Jump to content

LancasterOne

Beta Testers
  • Content Сount

    186
  • Joined

  • Last visited

  • Battles

    4931
  • Clan

    [BLAZE]

Community Reputation

245 Valued poster

1 Follower

About LancasterOne

  • Rank
    Chief Petty Officer
  • Insignia

Profile Information

  • Gender
    Male
  • Location
    Canada

Recent Profile Visitors

579 profile views
  1. Neg karma in CV equals good karma. Bad is good. The badder the bad, the gooder the good.
  2. LancasterOne

    Do WG listen to Supertesters and CCs

    One day I'd like to compile all the feedback that's ever been submitted by Supertesters and see what's the percentage that's got implemented in the release build lol
  3. LancasterOne

    Mass and Bismarck

    Bismarck NEEDS FPE. I don't know what's with german ships but they get set on fire like no one's business. Mass meanwhile could just do with CE to sneak into secondary range.
  4. LancasterOne

    Should I build for full secondaries on my Georgia?

    Only do secondaries build on the Georgia for the lols. If you're in brawling range, you're in losing half your HP pool to citadels range. It's not tanky like the Germans and it's not agile. It's an American drag car. Goes fast. Die a virgin.
  5. LancasterOne

    Statistical proof that MM is NOT RIGGED

    The part about the researcher being part of the data and hidden stats are a fair point. However hidden stats can appear on both sides of the team and I don't have any evidence that suggests that people with hidden stats tend to be particularly bad or particularly good players nor can I say I am that special of a CV player that would skew the data everytime. It's an imperfect analysis but it's the best I can do without pulling WoWS matchmaker API if that's even possible. It is impossible to see hidden data or remove the self from the data gathering process.
  6. LancasterOne

    Statistical proof that MM is NOT RIGGED

    Well what I'm thinking is what is the correlation between the number of unicums you have on your team and the likelihood of that match coming out as a win. Though to think ima have to MANUALLY INPUT every single player's WR...
  7. LancasterOne

    Statistical proof that MM is NOT RIGGED

    tbh I never see people quote where they found that 49% number. I was on wows.stats and na.wowsnumber and both of them don't give you that 49% thing.
  8. LancasterOne

    Statistical proof that MM is NOT RIGGED

    go back to the opening post. You see the test data? That and your personal WR. For my next research I kinda wanna research the correlation between an individual's performance and the outcome of the game. Say..... do unicums really help you win or is it all down to randomness in the end? Vice versa.
  9. LancasterOne

    Statistical proof that MM is NOT RIGGED

    well if you got the data I'd certainly like to take a look.
  10. LancasterOne

    Statistical proof that MM is NOT RIGGED

    The data comes from WoWs Monitor, and they pull that data from WoWStats. I then selected the win rate for their whole account since I think that is more representative of a player's "skill" as a whole than individual ships which oftentimes only have maybe a few dozens of battles in. Oh yeah I do recall the required sample size for it to within say 95% confidence interval of a certain standard error but to do that I'll need the SD of the population which unfortunately I do not have since I'd need access to the entire database. A big reason of why I even did this research in the first place is in this... what...3...4 years of people constantly whining about how WG rigs the MM to keep players at 50% and all these anecdotes and conspiracies....no one, not a single person has ever put it to the test and did the research to fundamentally confirm or deny the hypothesis. I'm no statistician by any means but my uni stat class was only a few semesters ago so I decided to put it to use and answer this question once and for all. Pretty fun and rewarding to finally be able to solve a mystery like this and put down the conspiracy.
  11. LancasterOne

    Statistical proof that MM is NOT RIGGED

    I've been in the Kaga for over 12 hours sitting through 6-5 losing streaks. Tasting lead laced paint would be more fun lol. Rekindling a lost passion for statistics though. I think ima go learn R and Python now.
  12. LancasterOne

    Statistical proof that MM is NOT RIGGED

    Honestly for my next research I wanna see how much unicums' performance can affect the outcome of a game though I'm not a unicum so I don't really have data for the next research. Someone already did a research on potatoes so.
  13. LancasterOne

    Toxic player of the day

    You play CV and do too well? You're losing karma. You play DD and dev strike a BB? You're losing karma. You play CL and smoke HE spam? Neg karma. Karma really doesn't mean anything. It's just a punching bag for people.
  14. LancasterOne

    Statistical proof that MM is NOT RIGGED

    The way I see "skill" is that a skillful play should lead to positive outcomes for a team. It doesn't matter how much damage you do or how much you tank, if you lose in the end, it was all for naught. That's why I chose WR as the prime indicator for player performance. The avg wr only reports to the mode we are playing so there's no contamination. We are only talking about ranked in this test but I can imagine random to be pretty similar.
  15. LancasterOne

    Statistical proof that MM is NOT RIGGED

    The team WR is the mean of all the players' mean account WR.
×