• Announcements

    • rei

      Battle of the Anime Characters   02/10/17

      THE BATTLE OF THE ANIME CHARACTERS HAS BEGUN   Find it here http://duelistgroundz.com/index.php?/forum/615-battle-of-the-anime-x/

ACP

+Administrator+
  • Content count

    14501
  • Joined

  • Last visited

Community Reputation

33012 RAWR means ilu in dinosaur

About ACP

  • Rank
    Not everyone can be a master
  • Birthday 11/28/90

Profile Information

  • Gender fucking white male
  • Fav YGO Deck Is Gishki still good?
  • Location FL
  • Interests Creating ridiculous profile fields

Recent Profile Visitors

10666 profile views
  1. It's interesting that you bring up spotify, because music playlists are intentionally programmed to be quite non-random due to the way that humans perceive "randomness." The issue originally came about with the "shuffle" feature of the iPod and similar apple products. Customers kept complaining to Apple that their shuffler was "not random," while Apple's computer scientists insisted that it was in fact, quite random. Eventually, someone figured out that what the customers wanted was not a shuffler that was truly random, but one that focused on eliminating "clumps." That is, when the customer hears 5 songs in a row by the same artist on his supposedly "shuffled" device, he thinks that there's some bug in Apple's shuffler, when it reality, what he doesn't realize that a truly random shuffler could play 5 songs in a row by the same artist quite often. Apple decided to take a "customer's always right" mentality and restructure their algorithm such that it was less likely to play a song by the same artist twice in a row, and in most circumstances impossible for it to play a song by the same artist three or more times in a row. From what I understand, most other music-streaming companies followed suit.
  2. Lol Don't let Atem hear you say that.
  3. Obviously it matters to some extent. Take an extreme example: Suppose a "shuffler" gives you the exact same 5 card hand every game. You can't say, "Well it doesn't really matter since irl shuffling is flawed anyways." Clearly, no one wants to test the same 5-card hand every game. Of course, in reality, the problems will be a lot more subtle, but that doesn't mean that they should be ignored. It would be one thing if online shufflers were flawed in the same way that irl shuffling was, but in reality, you're just substituting a certain set of problems for a different set of problems, which accomplishes nothing. It's better to just eliminate the problems entirely, even if it makes your game technically less of a simulator. Irl we should practice good shuffling methods too. A number of articles have been written on the subject.
  4. I don't why players couldn't just say "hard left," "slight left," "mid," "slight right," or "hard right" when they place a card on the field. Once zones become important, players will naturally become more careful with placement.
  5. There are 40! permutations of a 40-card Yugioh deck (approximately 1047), so having a good periodicity is going to be pretty important. There might be some other factors that come into play as well, including whether the RNG is done client side or server side, and whether or not the order of cards in the deck is at game setup or as cards are drawn. An expert in this area would need to look at the full code to identify any potential flaws. I recall that BLS didn't want to show any of his RNG code for fear that players might find a way to exploit it somehow. Of course, if his RNG was good, it should be impossible to exploit, which demonstrates a lack of confidence on his part. Note that I come from poker, where players will an many cases simplify refuse to play on your site if you aren't willing to make your RNG code public. I do agree though, that most players don't actually care about RNG, and even if it turns out that DB's RNG does not pass certainly scientific standards, it still won't be flawed in a way that is obvious to its users.
  6. Not really, because most people don't actually care about RNG as much as they say they do. Also, DB could easily implement Mersenne Twister if it wanted to. The code for it is only about 30 lines.
  7. I had a job interview today, and one of the topics that came up was random number generation. I, like most other people who had studied computer science, assumed that whatever standard number generation packages came with a language were fine. It turns out that this is not the case. The bad news is that I was naive and not able to answer any of his questions about RNG; the good news is that I learned something new today which is going to prove to be pretty important to whomever is designing future TCG simulators. "Old" methods of RNG that are implemented in the libraries of C, Java, and many other languages (including Flash) are not sufficient and do not pass most modern tests for randomness. There are issues with periodicity (how long until a cycle of random numbers ends and starts repeating itself) and correlation (the fact that successive random numbers are determined by previous numbers). In "Numerical Recipes," the author writes that about 20% of all scientific simulation work is suspect due to poor choice of RNG. One of the best RNG methods that we have currently is the Mersenne Twister method, with periodicity of a whopping 219937 − 1, which has 6002 digits (recall that the number "googol", known for its large size, only has 101 digits). In other words, for all practical purposes, you will never run into any periodicity problems with Mersenne Twister unless you plan on generating numbers until the sun explodes. While there are certainly some disadvantages to this method (for example, it's not very good for cryptography), it is great for generating high-quality psuedo-random numbers that a shuffling algorithm relies upon. It's purely coincidental, but Lackey happens to be written in C++, which by default utilizes the Mersenne Twister method for RNG. On the other hand, flash-based programs like DN and DB are written in ActionScript, which does not use the Mersenne Twister method. Thus they are susceptible to potential problems with the shuffler, unless the person who designs the software goes out of their way to implement a superior method of RNG (unlikely, unless that individual has studied numerical analysis and scientific computing). Discuss? lol
  8. Of all of the things that could possibly kill Yugioh, link summons is far from one of the most likely.
  9. I'm not a mind-reader, so I don't know what people's plans are. I do plan on keeping LackeyDGZ up to date, though. Realize that clearly the owner of DB doesn't know what's he doing if he still hasn't been able to keep a server stable for more than a few hours. The longer that he waits to bring it up, the more time that just gives NAS to send him a C&D, which we already know that they will. So there's a good chance that DB is going to be short-lived. One of the advantages of playing Yugioh on a plugin-based program like Lackey is that it is structured in such a way that we are not breaking any laws. Lackey is perfectly legal, and a list of Yugioh images being stored on a remote server is also legal (see yugioh.wikia for example).
  10. JC. is making a joke about Peter Cheng, infamous for using Gadgets to cheat in 2010 to use search cards that weren't gadgets, most commonly Future Fusion. In most of the formats that I played gadgets in (2007-2010) I would argue that Gadgets were much more of an aggro deck than anything else. Aggro is definitely a strategy that has not existed in Yugioh for some time, and probably will never exist again. There is still some hope for control, though.
  11. @Starlight Warrior the term "control" as it is applied to TCGs was first used in MTG to indicate a strategy that tries to win a long-term battle of attrition by accumulated card advantage. When you say "control" what you mean is "combo": a deck whose goal is to execute a sequence of plays that either cause the game to be won immediately or place the game into a state where it is nearly impossible for the other player to have any hope of winning (even if it may be several more turns until the game has actually ended). In modern Yugioh, nearly every top tier deck is a combo deck. True control decks have not been seen in the meta for quite some time. I haven't played too much in the past 4 years, so I can't really say when we last had a deck that would qualify as control. Some decks were combo/control hybrids. Nekroz, for example, could Djinn-lock for a victory, or it could try to grind out the opponent with its brutally efficient ritual summons.
  12. Tbh I'm kind of jealous of Starlight Warrior. I would love to play in a metagame where Link monsters make Darkness Approaches seem appealing to even a few players, where we need Giant Trunade and more boss monsters to slow the game down, where too many traps are being played, and where playing a rank 4 toolbox deck guarantees you a 50% winrate. Sounds way better than actual Yugioh that gets played here on planet Earth.
  13. No, I don't remember the last time that they unbanned Giant Trunade given that it's literally never happened.
  14. Probably tbh
  15. It's not an admin cp setting as far as I can tell, already checked.