Announcement

Collapse

Forum Rules

  • No flaming or derogatory remarks, directly or through insinuation.
  • No discussion, sharing or referencing illegal software such as hacks, keygen, cracks and pirated software.
  • No offensive contents, including but not limited to, racism, gore or pornography.
  • No excessive spam/meme, i.e. copious one liners in a short period of time, typing with all caps or posting meme responses (text/image).
  • No trolling, including but not limited to, flame incitation, user provocation or false information distribution.
  • No link spamming or signature advertisements for content not specific to Dota 2.
  • No Dota 2 key requests, sell, trade etc.
  • You may not create multiple accounts for any purpose, including ban evasion, unless expressly permitted by a moderator.

  • Please search before posting. One thread per issue. Do not create another thread if there is an existing one already.
  • Before posting anything, make sure you check out all sticky threads (e.g., this). Do not create new threads about closed ones.
  • It is extremely important that you post in correct forum section.

  • Balance discussion only in Misc.
  • All art related (such as hero model) feedbacks go to Art Feedback Forum.
  • All matchmaking feedback should go here: Matchmaking Feedback
  • All report/low priority issues should go here: Commend/Report/Ban Feedback
  • No specific workshop item feedback. These should go to workshop page of that item.
  • When posting in non-bugs section (such as this), use [Bugs], [Discussion] or [Suggestion] prefix in your thread name.



In case you object some action by a moderator, please contact him directly through PM and explain your concerns politely. If you are still unable to resolve the issue, contact an administrator. Do not drag these issues in public.



All rules are meant to augment common sense, please use them when not conflicted with aforementioned policies.
See more
See less

Out of curiosity, what RNG algorithm are we using

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Out of curiosity, what RNG algorithm are we using

    When SK or mortred crits six times in a row i always find it a bit odd. And it's not like that happens once every 100 games, there is a point it happens almost every game.


    15 percent chance of something happening 6 times in a row is pretty slim. Yet it happens every game.

    I assume we're using some form of pseudo random number generation so with every non-proc the percent chance of a proc goes up. But what's the need. Can we just get a really solid random algorithm in place.

    By leaving it up to pseudo random number generation you're manipulating the chance of something happening.

    My opinion on the matter is either remove RNG all together or find a decent algorithm, the in between shit we have going just isn't very cool.

    http://www.codinghorror.com/blog/200...f-naivete.html


    Opinions?

  • #2
    use the search function (it has been discussed more than once already).
    post in the right forum section (if you want to discuss -> brainstorming, if you are not sure -> brainstorming, if it is no bug -> brainstorming).
    post balance discussions on playdota.com rather than here.
    Make sure to read the Forum Rules as well as the stickied Threads of the Forum Section you are posting in.

    Contributions i'd like to highlight:
    My Suggestion: Coaching System
    My Sticky: Intended Changes List
    My Challenge: Completely Fixed Hero Challenge: Skywrath Mage

    Comment


    • #3
      Would you mind showing us some Match IDs from games where Mortred happened to proc coup de grace 6 times in a row?

      Comment


      • #4
        http://dev.dota2.com/showthread.php?t=20582
        That's the main thread to discuss this, though it's been kinda dead for a while now.
        Between pages of ~5-13 is probably worth skimming through as MDuh made some RNG tests with Slardar's bash there, additionally at the last page MDuh has a test with Mortred in which he gets a 5-crit streak.

        It's without a doubt that DotA 2 uses true RNG rather than PRNG, however so far they've completely ignored us.

        Comment


        • #5
          Originally posted by Arxos View Post
          http://dev.dota2.com/showthread.php?t=20582
          That's the main thread to discuss this, though it's been kinda dead for a while now.
          Between pages of ~5-13 is probably worth skimming through as MDuh made some RNG tests with Slardar's bash there, additionally at the last page MDuh has a test with Mortred in which he gets a 5-crit streak.

          It's without a doubt that DotA 2 uses true RNG rather than PRNG, however so far they've completely ignored us.
          You know that creating specialized hardware just to use true random for a video game is totaly unecessary and overkill lol (altho i must say that there will be processors with integrated true random hardware).
          Games use Pseudo random generators because you can't make a true random out of a determenistic system... and because probably nobody will die/loose gigantic sums of money just because your random has a finite loop.

          Also we can't deduce anything out of n times in a row strikes (in a true random it could happen that your every strike would be critical), unless somebody makes a detailed statistics with 1000+ games with critical in them, nothing can be concluded.


          Anyway i suspect a standard linear distribution (pseudo) random number generator, perhaps even the one that is already in the programing language as is (tho making a custom one is a 1min job...).


          P.S. human beings are incredibly bad at predicting/interpreting chances. If somebody had 50% chance of doing critical he would feel like he doesn't crit 50% of the time, and in contrast, if his opponent had 50% crit chance, he would have the feeling that it was definately higher.
          It's just in our nature. One can find quite interesting stuff regaring this matter :3


          edit: as you stated that some peps made some tests will definately check it out. Ofc there is always the chance that they use a poor random generator
          Last edited by zinus; 07-09-2012, 04:00 AM.

          Comment


          • #6
            there is a clear tendency in this algorithm: if it hits, it hits several times in a row. It basically works like this: *pause* hit several times in a row *pause* hit several times in a row *pause*

            even for an amateur eye it is very poorly written algorithm.

            Comment


            • #7
              Originally posted by zinus View Post
              *Long*
              I didn't mean true RNG as in 100% random. I just meant that it's not psuedo-random, and that streaks of any length can occur if given a large enough testing sample.
              And the thread I linked is kind of spammed with people giving unscientific reports, particularly in regards to the hero randoming. With a pool of almost 100 heroes, you'd need to do at least 10,000 randoms for your tests to have even the slightest significance.

              However, the tests with Slardar and Mortred depict quite clearly that the chance of long streaks occurring is many times bigger than in WC3. If I remember it's completely impossible to NOT get a bash proc after ~20-25 hits on r3/4 Slardar bash.
              And doing more than 3 Mortred crits in a row is incredibly unlikely. The chance for the 4th crit will be maybe 0.1% at most, probably less. The chance of a 5th crit would be astronomical.


              Originally posted by beinbliss View Post
              there is a clear tendency in this algorithm: if it hits, it hits several times in a row. It basically works like this: *pause* hit several times in a row *pause* hit several times in a row *pause*

              even for an amateur eye it is very poorly written algorithm.
              I don't think it works that way, because I don't believe they've written any PRNG algorithm at all.
              If you think about it the chance of it doing a perfect 1 bash every 4 hits is a lot less likely than it seemingly tying together in strings like that.
              That comes down to what zinus said with human perception.

              Comment


              • #8
                human beings are not good at compiling and analyzing information like this. we (everyone) tend to notice when something unusual happens. 2 entangles back-to-back or motred critting many times in a row are things that anyone will notice. however, we tend not to notice when bear roots 1/5 attacks or mortred crits 1/6-7 attacks as is average.

                this is why there are so many threads questioning the RNG system, because people do not look at events like back-to-back bashes/crits objectively

                Comment


                • #9
                  Originally posted by Arxos View Post
                  I didn't mean true RNG as in 100% random. I just meant that it's not psuedo-random, and that streaks of any length can occur if given a large enough testing sample.
                  And the thread I linked is kind of spammed with people giving unscientific reports, particularly in regards to the hero randoming. With a pool of almost 100 heroes, you'd need to do at least 10,000 randoms for your tests to have even the slightest significance.

                  However, the tests with Slardar and Mortred depict quite clearly that the chance of long streaks occurring is many times bigger than in WC3. If I remember it's completely impossible to NOT get a bash proc after ~20-25 hits on r3/4 Slardar bash.
                  And doing more than 3 Mortred crits in a row is incredibly unlikely. The chance for the 4th crit will be maybe 0.1% at most, probably less. The chance of a 5th crit would be astronomical.
                  pseudo-random refers to the practice of getting your random numbers from something that technically isn't random. for example a common way to get a random number is to take the time in milliseconds as a random number when it's not truly random. most games use pseudo-random methods to generate random numbers, and it works very well and doesn't need to be improved upon. people see "pseudo" and think it's sloppy or bad, when in reality there are many pseudo random methods that work very well.

                  when he refers to actual random numbers he's talking about something like this: http://www.random.org/ which uses atmospheric noise (can be any physical measurement that we expect to be random) in order to generate numbers that are truly random.

                  you seem to be confusing psudeo-random with the system that wc3 used

                  Comment


                  • #10
                    Based on the data analysis from the older thread ("Valve fix your RNG"), there seems to be a lot of evidence indicating that the RNG or PRNG isn't working very well; some of the data put forth seems to indicate that there may be additional issues with chance based effects such as slardar's bash which procs a lot less when his attack speed is higher and over procs on the lower levels.
                    The developers should seriously consider investigating this, it makes a huge difference to gameplay. This might just rationalize my fear of heroes with chance based effects (I almost never get crits or multicast).
                    "Better to run, than curse the road." - Clinkz, The Bone Fletcher.

                    sigpic

                    Does anyone else think that Icefrog didn't make this update?

                    Comment

                    Working...
                    X