die, quit back to the garage and the team win without you. 

We as a species are wired to pay more attention to negative events. Those are the ones that are likely to be life threatening, over good times.

 

You might want to turn on replays, for self reflection if nothing else. 

 

 

Edit: that looks like a generally smooth WR curve.

 

There were I think 3 such streaks in 4000 games. Out of so many battles, these losing streaks of about maybe 15-25 battles each would not even register. However, if there was a way to unsmooth the plot, they might be noticeable. One happened the day after I had reached the 47% winrate mark and was harrowing but as a beginner I took it in my stride and another one very frustrating streak was when I passed for the first time the 48% winrate mark and my overall ranking was I think around 2000 (though going up, still quite low and so I imagine were the rankings of most other players). Maybe that's probabilistic. However, if I was dying during my winning streaks, they would have probably not been so one-sided as when I was losing. In those losing streaks I was feeling I could not catch my breath, the battle would start and we would get immediately overran. Many were practically over in 3 minutes.

 

There is little point in recording over a long time to catch something that may happen only maybe once every 2000 battles or maybe only when you are a beginner player. Maybe it only takes one extra experienced player in the enemy team for a string of battle catastrophes, until you are knocked back down the rankings well enough for things to ease out. Maybe when there is just a selection of a lot of bad players and some ok players at the low tiers where there are lots of beginners, maybe a little change can make a big difference. I am not saying I am suspecting anything else. It could also be pure chance, as you say. Still they happened and were unpleasant and discouraging and I wanted to ask. You only call up a lawyer a few times in your lifetime, probabilistically speaking lawyers are never needed, or hospitals, etc. You mostly have car accidents as a new driver, and so on.

 

 


Edited by Geoffrey_Ironfist, 14 August 2018 - 09:26 AM.


Baldrickk #19 Posted 14 August 2018 - 12:03 PM

    Field Marshal

  • Player
  • 29530 battles
  • 13,766
  • [-TAH-] -TAH-
  • Member since:
    03-03-2013

View PostGeoffrey_Ironfist, on 14 August 2018 - 09:22 AM, said:

 

There were I think 3 such streaks in 4000 games. Out of so many battles, these losing streaks of about maybe 15-25 battles each would not even register. However, if there was a way to unsmooth the plot, they might be noticeable. One happened the day after I had reached the 47% winrate mark and was harrowing but as a beginner I took it in my stride and another one very frustrating streak was when I passed for the first time the 48% winrate mark and my overall ranking was I think around 2000 (though going up, still quite low and so I imagine were the rankings of most other players). Maybe that's probabilistic. However, if I was dying during my winning streaks, they would have probably not been so one-sided as when I was losing. In those losing streaks I was feeling I could not catch my breath, the battle would start and we would get immediately overran. Many were practically over in 3 minutes.

 

There is little point in recording over a long time to catch something that may happen only maybe once every 2000 battles or maybe only when you are a beginner player. Maybe it only takes one extra experienced player in the enemy team for a string of battle catastrophes, until you are knocked back down the rankings well enough for things to ease out. Maybe when there is just a selection of a lot of bad players and some ok players at the low tiers where there are lots of beginners, maybe a little change can make a big difference. I am not saying I am suspecting anything else. It could also be pure chance, as you say. Still they happened and were unpleasant and discouraging and I wanted to ask. You only call up a lawyer a few times in your lifetime, probabilistically speaking lawyers are never needed, or hospitals, etc. You mostly have car accidents as a new driver, and so on.

 

So a streak that should happen statistically rarely happens statistically rarely and yoy want to attribute it to something else?

 

I'm unsure as to what you mean by "unsmooth the plot".  If you mean you don't like the spline curves, feel free to plot those datapoints in excel with straight lines connecting them.  It's not going to change the overall shape of the graph. 

 

It seems to be a common occurrence that losing streaks are mossy commonly spotted when players are near a WR boundary. 

It wouldn't be because that's when people actually take notice, is it?

 

Even something that happens infrequently adds up over time.

WOT results are very "noisy" - a streak, win or loss in the short term is very hard to attribute one way or another.

In the long term, over many battles, the effect of random noise averages out. Even small, infrequent trends add up and will show.



mpf1959 #20 Posted 14 August 2018 - 12:03 PM

    Warrant Officer

  • Player
  • 13618 battles
  • 633
  • [T-D-U] T-D-U
  • Member since:
    10-29-2017
On the whole I prefer fast battles to slow ones, and being involved in a draw is the absolute pits, especially if I am dead but want the tank back.




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users