Jump to content


Confirmation Bias or a Game Feature?


  • Please log in to reply
31 replies to this topic

TRK213_Turkey #21 Posted Aug 23 2019 - 21:10

    First lieutenant

  • -Players-
  • 6348 battles
  • 696
  • [WCTNT] WCTNT
  • Member since:
    01-02-2017
I suspect that there is some sort of last chance function that keeps a tank from succumbing to a hit that otherwise should have killed it.  I once hit a tank which had less HP left than 75% of my own alpha.  The tank survived with a few HP left.  I went back and looked at the replay just to be sure (this was a long time ago and I don't know if I still have the replay).  I have had the same thing happen with hits on my own tank.  So, yeah, I think there is some sort of "you got lucky" code in there somewhere.

Markd73 #22 Posted Aug 26 2019 - 18:22

    Major

  • Players
  • 32795 battles
  • 5,095
  • Member since:
    04-20-2011

View PostSaffron_Blaze, on Aug 23 2019 - 02:08, said:

That does seem like a more correct concept, but I am not convinced what I described is just some sort of bias. It is just too frequent of an occurrence given how often it should occur statistically. Not bothered if they do have such a scheme, although I would be interested in knowing what purpose it serves.

 

Perhaps. There is also the chance that your response was confirmation bias about your existing belief. Without tracking the frequency of occurrence and tracking this against expected (from a verifiable and non-biased source) and then publishing the raw data; then this is simply an assertion without evidence. See all the work that Neatoman has done on Matchmaking and his raw data for an example.

 

I swear I am fun at parties...


Edited by Markd73, Aug 26 2019 - 18:22.


Saffron_Blaze #23 Posted Aug 26 2019 - 18:36

    Sergeant

  • -Players-
  • 22369 battles
  • 189
  • Member since:
    09-23-2017

View PostMarkd73, on Aug 26 2019 - 18:22, said:

 

Perhaps. There is also the chance that your response was confirmation bias about your existing belief. Without tracking the frequency of occurrence and tracking this against expected (from a verifiable and non-biased source) and then publishing the raw data; then this is simply an assertion without evidence. See all the work that Neatoman has done on Matchmaking and his raw data for an example.

 

I swear I am fun at parties...

 

Right, no argument there, but since I hadn't done any such research I came here to see if someone else has. They haven't. Neatoman stuff is interesting, but it doesn't have anything to do with this.



AndrewSledge #24 Posted Aug 26 2019 - 18:43

    Major

  • -Players-
  • 36158 battles
  • 2,018
  • [1_FFD] 1_FFD
  • Member since:
    04-14-2014
There was a game patch a few years ago that increased the chance of a saving throw to leave you on 1 HP.

Saffron_Blaze #25 Posted Aug 26 2019 - 19:07

    Sergeant

  • -Players-
  • 22369 battles
  • 189
  • Member since:
    09-23-2017

View PostAndrewSledge, on Aug 26 2019 - 18:43, said:

There was a game patch a few years ago that increased the chance of a saving throw to leave you on 1 HP.

 

If it exists, I doubt it is set to just 1 hp.



Markd73 #26 Posted Aug 26 2019 - 19:13

    Major

  • Players
  • 32795 battles
  • 5,095
  • Member since:
    04-20-2011

View PostSaffron_Blaze, on Aug 26 2019 - 17:36, said:

 

Right, no argument there, but since I hadn't done any such research I came here to see if someone else has. They haven't. Neatoman stuff is interesting, but it doesn't have anything to do with this.

 

It does in the sense that in order to guard against confirmation or negativity bias, that properly constructed and open data collection and analysis can help identify and/or rule out these bias challenges.



Saffron_Blaze #27 Posted Aug 26 2019 - 19:29

    Sergeant

  • -Players-
  • 22369 battles
  • 189
  • Member since:
    09-23-2017

View PostMarkd73, on Aug 26 2019 - 19:13, said:

 

It does in the sense that in order to guard against confirmation or negativity bias, that properly constructed and open data collection and analysis can help identify and/or rule out these bias challenges.

 

Again, no argument that research is good. My point is that none has been done to prove or disprove there is a feature of the game that conspires to deny kill shots in the manner described. However, one could argue dispersion and RNG do this already, but I would suggest without direct intent to meet that particular aim.



Markd73 #28 Posted Aug 26 2019 - 20:44

    Major

  • Players
  • 32795 battles
  • 5,095
  • Member since:
    04-20-2011

View PostSaffron_Blaze, on Aug 26 2019 - 18:29, said:

 

Again, no argument that research is good. My point is that none has been done to prove or disprove there is a feature of the game that conspires to deny kill shots in the manner described. However, one could argue dispersion and RNG do this already, but I would suggest without direct intent to meet that particular aim.

 

Completely agree. Then IMHO the only intellectually honest position to take is "I don't know" until said research/evidence is presented and tested.


Edited by Markd73, Aug 26 2019 - 20:44.


Saffron_Blaze #29 Posted Aug 26 2019 - 21:00

    Sergeant

  • -Players-
  • 22369 battles
  • 189
  • Member since:
    09-23-2017

View PostMarkd73, on Aug 26 2019 - 20:44, said:

 

Completely agree. Then IMHO the only intellectually honest position to take is "I don't know" until said research/evidence is presented and tested.

 

That is fundamentally the scientific method. However, to properly use it one should discuss the issue and raise questions that can lead to not only an appropriate hypothesis to test but the potential means to conduct the research. I am certainly not there yet. I hope you aren't counselling to not discuss issues unless you already have the answer.



Markd73 #30 Posted Aug 26 2019 - 21:43

    Major

  • Players
  • 32795 battles
  • 5,095
  • Member since:
    04-20-2011

View PostSaffron_Blaze, on Aug 26 2019 - 20:00, said:

 

That is fundamentally the scientific method. However, to properly use it one should discuss the issue and raise questions that can lead to not only an appropriate hypothesis to test but the potential means to conduct the research. I am certainly not there yet. I hope you aren't counselling to not discuss issues unless you already have the answer.

 

No I am counselling to not adopt a specific position (pro or against) without sufficient credible evidence to support said position.



Saffron_Blaze #31 Posted Aug 26 2019 - 22:34

    Sergeant

  • -Players-
  • 22369 battles
  • 189
  • Member since:
    09-23-2017

View PostMarkd73, on Aug 26 2019 - 21:43, said:

 

No I am counselling to not adopt a specific position (pro or against) without sufficient credible evidence to support said position.

 

Indeed! I am sure that never happens here on the WoT forums anyway.



Markd73 #32 Posted Aug 26 2019 - 22:41

    Major

  • Players
  • 32795 battles
  • 5,095
  • Member since:
    04-20-2011

View PostSaffron_Blaze, on Aug 26 2019 - 21:34, said:

 

Indeed! I am sure that never happens here on the WoT forums anyway.

 

I am drowning in the sarcasm. 

 

+1






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users