legend

Member
  • Content count

    20,297
  • Joined

  • Last visited


Community Reputation

1334

About legend

  • Rank
    Dreams of Electric Sheep
  • Birthday 11/05/1983

Contact Methods

  • Website URL http://www.onesouthdesign.com

Profile Information

  • Gender Male
  • Interests AI, Machine Learning
  • Xbox Live Alerus
  • PSN ID Alerus
  • Steam ID alerus

Recent Profile Visitors


32,779 profile views

legend's Activity

  1. legend added a post in a topic Back to the Dorm - Who else?   

    I last lived in a dorm (or on campus living) 9 years ago.
    • 0
  2. legend added a post in a topic What aspect of game creation would you want to take part in?   

    For Geralt or Shepard, am I right?
    • 1
  3. legend added a post in a topic What aspect of game creation would you want to take part in?   

    World simulation and AI. I'd also probably like to direct some of the gameplay design because I have a general interest in the philosophy of what makes gaming fun.
    • 0
  4. legend added a post in a topic Nvidia DX12 drama continues as Ash's dev explains more...   

    How about that. We'll see how the next iteration of cards go. Maybe I'll find my way back to AMD again!
    • 0
  5. legend added a post in a topic Warshift: the sci-fi RTS-ARPG hybrid made by one guy that you've got to see   

    1 guy? That's damn impressive.
    • 0
  6. legend added a post in a topic GRIP (Rollcage remake!) - Kickstarter   

    Makes sense.
    • 0
  7. legend added a post in a topic Divinity: Original Sin 2 Kickstarter is up! Uh   

    First, I think you're underestimating how much money the game brings in versus development costs and at some point they should get some of the profit. Second if they're trying to do multiple projects, funding from the profits of one game is going to be even harder. Third some of the profit is going to funding the game (and their other projects), but this money will allow them to expand it on top of what they would do regardless. Fourth, a few million dollars really isn't that much for a game budget. Fifth, it's not like people are funding it and then buying the game. They're generally getting the game at a discounted rate with a delay while ensuring the game is even better, which seems like fair deal. Sixth, it's not like these people are greedy; they're giving away the EE of the first to all existing owners for free, which includes very significant expensive changes (like having it entirely voice acted, updating the models for all environments, adding controller support and split screen multiplier, changing the story and quests etc.).
    But if you're just trying to get a rise out of us so be it
    • 0
  8. legend added a post in a topic How should game scores be assigned in reviews?   

    A recommendation is what you think someone would enjoy/prefer. "I think you should play this at some point" is rather wordy for the summary when  you could just say "Play" with the understanding that the score is a recommendation. If really necessary the "Play" or "Play Now" label could have this clarification in an about somewhere, but do you really think people's default assumption from seeing a score like "Play" is going to be "this person is trying to spread their opinion as dogma and isn't just making a recommendation"? I think only the most juvenile of individuals try and claim authority in their opinion about entertainment and we shouldn't be expecting the worst from what is meant to be helpful.
    There isn't really much a of high score with three labels. I'm also not sure people will be more or less inclined to read a full review if there isn't a score. Do you have any evidence of that? Even if there is though, should we care? I don't think our goal should be to try and manage people unable to make reasonable judgement from a review and score; that seems like a losing proposition. Nor should we try and force people to read something they otherwise wouldn't. Instead I think we should focus on providing reasonable people information tools.
    You might question then why I suggest three labels if not to try and manage the unreasonable if that isn't my goal. My response to that is that I don't think there is a good way, whether you're reasonable or not, to provide fine grained numeric scores to games. There are too many dimensions and qualifications along which to analyze games.
     
    I'm cool with a comments section though. That could potentially even feed into the boards in a natural way (though anonymous users might be hard to manage with that).
    • 0
  9. legend added a post in a topic Final Dragon Age Inquisition DLC due Sept 8th, trailer here   

    Hmm This one looks like it might be good. I've waiting for solid story DLC and first that really gets me interested is the final one?  
    • 0
  10. legend added a post in a topic How should game scores be assigned in reviews?   

     
    This was back in 2009, so maybe the technology is now better able to find patterns, but I was at a session at an academic conference that was discussing collaborative filtering algorithms: algorithms that predict what a user will rate something based on how other people rated it. At the time, their conclusion was that due to the variability in 5 points scales they couldn't find any more signal in a 5 point scale versus a 2 point (thumbs up thumbs down) scale. They said users often preferred using a 5 point scale just because they did, but that there wasn't really information to extract. Rotten Tomatoes also seems to work fairly well and its main stat is the 2 point aggregate; same with Steam user reviews.
    Ultimately, 5 point is certainly better than 10 point with tenths, but think even there it starts getting into nit picking for things like 4 versus 5. Personally, I find this especially hard when trying to rate indie games along with AAA. Objectively, they are at a disadvantage with their engine, voice work, etc. Does that mean indies should almost never get a 5 star? Seems hard to manage to me. But Pass, Play, and Play Now: those I feel like we can assign without worrying about how indie games fit into it and making it ambiguous.
    That said, you also mentioned 5 point with strong definitions, so if you can suggest a good set of clear definition that are not the normal "star" rating, I could be open to that.
     
     
    Having the user ratings incite a more full "official" (whatever that means here  ) analysis is an interesting idea that could be worth doing. In particular, if a game is polarizing, that might be a good metric to get a deeper one.
     
    One other advantage for a three or two point system is that we can more easily express the frequency distribution over those values, which I tend to think is more useful than an arithmetic mean for things like these.
    • 0
  11. legend added a post in a topic How should game scores be assigned in reviews?   

    I don't think we're going to change how people reason, even if they're taking scores too seriously as something to champion. I'd prefer to provide the information that is useful for those of us who know how to use it and the rest of the community is going to do whatever they're going to do. '
    Ultimately the point of reviews to enable a person to decide whether to get a game, so I don't think there is anything wrong with making a recommendation that goes along with the analysis plain and obvious. But numeric scores I think are simply more confusing and turns things into a nearly impossible to manage pissing contest between games, which is why I prefer the 3 category system that I suggested.
    • 0
  12. legend added a post in a topic How should game scores be assigned in reviews?   

    It can certainly serve that role too. A discrete "score" (using scare quotes since I don't think we should use numbers) is more easily index-able though, so why not both?
    • 0
  13. legend added a post in a topic The Witcher 3: Wild Hunt (PC, PS4, XB1) discussion thread   

    Yes, you can continue to play.
    • 0