riadsala

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.

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.

I second this proposal! Troy - get your people to talk to his people. Rob - make sure both Bruce and Tom are available as panellists.

It depends how you listen. I usually listen to podcasts while on the train, or walking to and from the shops, so looking up background for a game isn't really convenient. Sure, I agree they don't need to walk all the listeners through all the mechanics.. i wouldn't want that! But at least a loose description (turn based? real time?) just to give the listener a loose frame of reference in which to interpret the following conversation. Sometimes, they miss that out completely. It would be like talking about Hearthstone for an hour without mentioning that it's a card game!

I think this is something of a more recent problem? At least, I'm more frequently finding myself thinking the same thing. Bruce and Tom are usually pretty good at backing up a little to explain things to people who might not be familiar with the setting (or, at least, that's my impression. I could be completely wrong!)

Yup. I find it depressing how little effort goes into accuracy into most games. Why contradict what little we do know, when there are huge gaps that could imaginatively be filed in

Oh, I know that it DID happen. But I was under the impression that it wasn't the norm. Or as easy as it is in the game. The AI always feels incredibly passive in agreeing to marriages. I am guessing that there was a fair bit of diplomacy and various factors playing into the historical examples you cite. It just feels weird that a lowly Irish count can court European royalty. Maybe an easy way of tackling this would be to more closely align marriage decision making to your prestige, and connections between the courts? So, sure, your humble Irish count and still still marry into germanic royalty, but he'll have to have a hell of a lot of prestige and great reputation?

Interesting post. On a similar (but more less well informed and sophisticated) note, one of the issues that has always bothered me is just how easy it is to marry your ruler off to your pick of eligible women. Was medieval Europe really so international in its family ties? I know there will be plenty of examples from history, but I'm pretty sure that not every king of Scotland was married to the top lady from continental Europe? It just seems a little odd how this is implemented. Should it really be that easy to sort out a marriage alliance with the ruling family of the Holy Roman or Byzantium Empires? I would be interested in people's thoughts (both those who know how to play the game well, and those who know history well).

Civ IV is also a very good game in its own right (especially when you consider the wonderful mods). Xenonauts is a very good modern (more more faithful than XCOM) remake of X-COM. Endless Space is a not too complicated space 4x that I quite like. It's not as flashy as Endless Legend, but has a tight rule set and shouldn't be too taxing for a laptop. Conquest of Elysium is a little ugly, but it's a small rougelikelikelike strategy game crossover set in the Dominions Universe and is worth playing if you can get past the lo-fi graphics.

They've never discussed it in much depth. If I do get round to covering it, it would be great to get the designer on (or a high-skill player) to give a different view point. War of the Ring is a fantastic game. To me, it's main strength is that it attempts to keep itself as simple as possible. Don't get me wrong, it is still a long and reasonably complicated game. But the design decision to resist the temptation to give the different races different stats was excellent. Sure, after a background in Warhammer and what not, it was a bit odd and, at first disappointing, that one elf = one orc = one dwarf, etc. But, it turns out you don't need fiddly stat lines to make the races feel different. It all works fantastically well.

A few eurogames in which the player out in the lead is penalised at every turn. Oh, another game that deserves mention is Solium Infernium. I haven't played it in ages, but I remember power being very hard to hold on to, and there were a lot of ways of getting back in to the game. It was very difficult to protect all your weaknesses, even if you were ahead, so there were plenty of opportunities for weaker players to mess with you. [These weren't really comeback mechanics though, as the winning player could use the same tools to win more, depending on character build etc. Just different routes to victory interacting with one another]

hey Sorb, we agree! That doesn't always happen I'm going to put this out there... I'm not sure I agree with it entirely, but..... if a game requires an explicit comeback mechanic in order for the players who aren't winning to enjoy themselves, then it probably isn't a good game. I like the Dominions and Street Fighter examples that have been given. Both work pretty well. And at the other end of the spectrum, Paradox games are in some ways superior to Civ as they games are often still fun to play even when you're not winning!

Hmm. An interesting topic. I found it a little frustrating that there was so little talk of a player using skill to get back in the game. I know i always mention Go, but hey, it's a good example. With games at a "average" amateur level, big swings happen fairly often, as it's easy to misread a situation, and it only takes a one wrong move for a group of stones to die, or an invasion to end up successful. There's no support for this hard coded into the game mechanics. It's just that the game is sufficiently deep to allow these things to happen. Also, the type of skill required for the opening, mid-game and endgame are quite different, so even at professional level, it is possible to see players coming back from behind. Another weird thing was the lack of discussion of enjoying the game for the sake of playing and learning. When I play a stronger opponent at go, I know that I am very unlikely to win. Instead, I try to play well and make him/her have to think a little. If I can give my stronger opponent an interesting game, then I count that a success. And I'll likely learn something (as losing to stronger opponents is a great way to improve ones own game, in many different games and sports). I am comfortable enough with my own skill level that I don't need the illusion of some comeback mechanic. There also wasn't a discussion of how often these miraculous comebacks should happen? If the long-odds victory happens a little too often, then the odds can't have been that long! Also, there's nothing wrong with admitting defeat and retiring from a single player or two player game once you have played out all the likely interesting options. Obviously, things change somewhat with social multiplayer games.

Hello, I'm might tempted to pick up Distant Worlds Universe in the sale. Does anybody else here play this, and can anybody offer any advice on how to get started? Are there good tutorials? I remember Fraser offering some advice on one of the podcasts.

True. Perhaps once you've weaned yourself of Star Wars, we can look into picking this up. One for the holidays I think

You said it had a-sync multiplayer? I may pick this up over the holidays for playing with friends. Is it too much to hope that it offers the same deal as Frozren Cortex/Synapse?

Just wanted to drop by to +1 Martin Glaude as a fantastic guest. Make sure you invite him back! A good episode all round, I even found myself agreeing with Rob more than I usually do! [in particular, the conversation at the end of the game about what CiV lost when it decided to remove large empires from the game]

I'm all for a patreon, but I dislike the idea of exclusive content. The public face of Rock Paper Shotgun has noticeably gone downhill since they started their supporter-only posts. [They've made a few of these articles public, and they're excellent - but the normal content has suffered]. I would be worried that this would happen to 3MA. As I like the podcast. I'm not really sure I can justify spending £60 a year on it though, as this is more than I spend on games these days (I'm working through games via steam sales.. for example, I'm currently playing Dragon Age Origin and Napoleon:TW). And there's a whole load of other great free content that I try to keep up with. Mental note to self: go crazy with Patreon when I finally get a permanent job)

I don't think so. I don't even think its meaningful to talk about pvp and single player (or comp stomp) as being at all similar. As I've said before, I've been learning how to play Go (a bit like chess) properly, and it's probably broadly similar to trying to get good at an RTS, There's a huge learning curve and it takes a lot of hard work to reach a level at which your a competent player. Go is also played professionally in Korea (it may be related to why they were so quick to embrace e-sports), China and Japan. If i want to get better, I have to study, play games, review my own games to find my mistakes, study professional games, do exercises, etc. All that hard work which Rob is so dismissive of. But the pay off is huge, as you end up with a really engrossing game. Single player strategy games are nothing like that (for me at least). I kid myself that I'm playing a deep strategy game when I play Civ or TW or whatever, but really, its what I do to relax and unwind. I just make some fairly shallow strategic decisions and push units around a map. There's no trying to out think the opponent, or reading ahead to check what works and what doesn't. This isn't a value judgement: I enjoy playing these games a lot. And I don't want them too simple (DoW2 was a big let down for me as the tactics felt really one dimensional), but I know that I'm not going to invest the time to get good at playing them in multiplayer as I don't really have time. I guess this is a round about way of saying that I agree completely.

Thanks for reminding me! AI War is great, I really need to play it again sometime.

Howdy people I recently read a fascinating article about Go. One quote stood out to me: "However, the point of playing is clearly understood as not that of winning games (when Go is played properly, you lose about half of your games), but of exploring the possibilities to be found in particular arrangements of stones." I thought this was pretty interestingly in relation to some of Rob's recent comments about SC2, and how he finds the ladder really unsatisfying as you're always playing against people close to your skill level. I think, in general, people perhaps are more likely to have this mindset when playing single player games (campaigns in CivIV, CK2, etc), but as soon as they have an opponent, they find the game stressful and get themselves into a really competitive mindset. Some interesting psychology at work? I've certainly found Go, and other games, more enjoyable after switching to the "explore possibilities" mindset. I still have a clear motivation to get better at the game... the stronger I get, the more sophisticated "conversation" I can have with my opponent through playing the game. I can say things (ie, play a stone) that have multiple meanings and repercussions, and I will be better at visualizing the potential futures that are generated by any given action. That in itself is pretty satisfying in a kind of abstract way. I should add that I don't think this is specific to Go. Go really really is a competitive game, and is played competitively at a very high level by professional players who earn a lot of money.

I'm puzzled as to why this show was done before any of the panel tried multiplayer, given that seems to be the main focus of the game? Or is the game being sold on its single player?

I would say that I enjoy winning against opponents who are equal (or above) me in terms of skill and/or technique. This applies to single player games for me too... if I just wanted to win at Civ all the time, I could put the AI on easy, hit End Turn a few hundred turns, and the victory is mine. But it's a completely hollow victory and I would feel as if I've just wasted a few hours of my life. With Go, I don't enjoy playing weaker players so much. I'm happy to play them, as they will learn from the game, etc. But playing against an even/stronger opponent player is much more interesting, as they will challenge the points you try to make, and the "conversation" is far more nuanced. I'm not sure if these ideas apply to all games though. From what I've read, it sounds like players who get stronger at Paradox games end up simple breaking the game mechanics.

I'll give it a shot when I get a chance (just now the PC is getting repaired). On the topic of titans, I also need to give Sins:Reb another shot and see how that turned out. I didn't really like the balance or AI with the release version (thought it was a def step down for a brilliant series)

I thought Planetary Annihilation was the spiritual successor to Supreme Commander which was the spiritual successor to Total Annihilation?