Jump to content


Clan Test Coordinator
  • Content Сount

  • Joined

  • Last visited

  • Battles

  • Clan


About rnat

  • Rank
    Sub Lieutenant
  • Insignia

Profile Information

  • Gender
    Not Telling

Recent Profile Visitors

2,061 profile views
  1. rnat


    The patch notes only mention the Atlanta getting the range buff.
  2. rnat

    Anyone know how long the MM-****-list lasts?

    The Shchors actually needs to be played pretty safe due to how squishy she is (and because everybody knows that fact), so don't go overly aggressive in her. That doesn't mean it's best to just max-range snipe if you can e.g. use cover or smokes to get closer to the front or attack from an off-angle. Usually medium range and above is what you aim for, the closer you can get safely the better. What might be a better idea is to try and make sure that from your spamming/kiting ect. position you aren't too far from your DDs and that you have angles on the likely locations where they might run into enemy DDs since her Soviet railguns have a relatively easy time hitting DDs even from medium range and slightly beyond. Additionally if you have a choice of targets (and none is a smol-boat you might hit/kill) try to exert pressure on enemy cruisers especially radar cruisers rather than just farming damage. (I assume you're already going for a kill if you have angles on a low hp ship, so I won't bother mentioning that here)
  3. rnat

    Anyone know how long the MM-****-list lasts?

    emphasis by me Since this is a known fact please provide the presumably easily accessible and surely plentiful irrefutable evidence for that assertion to enlighten the morons such as myself that have never heard of this fact.
  4. rnat

    killed by torps past their distance

    No offense intended but that doesn't sound like great achievement and more like they were absolutely useless. The 20km torps have a 2.7km torpedo detect, which is ~50% worse than the next worse torpedo. Also behind an enemy is the worst possible bearing to torpedo them from if I can interpolate that from your message. Assuming those 4 didn't have either active DD or CV support or more than 1 radar at their disposal you'd most likely have fared better with the 12km set in that situation.
  5. rnat

    the "carry harder!" thread

    Oh well, I tried. Shout out to our Smol, that iirc managed to ambush their Shima with neither radar nor DD support and out-brawled a Hinden. Got 3 caps in my BB and we lost on points because of course we did.
  6. So a system that incentivizes people to shoot DDs since they gain more rewards for the same total amount of damage dealt (due to the lower HP-pool) makes it easier for DDs because... ?
  7. rnat

    Bad example matchmaking

    Look through large amounts of battles and look for any significant deviation from what is expected in any parameter or group of parameters (something will always be slightly off, you always have that when you check a lot of parameters of any random sample so the difference needs to be significant) While this doesn't sound very exiting short of someone actually leaking the algorithm or internal documentation this is the most reliable way to find manipulation. Also note that if such a statistical analysis fails to find any abnormalities that means any manipulation (if present) can only have a tiny effect since that's literally what the statistical analysis is measuring, i.e. it would be subtle/hidden to the point of barely having any effect. You are underestimating how powerful randomization is if what you're looking for is a broad sample with as little inherent bias as possible. Compare the effort to write an algorithm that tries to perfectly sample the player base every time or tracking and trying to keep averages over long periods (with the added conga line of issues ranging from longer queue times through selection bias of different player groups, large amounts of data needed for each matchmaking session on a per player or ship basis ect) to just picking the first player/ship that comes along in the correct ship & tier and letting statistics/the law or large numbers do the job for you.
  8. rnat

    Bad example matchmaking

    Iirc no one attempting to check whether their is any unexpected pattern/rigging/manipulation in MM has found anything significant until now. Thus in the absence of evidence to the contrary the null hypothesis of MM being random (first come first served) within ship/tier restrictions (since that's the easiest to implement) holds. Naturally if you happen upon compelling statistical or trustworthy leaked evidence to the contrary ppl here would be all ears.
  9. rnat

    Match making with Winrate

    Are you deliberately ignoring the consequences of your proposal even when they've already been pointed out to you ? Yes, going by pure skill if you're good you'll stay good and bad players won't magically become decent either. It will no longer be accurately reflected in the winrate. A 60% WR guy will give his team an approximately 10% better chance to win over a random replacement because he's that much better than his counterpart on the other team (and the rest of that team). If you always match 60% guys on either side they will cancel each other out because suddenly the guy that helps his team win by outplaying the opposition will be faced against someone about equally good at outplaying the other team so to keep winning he'd have to be significantly better than any other player with the same winrate which we can safely assume is quite rare. And if that player was brilliant and kept winning anyways, as I've already noted he will then be matched against other players who kept on winning despite the odds, i.e. the cream of the cream. At some point everyone will face players so good that it's essentially a draw between them and their win rate will approach 50%. Same for any absolutely abysmal player. The game will keep on matching him against worse and worse opponents until in the end both are equally useless and his win rate will approach 50% as well no matter how bad he is. Keep in mind that no matter how bad a player is with your proposed system he would see "improvement" in his stats (dmg/wr ect) just because he will be matched against more and more useless enemies, giving an illusion of skill-progress and less incentive and opportunity to learn. And ofc the same holds for any given win rate player, the game would match him against opponents of equal skill which makes the outcome statistically (since this happens for all players) a 50/50. Unless you can give a cogent argument as to why this won't be the case in your system besides "it just won't be because reasons" or "you're all just elitists and afraid of challenge or change" or an explanation why the benefits are worth both punishing skill, rewarding bad play (MM wise) and making wr (which you base the selection on) a useless metric stop pretending it would make things better. Now please give such an argument and make all of us look foolish or admit that your proposal still needs some work to iron out those problems. I'd hate to loose the illusion that people actually try to engage in good faith arguments and discussions on the internet.
  10. rnat

    Research Bureau Questions

    To your first question: Yes. Any tech-tree branch that terminates in a T10 counts as a line for this purpose. Note that if you reset 2 or more lines that share common ships, e.g. you reset both the Khabarovsk and Grozovoi you will only get research points once for each shared ship, i.e. the Podvoisky, Gnevny and Minsk would only give you the research points from resetting the first line rather than both, unless you grind back to the Minsk before resetting the second line. The answer to the second question is also Yes. The game doesn't care how you unlock a ship, as long as you have enough distinct tech tree T10s in your port you get access to the bureau and the first win on the T5+ ships after a reset will give you RP, no matter if you ground them out or unlocked them via FreeExp.
  11. rnat

    Match making with Winrate

    Some numbers on why that's not a workable idea: 60%+ players are less than 1.5% of the player base, 55-60% players make up another ~5% At the same time 1 player is ~8.3% of the team. So while you might make this work with the 55-60% guys there's no way you have enough 60% ppl to fill even 20% of the matches (even disregarding tier and ship class). (And now try to find opponents for a 3 man div with wr in the 60s, the queue times would be ludicrous) As to why I think that idea is problematic: Currently the MM is fair in that it doesn't give a toss. You have an almost equal chance to be on team with on average better or worse players, the only difference coming from how you yourself measure up against some random guy that would otherwise be in your place, i.e. your own skill. Sure individual match ups might be very lopsided but usually both teams are on a more or less similar level stats-wise. Your proposal (and any similar one I've seen so far) neither changes the fact that a good player will most likely dumpster you but make the MM actually biased, in the sense that the worse you are the more good players will be allocated to your team to carry you and the better you are the more clueless people you have to drag to victory kicking and screaming. Given the pool you have to pick from it makes little sense to actually make the teams actively worse for a bunch of people while not helping those that were supposed to benefit grow. Not to mention Petes point that ofc winrates would start converging to 50% since you'd always get matched with someone of similar skill level your own skill level would mostly cancel out, because you'd always have an opponent that was as good or bad as you on the enemy team, the same being true for all players. Even if you kept on winning in that system at time of change because you're god himself the MM would simply pick better and better opponents for you until you start loosing as often as you win, same argument for those that still would keep loosing. That's an easy straw man. Yes when you play both teams are on average not equal because you're in one them. If you're a horrible player on average your team will be worse, if you're good on average your team will be the better team, it's as easy as that.
  12. rnat

    General CV related discussions.

    I'm sorry to hear that, but at the same time glad that it didn't cause any major damage. Hope it will have been your last stroke.
  13. rnat

    General CV related discussions.

    You prolly meant exiting loss ? I agree that winning is fun, but I personally prefer an exiting match that is lost fairly to an as-good or better opponent over a boring win any time of the day. (Not included are those matches where my team is completely and utterly pepega and we still barely loose because the enemy brought less glue)
  14. rnat

    "Defense" Ribbon

    The Ognevoi was under too much fire for most of its life to accumulate enough cap-points to give you a ribbon. There's only 1 salvo where I can't say for sure it either doesn't have enough cap points or some other ship resets her before your shells impact and on that you have a salvo from you and an ally arrive at about the same time, so it might have been a slight de-synch.
  15. rnat

    General CV related discussions.

    Almost makes me want to demount that flag and not go full try-hard pleb all the time. almost.