riadsala

Members
  • Content count

    339
  • Joined

  • Last visited

Posts posted by riadsala


  1. That was an incredible game.

     

    in case anybody is wondering, it was pretty close. I think the estimated margin of victory was less than komi. And, I'm stunned that AlphaGo took more time to think than Lee. Truly incredible the amount of processing power required to beat the human brain.

     

    I expect the rest of the matches will be similar, but perhaps there is still hope.


  2. So Google DeepMind's AlphaGo program is playing the #2 ranked Go player Lee Sedol in a few hours. You can

    if you're so inclined. If it beats him, there's theoretically only one person in the world who it might not be able to beat.

     

     

    Yup. I'll be setting my alarm for 4am (UK time) to watch it tonight. My money (a whole £5) is on AlphaGo, but I'd love to see Lee Sedol win. I read earlier that Google have just given their biggest ever press conference about this (they're flown over to Korea, and there's a lot of interest in Asia generally).

     

     

    If AlphaGo manages to win, I doubt anybody can beat it.. it would be a huge coincidence for the technology to reach a level that is exactly between the current number 1 and current number 2 player in the world. And, Lee Sedol has 18 international titles to his name (I think that's only one player has ever managed more!).

     

    Whatever happens tonight, he still has better hair than the bot.


  3. I bought SmartGo, just to have something for on the run, and man I am bad. I knew I was bad, but not quite this bad.

     

     

    Don't worry about winning and losing just now. Don't even worry too much about trying to play "the best" move. Just get used to playing moves and your brain will automatically start to recognise patterns (as long as you're paying attention).


  4. Finally, if I had to give one piece of future advice also related to last week's episode, a group of relative amateurs guessing at the putative top-level optimization of a game that they've just learned doesn't make for the best design roundtable. If you look, for instance, on BoardGameGeek, there is almost no consensus over multiple threads about "optimal" strategies, except Ishida taking Kiyosu and Tokugawa taking Aizu in the first couple of weeks. Even then, the randomization of starting forces and card hands can make those actions sub-optimal, too. Woe betide the Tokugawa player who has to face four Uesugi blocks the first turn and reinforcements there afterward, making a march on Kyoto a huge risk. The genius of Sekigahara is how it demands flexible awareness of overall strategic situation while still knowing the line of best fit (a gradual consolidation of forces and territories over the first four or five weeks, followed by an advance along one or both highways and a climactic battle in the sixth or seventh week).

     

     

    Agree 100% It sounded like Rob has only played the actual game once (with the previous two games being him figuring out how the game works). It was a good show, but it did leave me wanting more - the difference between the game being good, and the game becoming a classic part of the canon is how it stands up to repeat playing when both players know what they're doing. Does the game eventually turn into a bluffing game, rather than a "strategy" game? As that would be cool an interesting. But perhaps the mechanics only support a relatively low skill ceiling.

     

    Anyway, sounds like a cool game. Maybe if I ever find somebody to play WotR with again, I'll think about picking it up.


  5. That AI training video mentioned this:

     

    https://www.gokgs.com

     

     

    yup. that's the main western server, and where I mainly play (tygem is where a lot of the Japanese/Chinese/Korean players play, but I wouldn't recommend a new player it to a new player). KGS is pretty good, fantastic community, but aging and slightly crap java interface.

     

    OGS has a much better web interface, but is quite new and the community is nowhere near as large. But it does offer correspondence games, so I play the odd game there. And DGS is another super simple correspondence site which I've used for years.

     

    If you want to learn the basics, then this site is useful.

     

    And there's a very good old proverb for new players: lose your first 100 games as quickly as possible. ie, don't worry too much about playing well, or winning. Just play. You'll start to recognize patterns and whatnot automatically.


  6. That's definitely preferable, though it can be intimidating to go ask for matches against humans.

     

     

    I can be, but it shouldn't be, and there's no (external) reason for it to be. So the sooner you get over it, the better.

     

    How about, I start: I'm more than happy to play correspondence games (ie, a couple of moves a day) games with anybody who asks. We can play either handicap games, or a teaching game (where I'll comment on your moves after you play them). I can also help tell you about servers, how to enter your rank, and find other beginners to play against.


  7. If you beat the AI you wouldn't learn anything. That the AI is actually good now means that you can try different things and see how it responds. For learning though, I'd pick up a program like SmartGo, work through problems, and play 9x9 games with the AI. The problems in SmartGo are ranked by difficulty and I think do a good job of teaching basics all the way up to a competitive level. It doesn't even burn up the CPU any more trying to come up with a good response for a move either, so if you're using a battery-powered device you'll get hours of play time in. Heck, I have it on my phone.

     

     

    Better yet, just play against other real people. And ask stronger players to review your games.


  8. man there's always a couple people round my level online when I'm on but none of them actually want to play/respond to my challenges at all. Finding a real promising looking multiplayer game but showing up too late to the community is such a downer :'(

     

     

    Well, Sorbicol has a copy... I'm sure we could find more people if we looked hard. RPS can be a good place to find people: https://www.rockpapershotgun.com/forums/forumdisplay.php?10-Game-Clubs-And-General-Sociability-Ok


  9. Same here. Although I think my score against Riad is something like Riad 27 - 7 Sorb so if Riad thinks he's rubbish I must be the pits.

     

    That's in Synapse. We're talking Cortex here, a whole different ball game [smug pun very much intended]


  10. If you go back and listen again I think both Fraser and Jonathan give a much more balanced appreciation of both what XCOM 2 is doing and what XCOM 1s flaws were. I think the issue is that some of the stuff David Heron and Rob are saying is just so far out of kilter with what would appear to be everyone else's experiences with both games that they stand out a lot more.

    I too would love an episode which looks at the evolution of the Firaxis XCOM games and the influence of The Long War mod (which at the end of the day is probably one of the greatest game mods ever made - and I include the likes of Counter-Strike and Fall from Heaven in that statement) on XCOM 2. Because I've just started a new campaign on Commander level and boy can you feel the influence of LW at that level!

    I would also dearly love that conversation to include Jake Solomon or Garth from Firaxis as well - or whoever else the lead designer is. We had that for both XCOM:EW (Jake and Garth) and for XCOM:EW (Ananda Gupta) - in fact over the last couple of months this seems to be something 3MA doesn't do any more - have those conversations with lead designers - for any of the games they are reviewing unless it's one of Bruce's more obscure wargaming titles. Now as unlikely as it is I'll ever play one of those games I usually listen to the podcast because it's fascinating listening to the designer talking about how they came to their decisions. I distinctly remember both Jake and Garth saying how pleased they were to be on the podcast too - how they all considered it to be gold standard and that everyone at Firaxis listened to it. Or words to that effect.

    Now I completely appreciate that Rob and Troy and Bruce are all very busy individuals, and that 3MA is very much a labour of love for all of them, but is it really that hard to get hold of those people any more? The last time I remember you guys getting a designer on to talk about their game (which wasn't a war game) was probably Will and David talking about Beyond Earth.

    Does nobody want to talk to you guys any more?

     

     

    Ha - Sorb, this is one of these rare situations in which we end up agreeing 100%. Chris Park would be another great guest designer to get back on and interview. Arcen may not be all that successfully (both in making "good" games and commercially) but they are certainly one of the most important and interesting strategy game studios around just now.

     

    Both the Kingdom and the Thea show would have been way more interesting if a designer had been invited on too. They were already good episodes, but I would have loved to have heard the designers responses.


  11. Maybe I'm oversimplifying or misunderstanding something, but I always felt a lot of cyberpunk (sort of driven by the "punk" part) was the opposite of competence porn.  Broken or near broken people stumbling their way to maybe a temporary victory in a system that was way more powerful than them. 

     

    Gibson, in particular, is very stingy with true success and more often it seems like characters almost completely fail, and then the system tears itself apart from their disruptive influence more so than their competence. 

     

    The success is just the punk ideal that absolute corporate power will be broken down by sort of haphazard disruption more than protagonists actually accomplishing specific goals/tasks.  That comment about neon and noir sclpls made resonates with this reading for me.

    Maybe that's just the type of cyberpunk I like, and so those themes stand out a little more to me.

     

    I think I agree (although I don't know that much of the genre). This is certainly the set up that I find more interesting. I'm starting to think that where my taste diverges from Rob's is that he's enjoys power fantasies too much for me. Sure, I enjoy winning, but I like to think that the winning was well earned (I'm either playing well in the game, or i've practise, studied and mastered the game's systems over multiple games). And without the possibility of everything going horribly wrong at any moment, the suspense goes and I feel like I'm a bit bored mindlessly hitting buttons. This is course, is all just my personal likes and dislikes.

     

    ie, one shotting an alien with 100% success in XCOM is boring. one shotting an alien with 50% chance, in a situation in which if you miss, your prized assult marine is going to be eaten, is a great feeling. And then, in a later mission, having a good sniper who you can depend on it s great feeling, as I feel like I've learnt it via playing well and keeping the gal alive for long enough. [And I can then play deeper strategies and get risk elsewhere]/

     

    it does make me think that "the role of power fantasy in strategy games" would be an interesting topic for a show. I think they're fine in some genres (rpgs, fps?) but I don't enjoy them so much in strategy games. I find playing a game with easy systems and/or dumb AI similar to playing a weaker opponent at go or hearthstone, and its just not all that satisfying (for me).

     

    But then again, maybe I'd enjoy Ck2 more if I embraced the power fantasy and opted to play as a King/Emperor more often. So I could just well be wrong.

     

    Maybe it all comes from being Scottish, and constantly losing in the football and rugby? Perpetual underdog!

     

     

    And regardless of all this, it sounds like I really need to play Invisible Inc sometime soon.


  12. I haven't played XCOM2 yet (I decided years ago that life is to short to be playing PC games on release -even if there are no bugs, there will still be balance issues) . But i have to say, a few of the panelists on this show appear to have played XCOM1 far too much, and forgotten what it is like for a new player. Speaking as somebody late to the party, and still trying to complete the game the way it is meant to be played (classic ironman), I feel nearly all the criticisms the panel make of 2 could easily apply to 1. The game hides a lot of important information from the player (a good, if frustrating, design choice I feel, as it makes the aliens feel more alien and more dangerous - if the game told you how to go about beating the aliens and what all their strengths and weaknesses were off the bat, it would lose a lot). 

     

    Overall, disappointing show. I'd much rather you all waited a month or two and then gave us some in-depth insightful commentary. 3MA is at its best when it's happily behind the times and discussing a game that's been patched by the devs, and played properly by the panelists. For example, even though I haven't yet played EU4, I still enjoy those shows as the game is mature and the panelists generally know exactly what they're talking about.


  13. Something that was thought to be impossible for decades to come was achieved and it's not impressive because nothing's impressive. How quaint.

     

    Well said.

     

    Yes - a lot of AlphaGo's strength comes from cNNs, which have already revolutionized computer vision. That doesn't stop it being an incredible demonstration of how quickly the AI field is developing, and how generalisable a lot of the tech is.

     

    A friend made a video here outlining how some of the systems work:


  14. I think that's the core of it. I find myself put off SC2 because I find it's reflexes and gambits instead of strategy. I want chess, SC2 gives me speed chess with a minute on the clock.

     

    Then why not play a turn-based game?

     

    or, if you like the added pressure, you'd comparison to chess with a clock was an interestig one. Are there any (computer) turn based strategy games that use that mechanic? Feels like an interesting design space to explore. Imagine a game a little like Unity of Command, but you have a 10 min (+ 30s overtime periods)  for a level?

     

    I am guessing that a lot of players would be put off by the idea of time limits on a tbs (which I get, as people often play to relax and may be checking their phone, etc while playing) which is a little odd when you remember that both chess and go are always played with some form of clock time when it is competitive. So it is already baked in to traditional turn based games!


  15. Also, I suspect making an algorithm that played more like an amateur Go player would be reasonably trivial now. Instead of training on pro games, you train on the millions of kyu game records on KGS, Tygem, etc. then simply restrict the number of possibilities investigated during tree search. I suspect that would lead to an AI that plays the same hopelessly optimistic mistakes that I (and many others) make from time to time. You could also go further and simply add in an "imaginary penalty" when scoring so the AI over-emphasis capturing its opponents stones and not losing its own stones.

     

    And, if we can get back to my original point about computer game AI... we don't need a Civ AI that can beat the best civ player in the world. there would be little point in making such a thing. But an AI who can match most players without cheating would be great. So we're looking at something around Prince/King level in Civ right (I have no idea what most Civ-fans play on, but I'm pretty sure that most strategy fans would be able to beat civ on normal after perhaps one practise game?).

     

    And you can easily mess with the heuristics to give the AI different personality (as I suggested above).... the heuristic they're trying to maximise doesn't need to be the victory condition. Perhaps you set the aggresive faction up so that it doesn't care about winning, it simply wants to maximise the number of enemy units it can kill over the whole course of the game. (this is just a daft example, as clearly this would lead to an AI that would avoid winning the  game at all cost, as once the game has been won, it can't kill enemy units anymore!). but you get the idea right.


  16. See I understand, kinda, your lack of enthusiasm for a computer being good at Go. But your counter examples I don't understand. Human language processing seems magical but (as you admit) it's so much simpler a problem than playing Go (and then, of course, looking up answers in a database is trivial).

     

    Is what makes artificial intelligence "impressive" mostly to do with its immediate applicability to becoming more human-like? If so, I still think the Go thing is a more substantial step in that direction... picking a move in Go is achieved through attempts at emulating how a human thinks about playing Go (because comprehensive possibility trees are impossible), while NLP can be boiled down to recording a long list of situational rules and applying them over and over. Both are pretty narrow aspects of "humanity", but the former is a lot more complicated and (until recently) elusive.

     

     

    You sound like you know what you're taking about.

     

    Watson winning Jeopardy is, IMO, way less an impressive feat than this. I read somewhere (sorry for not having the citation) that most of the Jeopardy answers are available straight from wikipedia. NLP isn't all that hard these days (sure, it's not a solved problem, but it works reasonably well a reasonable amount of the time).

     

    if anybody is interested, there is some commentary from a 9dan pro on the alphaGo games here: https://www.youtube.com/watch?v=NHRHUHW6HQE.


  17. That is kind of what I was trying to say. This GoAI feels more incremental, they made a GoBot. The tech may be really impressive, but to me it just feels like a fancier chessbot. As opposed to making an AI that plays more like a human in a game that allows for more complex player actions. 

     

    in which case you are wrong, and misunderstood my point.

     

    On the first point, a number of the top players in the world have said how surprised they are by how human AlphaGo appears to be in its play. For example, Ke Jie, the current number one player in the world has this to say:

     

    Q: Can you tell from the records that AlphaGo is not a human but a computer?

     

    A: No. I reviewed all five records but I didn't see the names of the two players. I had no idea which player is AI. AlphaGo plays like a human. It abandoned stones that should be abandoned, and made a concession when it is needed. AlphaGo plays in good balance and it is hard to imagine that it is a computer. Previous Go problems such as Zen, may, from time to time, make some nonsense moves. AlphaGo won't. It always plays important moves and that is strong.


  18. I don't see why this means that strategy games will have better AI. The issue with AI in games like Civ isn't a technology one - the problem is time, money and labour. Deep Blue was 20 years ago and I don't think any modern games have AI that compares to that.

     

     

    Somebody correct me if I'm wrong, but I believe that the methods used by DeepBlue were not as easy to generalise to other domains. It was a chess AI and that was that. However, the methods that behind AlphaGo are incredibly generalisable (cNNs and MCTS). You can download the caffe library for python and have a close to state-of-the-art library (it's used a lot in cutting edge academic research) for machine learning. So yes, you still need time, money and labour, and perhaps, more importantly, willpower to actually bother (see the terrible state of AI in Total War games). These tools are only going to become easier to use and interface with.

     

    The way it is more likely to happen is via mods, if a developer opens up the game enough. Then I can imagine some phd students (with access to their university's computing cluster) giving it a shot as a pet project. See this for a fun example of what people have done previously.  And this was in 2011, a year or so before AlexNet blew the previous state of the art out of the waters.

     

    There are plenty of people smarter than I am who believe that these recent developments will usher in a step change similar in scope and magnitude to the internet. Have a read. http://www.computervisionblog.com/2015/11/the-deep-learning-gold-rush-of-2015.html.

    Another really cool application would be for procedural generation. Hand the cNN a high definition map of the whole globe, and it should be able to easily synthesis maps, globes, etc.

    On a less excitable note: I am rooting for Lee Sedol in his upcoming match against AlphaGo! I'm not sure I'd place money on the result, but I may well try and watch some of it live.


  19. I disagree, for pretty much the reasons riadsala stated. The possibility space is not altered by the presence of randomness.

     

     

    Fantastic!

     

    like I said before, the difficult thing is to design a game which is fun when playing against a strong AI opponent. Civ and the ilk are games that are very much designed around the idea that the human player will win - the AI is there simply to through up some interesting challenges.

     

    With Go, I don't think I would enjoy playing a bot, as simply losing is no fun (although learning from stronger players is great). But I'm always more than happy to play against a stronger human player, as I can set myself the goal of making them think (or laugh at an unexpected move, etc).


  20. I think there is a major difference in complexity between reacting to another players moves and reacting to their moves, what random card you drew/roll you got and what the did along with what they maybe drew or will roll.

     

    Not really. If you think about how the new Go AIs work, they use monte-carlo tree search algorithm, which has randomness built in, as it is impossible to read out every possibly move. So when the AI "thinks" about a move, it then randomly simulates loads of possible games from that point (and this is often done in a pretty crude manner) and concludes that it wins xx% of the games if it plays that move. It then does this for other moves, and then picks the move that gives it the best odds for winning. [i'm simplifying a lot, but this is the general gist as I understand it]. So fundamentally, there is no difficulty in adding in stochastic elements such as dice rolls and card packs. At all. I actually believe that this would make it easier to match human performance, as people are so bad at dealing with probabilities and rational decision making [and we're talking AI for consumer games here - it doesn't have to be able to beat the world champion, it just needs to be good enough to give a good game to most people without cheating]. If you'd like, I'd be happy to share some of my research on human decision making and how people fail at even the simplest choices when probability comes into the mix. Or, if you've heard of the Monty Hall Problem, you'll have the right idea.

    Anyway, lets not argue past each other any more. We can agree to disagree.

    If anybody is interested in learning go, a good place to start is here. Then after that, you can play at OGS (easier to use, but new and small community), KGS (java, but probably the main western online community?) or Tygem (crazy big, and where a lot of chinese, korean and japanese players play. Maybe not a good place for a new western player to start!)


  21. Yeah, and to add to that: comparing 4X game AI (usually a set of if-then-else decisions written by a human) to a machine-learning system that has actually trained to win entire games of go (instead of losing entertainingly to a human player) is not exactly correct.

    I've also read exactly that kind of post when Deep Blue won but with Go as the "I'll worry when X happens" bit by the way.

     

     

    But the machine learning system is very general. it's based on cNNs which were developed for object recognition in computer vision. I think this is a really good step forward. 

     

    And, people are already doing some cool ML work with games (there's an AI for natural language processing that can play CivII reasonably well based on reading the manual. Which I think is very cool. And there's some cool work being done on Starcarft1. Exciting times. the only downside is that a lot of the experts in the field are being snapped up by Google, Apple, Microsoft, Facebook, so it will be hard for the much smaller game developers to recruit people with the know-how (not to mention university departments which are finding it hard to hang on to their best staff in these fields). So it's a question of whether these companies decide it's worth cracking more computer game AI projects for the free publicity. As, you know, there's a lot more money to be made with self driving cars and image understanding.

     

    Did you read about Deepminds previous project, learning to play old console games just through trial and error?

     

    but like I said, the trickier part will be making FUN AI.


  22. Games like checkers, chess, go etc are pretty solvable for AI. I will get a lot more worried when AI can beat champions at games that have a lot of chance involved. Being able to build a winning strategy out of a bad set of rolls or draws is much more impressive/terrifying. Daft Souls latest episode had a good discussion about how AI in 4X don't hold grudges or really think long term. You can declare war on them over and over, betray them, steal from them, but if an alliance looks good to them in the short term they will make it again.

     

     

    Trust me, computers are better than people at dealing with probability. I am speaking professionally as a Post Doctoral Research Fellow in Psychology. People are really bad at dealing with probabilities. Really really bad. And even worse at rational decision making when chance is involved.