test content
What is the Arc Client?
Install Arc

Number Crunching with Hilbert

24

Comments

  • dukedom01dukedom01 Member Posts: 462 Arc User
    edited October 2012
    Incredible work so far Hilbert.
    Ceterum censeo Otha supplendum in praemiis.
  • zerobangzerobang Member Posts: 0 Arc User
    edited October 2012
    It's no fun for a player that has worked hard to unlock a certain power or effect, to be told they cannot use it in a certain circumstance. It's a more user-friendly option to tune such effects such that turning them off is never a necessity.

    *scratches head*
    you mean like operational assets that only work in fleetbase missions?

    *uhm*... yepp that happened
    [SIGPIC][/SIGPIC]
  • fakehilbertfakehilbert Member Posts: 252 Arc User
    edited October 2012
    Some EPtS charts are now available at http://hilbertguide.com/datamining/
  • snoge00fsnoge00f Member Posts: 1,812 Arc User
    edited October 2012
    A 3 hour tour...

    A 3 hour tour...

    I guess that game was full of Odysseys running extends.
    [SIGPIC][/SIGPIC]
  • edited October 2012
    This content has been removed.
  • carl103carl103 Member Posts: 1 Arc User
    edited October 2012
    mewi wrote: »
    Things like overhealing cannot really be accounted for on my side, since I have to work with the information that is in the combat log. This also means that I won't be able to analyse the decline of CPB/TachBeam since these abilities don't show up in the log.

    The first preview chart shows average magnitudes over time (something like a ship equipped with one normalised copy of EPtS/ES/TSS/...). The situation will improve a bit towards non-BFI healing when we look closer at absolute healing numbers and start adding some things instead of averaging them.

    I'll also be able to look at some data from premade matches prior to BFI/TT for comparison, although I'm afraid that the earliest data points I have (Dec 2010) won't be particularly meaningful since these are the super quick SW wins in the KT league. But I am confident that there will be some useful mid-2011 logs.

    Cheers for that. Fgured as much, unlike others i agree that BFI is a destinct issue, just thatthis chart is almost certianlly exagerating things.

    p.s. is it me or is the quote button broke,
  • havamhavam Member Posts: 1,735 Arc User
    edited October 2012
    Boy would i love GEcko to come in here and comment on these numbers.

    I guess we all just don't spec into the right counter skills, remember, there are always counters ;)
  • ghostyandfrostyghostyandfrosty Member Posts: 864
    edited October 2012
    Gecko would say something like:

    "Urrrr herpaderp derp derp!" Translation as I am quite fluent in Stupid Speak: "I AM GECKO MY MATH IS NEVER WRONG! THIS IS ALL A LIE!"

    Also Hilbert, repost your BFI #s too :P You deleted the link when you put up EPTS.

    I hope you don't mind me inserting some #s here of my own. I'm working on doing weapon tooltip data correlation.

    But in the mean time. Here is why brace for impact is broken.

    at full aux, when you smash it, you end up with an 11k Shield heal.
    To put this in perspective the best shield regen in the game (borg) at full shield power only regens a little less than 6k shields in that same time span.
    ST3 at full consoles heals around 5k shields in every 15 with either Two copies of ST3 or 2 dev lab scientist.
    Transfer Shield Strength 2 heals for around 4500 after 15 seconds on average at full aux.

    Who seriously thought this was a good idea.
  • leviathan99#2867 leviathan99 Member Posts: 7,747 Arc User
    edited October 2012
    It's no fun for a player that has worked hard to unlock a certain power or effect, to be told they cannot use it in a certain circumstance. It's a more user-friendly option to tune such effects such that turning them off is never a necessity.

    I realize Accuracy is one... But what about Resilience-style PvP stats?
  • leviathan99#2867 leviathan99 Member Posts: 7,747 Arc User
    edited October 2012
    hurleybird wrote: »
    I would tend to agree, in a perfect world at least. In the past I didn't think that Cryptic could ever pull it off, that they should cut their losses and go with something more GW-like, but now I'm not entirely sure ;)

    That said, some items (not gear) should be turned off. The fleet bonuses for example rightfully don't have an effect in PvP. The same should be true for bonuses you get from doff missions, consumables like ablative jevonite, and of course team batteries, but sadly isn't. I wouldn't mind seeing deuterium disabled in PvP either, I certainly wouldn't miss grinding it.

    There's a simpler way that doesn't knock achievement as hard:

    Instead of disabling those buffs, have an alternate way for PvPers to get all of them.

    My BIG idea here is a method of paid PvP tournaments, which I think is necessary to support PvP development under F2P.

    Here's the deal:

    Cryptic releases a variety of ships BALANCED AGAINST ONE ANOTHER. Not perfectly balanced against anything else. Just eachother.

    These ships cost a premium amount, more than normal ships.

    For 3 months or 6 months, Cryptic has a PvP tournament. Which can only be queued for using these ships. Rewards might include things like Fleet-tier upgrades of the PvP ships, a special hull texture for all your ships, etc.


    At the end of the tournament, the ships go down in price to 2500 or 2000 ZEN. This is more or less stated upfront than the higher cost is your "Tournament Entry Fee."
  • fakehilbertfakehilbert Member Posts: 252 Arc User
    edited October 2012
    The latest update includes an investigation of the claim that science heals like TSS and HE are not being used often enough.

    For most of us PVPers the idea that TSS and HE are not commonly used and overshadowed by engineering abilities was a bit strange. And indeed we find that the data from the last year does not seem to indicate a bias toward engineering heals.
  • edited October 2012
    This content has been removed.
  • husanakxhusanakx Member Posts: 1,608 Arc User
    edited October 2012
    sophlogimo wrote: »
    Remember, though, that you are only collecting from a very small subset of the player base.

    I don't know, but I would guess that the devs looked at the whole database in some way before judging that.

    If you mean they asked the people around the office that play STO... then ya your likely right. :)
  • edited October 2012
    This content has been removed.
  • hurleybirdhurleybird Member Posts: 909
    edited October 2012
    Data mining can only get you so far.

    If you just look at what the entire playerbase is equipping, you might get the idea that already weak skills like Jam Sensors and Boarding Party need to be nerfed based on the vast amount of PvErs who use those skills.

    On the other hand, if a player is even semi-serious about PvP he's going to learn quite fast how bad Jam and Boarding Parties really are and how life saving HE and TSS can be.
  • edited October 2012
    This content has been removed.
  • shimmerlessshimmerless Member Posts: 0 Arc User
    edited October 2012
    I think you're really overestimating the amount of complex work the devs perform when it comes to game balance.
    [SIGPIC][/SIGPIC]
    vids and guides and stuff

    [9:52] [Zone #11] Neal@trapper1532: im a omega force shadow oprative and a maoc elite camander and here i am taking water samples
  • edited October 2012
    This content has been removed.
  • shimmerlessshimmerless Member Posts: 0 Arc User
    edited October 2012
    sophlogimo wrote: »
    It is not as if the above was complex. It is actually just a no-brainer when developing a game like STO. Of course, it is assuming that the game uses an SQL-compatible databse that can be queried that way. Which I don't know.

    The game doesn't run on SQL, for some reason the devs had issues with their SQL package so they upped and wrote their own database language (no, I'm not kidding).

    I'd be sort of surprised if they cache arena results in the long term the way Hilbert did, I suppose it's not impossible. I just mean that I seriously doubt they mark players out or do much other than keep raw combat data.
    [SIGPIC][/SIGPIC]
    vids and guides and stuff

    [9:52] [Zone #11] Neal@trapper1532: im a omega force shadow oprative and a maoc elite camander and here i am taking water samples
  • edited October 2012
    This content has been removed.
  • mewimewi Member Posts: 0 Arc User
    edited October 2012
    It's no fun for a player that has worked hard to unlock a certain power or effect, to be told they cannot use it in a certain circumstance. It's a more user-friendly option to tune such effects such that turning them off is never a necessity.

    Logic Fallacy, it is also no fun to unlock a power and have it completely useless due to OTHER skills being unbalanced.

    The best games with the best PvPs know that two sets of calculations for skills is better than one ;3 I'd even like to see a 3rd set of calculations for Team PvP.
    [SIGPIC][/SIGPIC]
    | Join Date: January 2009 | Computer | Fleet: Broken Wings |
  • edited October 2012
    This content has been removed.
  • thishorizonthishorizon Member Posts: 1,158
    edited October 2012
    ....................

    Sprinkles?

    Sprinkles is that you?

    .....................
  • fakehilbertfakehilbert Member Posts: 252 Arc User
    edited October 2012
    sophlogimo wrote: »
    Remember, though, that you are only collecting from a very small subset of the player base.
    I believe my data is not as bad as you suspect. In fact I'm beginning to see why cryptic doesn't seem to care about balance and doesn't take our talk of excessive healing and pet spam serious: it's simply not visible in the averaged data.

    Maybe I'm still looking at the wrong things to detect these issues, but I'm fairly certain that there is a huge gap between matches with equally matched sides (the ones we remember the most clearly and which show us the decline of pvp balance) and the average pug matches where a single good player can turn the tide because the average player has vastly less damage or healing than us few dedicated pvpers.

    These average matches with short time-per-kill values (~40s per kill) tend to completely drown out the increasingly excessive extremes with >5-minutes-per-kill and massive pet spam that have occured more frequently during the last year and led to our perception of declining pvp balance.

    To see what we see, the devs would have to play some high-level pvp instead of just analysing average data or dabbling in pug matches.

    Maybe there is a way to extract these extreme cases by filtering for matches with >x minutes total duration, but I will have to adjust my data compiling tools to do that. Currently I only have totals per logfile which allow for calculation of average values on a weekly or monthly basis. But I think it might be possible to parse the logs differently to get true per-match data. (Well, theoretically this is possible with the tools I already have, but it would probably take dozens of hours to just compile the raw data since I would need to do a lot of things by hand. I need to write a tool to let the computer do this automatically.)
  • naz4naz4 Member Posts: 1,373 Arc User
    edited October 2012
    I believe my data is not as bad as you suspect. In fact I'm beginning to see why cryptic doesn't seem to care about balance and doesn't take our talk of excessive healing and pet spam serious: it's simply not visible in the averaged data.

    Maybe I'm still looking at the wrong things to detect these issues, but I'm fairly certain that there is a huge gap between matches with equally matched sides (the ones we remember the most clearly and which show us the decline of pvp balance) and the average pug matches where a single good player can turn the tide because the average player has vastly less damage or healing than us few dedicated pvpers.

    These average matches with short time-per-kill values (~40s per kill) tend to completely drown out the increasingly excessive extremes with >5-minutes-per-kill and massive pet spam that have occured more frequently during the last year and led to our perception of declining pvp balance.

    To see what we see, the devs would have to play some high-level pvp instead of just analysing average data or dabbling in pug matches.

    Maybe there is a way to extract these extreme cases by filtering for matches with >x minutes total duration, but I will have to adjust my data compiling tools to do that. Currently I only have totals per logfile which allow for calculation of average values on a weekly or monthly basis. But I think it might be possible to parse the logs differently to get true per-match data. (Well, theoretically this is possible with the tools I already have, but it would probably take dozens of hours to just compile the raw data since I would need to do a lot of things by hand. I need to write a tool to let the computer do this automatically.)

    Keep it up buddy. Talk to you later on vent about help if you need.
  • edited October 2012
    This content has been removed.
  • broken1981broken1981 Member Posts: 0 Arc User
    edited October 2012
    marc8219 wrote: »
    So what? This just shows the BFI Doff are working as intended.

    Its the 3pc borg set that still is the main problem, and mutliple extend shields and RSP, not BFI Doff.

    Balance the borg set, make it require 4pc bonus to get either the hull or shield heal procs, leave the Doff alone.

    Nerf the DOff and you will just force more people to use 3 pc borg.

    3 sdo and 3 pc borg is the standard. thoes doffs need to be set to 1. aux should not boost the heal either.
    Join Date: Dec 2007Originally Posted by BROKEN1981
    I can throw [Fireworks] at you and hope you catch on fire and burn to death lol
  • mewimewi Member Posts: 0 Arc User
    edited October 2012
    sophlogimo wrote: »
    I am pretty sure that is not what they are trying to do.

    I'm pretty sure I just quoted the exact text of what a dev just said. And I'm pretty sure that is what they are trying to do. It is obvious, when you place the facts behind it, they state "We want everyone to have fun with x y z skill and we don't want to hinder x y z skill for the sake of PvP" yet hinder MANY skills based on the fact that they are trying to keep PvE balance intact. Hence logic fallacy, hence that is what they are maintaining, a massive logic fallacy.
    sophlogimo wrote: »
    I'd rather want everything to work the same whenever I play the game, regardless of whether I do PvE or PvP. Really. That whole idea of putting 2 games into one freaks me out.
    The two elements of the game are already too different. If anything, make them more similar.

    the game mechanics cannot support balance on both ends, the game mechanics need to have separate calculations in order to ever hope to have true game balance.

    "It freaks me out" it isn't that scary, or difficult to comprehend, as long as cryptic is open about the differences between the calculations then it is EASY to figure out.

    The changes aren't being proposed like "In PvE this does an explosion, and in PvP this does a Beam." The underlining mechanic is the same, only the damage balance is different, or heal balance, yadda yadda.

    ... if that freaks you out, then I'm sure there are some single player games you could be playing ;)....
    [SIGPIC][/SIGPIC]
    | Join Date: January 2009 | Computer | Fleet: Broken Wings |
  • hurleybirdhurleybird Member Posts: 909
    edited October 2012
    mewi wrote: »

    the game mechanics cannot support balance on both ends, the game mechanics need to have separate calculations in order to ever hope to have true game balance.

    Technically you can't break balance for something as retardedly easy as PvE (with the exception of NWS)
  • mewimewi Member Posts: 0 Arc User
    edited October 2012
    hurleybird wrote: »
    Technically you can't break balance for something as retardedly easy as PvE (with the exception of NWS)

    Another point to be made here is, the limitations you are placing on PvE by not attempting to make PvE too hard, or too easy, whilst "maintaining" PvP balance.
    [SIGPIC][/SIGPIC]
    | Join Date: January 2009 | Computer | Fleet: Broken Wings |
Sign In or Register to comment.