Jump to content

MaxL_1023

Members
  • Content Сount

    615
  • Joined

  • Last visited

  • Battles

    6066
  • Clan

    [KRAB]

Community Reputation

309 Excellent

About MaxL_1023

  • Rank
    Warrant Officer
  • Birthday 10/23/1992
  • Insignia

Profile Information

  • Gender
    Male
  • Location
    Toronto
  • Interests
    Randomly Generated

Recent Profile Visitors

708 profile views
  1. No I don't, but I highly doubt it is much different than WoT.
  2. MaxL_1023

    Dev blog announced CE change

    They are trying to create a larger concealment spread between the ship classes - CE becoming more effective for larger ships had the unfortunate tendency of making BBs stealthier compared to CA/CL than the base values would indicate, and CA/CL stealthier compared to DDs. I fully agree that a BB should be outspotted by a CA/CL by a considerable margin when both are using stealth builds - right now this does not always happen, especially for ships such as HIV, Moskva and Hindenburg compared to Lion/Conqueror, Iowa/Mo, etc.
  3. The problem is that people are looking at lower-dimensional subsets of the true dataset. The win rate probability distribution (x-axis is win rate, y-axis is normalized number of players at that value) will always integrate to 1, however there is no requirement for half the probability density to be above/below 50%. The only requirement is that the total number of wins in the dataset equals the total number of losses, and there is an infinite number of win rate distributions which can fulfill this statistic - you have more free parameters to vary than you are using to create your distribution. With extra degrees of freedom you can get intuitive results which are still perfectly valid. In the case of WOWS and WOT, the large number of players slightly under water (with a lot of games) in the 47-50% range is balanced by a smaller population extending into the 60% plus range, also with large numbers of games. The games played/WR correlation isn't really strong enough to skew the whole distribution, but it would make it even more asymmetric if the expected positive correlation exists. Something like a gamma distribution is exactly what you would expect given the underlying "physics" of the system.
  4. It seems WG is trying to widen the concealment spread between classes, so DDs have more of an edge over CA/CL and CA/CL have more of an advantage over BBs when running the CE skill. Within-class and consumable balance will probably need to be looked at subsequent to this change - they may simply buff the concealment of certain cruisers slightly (DM, Worchester, Mino, Zao perhaps) to return to the current levels while leaving slight nerfs to the others. It is not a massive change (4% for BBs, 2% for CA/CL) but is is large enough to have an effect on game balance. I hope WG remembers to test this one out before release.
  5. MaxL_1023

    Prinz Eitel Friedrich Preview

    With <356mm guns it hardly matters - the tier 8 CAs and tier 6-7 BBs you face will just angle-tank everything you can put out. It either needs meme-level secondaries or very high DPS to be competitive.
  6. MaxL_1023

    The HMS Daring

    It seems like the Jutland does better due to the concealment.
  7. MaxL_1023

    Harugumo Verses Other Destroyers Unfair

    Or any competent Gearing or Yueyang player running a full-gunboat spec. Gearing gets down to 2.4 seconds using BFT and MBM3.
  8. So a 7% dispersion buff turned out to be... a 7% dispersion buff. Why is this news again?
  9. MaxL_1023

    No boxes from Hall of Fame again?

    If what I am hearing is correct the HoF rewards are using the items from this patch - seems like they forgot to grandfather in the HoF for this one cycle.
  10. MaxL_1023

    Harugumo Verses Other Destroyers Unfair

    Khab and Groz can range-kite it, Z-52 can beat it in close if you know how to use hyrdo properly. Anything which can't put up a good fight in a gun battle can outspot it and get away. It is strongest in a support role but is not as broken as you think when it has to straight-up fight.
  11. This would basically remove the viability of half the battleships in the game unless they do nothing but sit still and bow tank. The Yamato/Musashi will be completely broken, as they will overmatch torpedo bulges for free 33% pens no matter what angle they are at instead of being stopped by the belt underneath or overpenning the upper portion of the ship. This is way too big of a change for WOWS at this stage of development.
  12. It is a complex combination of factors. In general, there is a positive correlation between games played and win rate. People do get better over time, even though outliers on both sides exist. In addition, it is nearly impossible to have a negative team contribution (without eventually getting banned). This puts a floor on the win rate you can get over a significant sample - in WoT it was near 40%. I suspect the value is similar for WoWs overall, although for CVs it can be much lower due to their unique impact and matchmaking. There is also a nonlinear relationship between team contribution and win percentage. What ends up happening is that the distribution of win rates takes on a different shape due to these factors than the distribution of player skill. The overall rule of (person-wins = person-losses with no draws) holds, and we get the skewed distribution which happens to be a good match to the Gamma family. I honestly have no idea why that particular shape is prevalent, but it is probably related to running a Gaussian through a non-linear transform taking everything else into account.
  13. Win rate follows a gamma distribution - we did a lot of investigation into this back at WoTLabs for tanks and it holds here as well Gamma distribution are skewed left with a long right tail, leading to a lot of players between 45-50%, very few under 40% and proportionately more then you would expect above 60%.
  14. MaxL_1023

    Add STS as a USN line module?

    I am top 5 on NA in the Mass and I can tell you it can stop anything if you know what you are doing. Not all Battleships need to be idiot proof with underwater citadels and HE-proof deck plating.
  15. MaxL_1023

    Add STS as a USN line module?

    12mm of extra thickness would make every USN BB broken overpowered. Heck - 2mm more on the T8-T10 would significantly increase their resistance to CA HE, enough to disrupt balance. USN ships already have better DCP uptime - if the tier 3-7 USN BBs need some help (I think they do need a small buff) perhaps just give them a slightly better heal (~20% instead of 14% per charge perhaps). The NC and Montana are pretty much fine as they are. Iowa should get Missouri's armor scheme - it isn't right that the premium is a straight upgrade.
×