Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Antsan]

The main reason why I think it's crucial for your game however, is that it allows some really awesome interactions. Don't address the lord by his proper title? instant rejection. What if the NPC asks you about your feelings on their rivals? you'll want to know that information and say the right thing. Use the proper greeting for that particular religion? gain a bit more leeway with your questions. Whilst that can be done by picking options, parsing is really easy and I'm absolutely positive it adds a solid layer of immersion. I'd strongly, strongly recommend it if you can.

I doubt that parsing it actually is easy. The player could be start asking perfectly valid questions (from their perspective and even from the perspective of the designer) that the bot simply doesn't understand, because the player is using synonyms of words unknown to the bot. That's way more immersion breaking than an option system in my eyes.
I also liked it perfectly well in the Realms of Arkania games.

[Robsoie]

I'd suggest instead a sort of chat bot for the user to converse with, maybe with some option prompts as you already have. I did this for a project I was working on and, whilst the rest of the game was awful, that worked very, very well.

Chat bots normally don't work well as the user can literally ask anything and the chat bot doesn't have a built in 'personality' (in fact, one that 'passed' the Turing test did have a built in personality of sorts https://en.wikipedia.org/wiki/Eugene_Goostman) so they seem lifeless. More than that though, with a game based chat bot, you're strictly 'in universe' - you don't have to worry about the bot not knowing things, as it should only know parts that are relevant to it. If you ask it a question it doesn't like, it can get irritated that you're wasting it's time without it seeming as though it's just covering up for a lack of sophistication.

They maybe don't work well for everything, but chat bots can be rather entertaining
https://www.youtube.com/watch?v=WnzlbyTZsQY
https://www.youtube.com/watch?v=W2JmbgbXX4E

[Retropunch]

I doubt that parsing it actually is easy. The player could be start asking perfectly valid questions (from their perspective and even from the perspective of the designer) that the bot simply doesn't understand, because the player is using synonyms of words unknown to the bot. That's way more immersion breaking than an option system in my eyes.
I also liked it perfectly well in the Realms of Arkania games.

Granted, it doesn't work perfectly all the time, but parsing really isn't too difficult to do and is very, very well supported in Python (https://wiki.python.org/moin/LanguageParsing). I had this up and running in my project (http://www.bay12forums.com/smf/index.php?topic=137403.0) and it worked fine in my own little universe - I'm nowhere near the level of competency of Mark, so I'd imagine he'd be able to come up with something truly special.

You can easily fill over any gaps by coaching the player a little bit or/and using autofill (or a list of suggestions when you start typing). However I honestly never found I needed that - as long as you're asking stuff 'in universe' then you're by nature quite limited in what you can ask but you don't feel to be.
In terms of synonyms you can get issues, but it's quite easy to set up a synonym list using a standard dictionary lookup. You can also fit in a cheeky cover method, where if you ask about something nearly right, the NPC tells you it's knowledge: 'well, I don't know how old the Red Church is, but I do know that the Red Church has a secret basement'. This can be used for pretty much everything - again, you'll get a few bits where the NPC overshares, but that shouldn't be much of a problem.

I'm not saying it's a very simple thing to implement, but URR has always been about breaking new ground, and I honestly think there is so, so much you can do with this that the normal option robot doesn't allow for.   

[Ultima Ratio Regum]

Ok, this is a really interesting discussion (glad you like everything else, though!). Sorry for slow reply, last week and this coming week are super-hyper-busy.

But, my initial thoughts are: no to chatbot, but this chat version will be sufficiently different to any other option-robot that it'll be interesting enough. Reasons:

1) Technical. I've never done anything with chatbots and don't really know anything about them, and I'm wary of downloading a massive new library with stuff to learn. In the short term this is exacerbated by the fact I *need* to get this version out before the very start of August, so I just don't have the time necessarily to learn this all. I also feel... you're right that I'm trying to innovate in lots of ways, and although I do use a library for printing characters and whatnot, everything else I've tried to build some scratch. It's possible I could try to build a chatbot myself (!!!) but I'd rather not download stuff for it and then build on that. Building from scratch would be a huge investment, but so would downloading-and-then-modifying.

2) Speed of production. A chatbot will take me so much longer to produce than this system, but I'm honestly not sure it'll be all that much better. Although I'm (obviously!) not allergic to long development times, I am approaching the conclusion of this release, finally, and I really don't want to slow things down any longer, or add another massive task. I mean this both in psychological terms, and in "getting the game out there for people to play with" terms.

3) I honestly don't like chatbots. I just... weirdly, I think they show their artificiality actually more than the option-robot does, or at least, the type of option-robot I have in mind. Also - and this is close to the technical reason, but a little different - I think it would be extremely hard to program a chatbot to drop the kinds of hints/clues, at the times I want them dropped.

4) Would it get annoying? I think it's one thing to quickly select a greeting for many conversations; I think it's another thing to type out a greeting anew every time. Conversations are a huge part of the game, so I think having something a little quicker and speedier will surely be useful.

5) I think the system will be so unlike most option-robots that it will still be very distinctive. a) You have a massive wealth of options and no explicit clues towards what you should say, or implicit "suggestions" for what you should say (i.e. games tend not to give you "pointless" options, but URR will effectively do so, but it's up to you to decide which are useful/pointless), b) you cannot exhaust all options and have to carefully choose options, c) you'll be speaking in and moving between dialects and different styles to say the same things. So (as I see it, anyway) the only thing "similar" to other option-robots is basically just having a list of options - but the nature of that list, and everything else, should be so distinctive that I really don't think it'll be an issue! I just... yeah, I just don't think a chatbot is the way to go here. I'm doing everything to make the system different, but I'm concerned about technical questions, the loss of "control" that a chatbot system would offer, and all the other things above.

Basically, I think: what if we can get SO many questions you can ask, that it's basically indistinguishable from the possibility space of what the player could ask to a chatbot?

^ that's what I want to aim for.

NOW! I won't rule out the possibility of altering it in the future. But for now, I do think this system, with the many unique/distinctive aspects above, is the way to go; but *obviously* as always I massively appreciate the critique, and it has made me feel I should look for some other ways to make it more distinctive, which I will ponder. But I just don't think chatbots are the answer!

Actually, in truth... a sixth reason is that I have signed up to either show URR and its conversation system, or give a talk on the conversation system, at almost half a dozen venues in the near future, so I actually need to have it show-able soon! 

[Ultima Ratio Regum]

This week has been absolutely bonkers with academic commitments, personal commitments, and travelling all around the country, and I just haven’t had time to program very much at all. However, there is a major question I’d like some input on, which is what this entry is going to be about – and, indeed, it’s actually important to resolve this question before going too much further.

Learning Dialects

I’ve got the conversation window opened up. You can see yourself and the person you’re talking to, browse the appropriate options, and access some special windows for challenging the person you’re talking to, exchanging or buying/selling items, etc. The next stage (this is the task for this and next week) is to add in the full database of possible sentences, add in a data structure to allow new sentences to be easily added (by me in the future) and easily “unlocked” (e.g. when you learn about a painting, you can then talk about that painting), and in order to do this, I need to think about the other half of the conversation system: being able to ask these questions in appropriate dialects. We now therefore face a rather challenging question: how does the player learn a dialect? I see several options, and I’m still not quite sure which one to go for.

1) You can only say a sentence in a dialect once you have heard that sentence. This would be the simplest; I’d add some colour variation in sentences said by the other person in a conversation to show whether a) you already knew that dialect bit or b) you were hearing it for the first time.

2) You gradually acquire % knowledge, and it is distributed randomly across topics. This could have a range of methods: maybe it’s the number of conversations you have, or the time you spend in a particular nation (though I’d have to make sure that couldn’t be exploited by just running around in the forest, so maybe settlements only!), or something else; as time goes by you can say more and more in that dialect.

3) You can acquire books, or blocks of knowledge, which teach you *everything* in a dialect for *certain topics*. So, for instance, a book called “Laws and Finance in the Nation of Mugoppe” will, once acquired, allow you to speak like a native on the “Laws” and “Money” topics; similarly, perhaps certain characters can be encouraged to tell you everything about a certain topic, which then triggers a little under-the-hood acknowledgement of the same thing, i.e. you can now speak on that topic as a native.

4) Some combination of the above?

5) Something I haven’t thought of?

So, from readers, may I request: can you think about these options, and any potential fourth options you might conceive of, and let me know your thoughts? The basic requirements are for there to be a truly MASSIVE volume of potential sentences, a volume of sentences which can be added to as the player discovers more about the world, and a “learning” system that is interesting, distinctive, and can be somehow “distributed” throughout the world – i.e. you will always learn languages gradually, rather than instantly. This might be fast if the player specifically sets themselves to quickly learning a dialect and doing X (the learning process) a lot, but it should never be instant; there needs to be the option for the player character to know a tiny part of a dialect, a lot of a dialect, some of a dialect, etc.

There’s also  an additional exciting technical question of how to code and store it all… but my intention is to get to plan that out in the next few days once I have some idea of how it should all work!

Also, for this release, I think the player will have full access to every dialect, but in the very near future (0.9 or 0.10) that will change, and the player will have to learn each form of speech for each dialect via whatever means we decide upon.

Chatbots or Robots

A very interesting idea came up in URR’s Bay12 thread this week about whether URR could perhaps instead use chatbots rather than the “option robot” approach. A very strong argument was made for it, and I did actually seriously ponder it for several days, but in the end I decided it wasn’t the right move. Succinctly, my reasons were:

Technical/speed – I’ve never used a chatbot before, and it would require a potentially lengthy process of up-skilling, and I’m really desperate to get 0.8 out asap.

Clues – how easy/hard would it be to program a chatbot to “drop clues” of the right sorts at the right moments in the right context? I have no idea, but I don’t think it would be easy, and given the above point, I don’t know where I’d even start!

Annoyance – a chatbot means the player has to type things, and selecting greetings is one thing, but typing out a greeting every time would quickly get quite infuriating I think. Don’t get me wrong, it would be super cool if the player learned things rather than the player and the player character learning things, but the corrolary there is that the player would have to type everything.

Uniqueness – I already think the system I’ve proposed here will be very different to any other “option robot”, although I am certainly open to further ideas about how to further differentiate it. This is because: a) You have a massive wealth of options and no explicit clues towards what you should say, or implicit “suggestions” for what you should say (i.e. games tend not to give you “pointless” options, but URR will effectively do so, but it’s up to you to decide which are useful/pointless), b) you cannot exhaust all options and have to carefully choose options, c) you’ll be speaking in and moving between dialects and different styles to say the same things. I feel the combination of those three will be very distinctive.

So, I decided not to go to with the chatbot, but I thought it was an interesting enough idea to be worth posting here, in case anyone else has some other ideas, or some other ways to make the system even more distinctive (but you’ll have to be fast, coding starts this week!). Of course, had I elected to go with the chatbot model, I wouldn’t have needed to think about the player character learning dialects as discussed above in this entry, since it would be up to the player themselves, but for the reasons above, chatbots just don’t seem like the best solution for URR.

[Scoops Novel]

Glad u weighed it up, leaving the rest to retropunch .

Dialects could be languages, strong accents, slang? Divvy it up. Slang you get on the streets, languages you stumble through with books or translators (and maybe practice with your travelling companions), accents you get used too.

I like the "you know it once you hear it" idea. We will inevitably be kidnapped/run away to/loose our books within somewhere with a strange tongue, and guessing the right phrase will be hilarious. Gestures should be included too.

Memory is the last thing. If it's a uncommon phrase you haven't bothered with for months, low chance to succeed. If you use a book, stay out of fights.

[Brightgalrs]

A variant of 1)/combination of 1) and 3):

The sentences you hear for the first time is scrambled or blanked out. You then have to ask the person who says says it what it means, which unscrambles/reveals it. This plays into the "conversation interest" mentioned earlier this month. An NPC who is asked to clarify every-other utterance will very quickly loose interest in the conversation. So to stop yourself from annoying every NPC you come across it should be possible to look up sentences in books to learn their meaning.

[Retropunch]

Very interesting. I can definitely understand your hesitation, and whilst I'm very keen I can see the downsides. My only request would be to give the user as much agency as possible, and get the NPCs to ask questions back/steer the conversation! Most game conversations are just one way streets of the player firing questions at the NPC until they eventually give out the information that the player needs, whereas in reality there's a lot of back and forth, wrangling for conversational leadership and a requirement to fulfill cultural needs.

I honestly believe that URR NEEDS that. Uncovering languages and the like is great, but if it's just used to fire questions out it quickly sucks the excitement away. I want - nay, need - to start off stammering through a conversation with a peasant about the way to the capital city, and with enough work and practice be able to trade my wit with the king's court, where as misplaced platitude, expressed belief or remark might get my head chopped off.

In terms of revealing the language I'd do it as follows:

• You have to learn the basics from books or a teacher. Alternatively this could be gained slowly with exposure
• You get (hidden) 'application points' from books and teachers, this is basically an xp pool that gets used up when you 'use' the language (mainly by speaking to users of that language)
• Once you use that pool up (although this could be done in tandem), you need to find books or teachers on specific subjects and study those - this means you can only speak about stuff out of the core topics once you've actually learnt about them them - again this, uses an application point pool
• You have a further 'mastery' variable - this goes up slowly, and once you get to 100, you automatically can talk about anything (meaning you don't have to go through EVERY topic to master the language).

Bonus points if you have a 'fluency' variable on top of that, and once that fills up they can't distinguish you from a native! This could be especially useful for more...dubious activities.

I believe this methodology would stop grinding behavior, and it'd also allow you to master the language without having to go through every single topic to be able to speak about them. It requires you to have to both find the resources to learn and apply them, and should end up being pretty natural (as well as having the added benefit of not having to learn stuff that you're not directly interested in). The hard part would be signaling when each pool of points is full up. A message of the 'you're ready to apply what you've learned' or 'you've exhausted all you can learn in this manner' would probably be ok, along with a few more pointers from teachers/books.

[NJW2000]

Something to note might be that grinding behaviour makes sense when it comes to learning a language.

Also, if URR has places in which more than one language will be spoken a lot, if there was some way that could be used to gain knowledge of a language faster, this would both be a bit realistic and present a more strategic gameplay approach than "pester random people".

[Scoops Novel]

If URR has places in which more than one language will be spoken a lot

Pidgin and loan words should be a thing.

[Daniel the Finlander]

I have a question unrelated to the last update: will there be more farms scattered around the countryside? Last time I played the map was surprisingly empty, consisting mostly of wilderness. The game's setting is based on the pre-industrial world, so shouldn't there be more farms and farmers? This little detail would make the game more immersive, at least for me.

[Ultima Ratio Regum]

Re: languages, great thoughts everyone. See the blog entry cross-posted at the bottom of this message!

I have a question unrelated to the last update: will there be more farms scattered around the countryside? Last time I played the map was surprisingly empty, consisting mostly of wilderness. The game's setting is based on the pre-industrial world, so shouldn't there be more farms and farmers? This little detail would make the game more immersive, at least for me.

Hmm, interesting question! Er... not in the next release, but I could certainly see myself adding this change. It would be a fairly small/easy one to make, so I'd definitely consider it.

---

This week's blog entry touched on lots of things (including my first live TV experience!), so the full thing can be read at http://www.ultimaratioregum.co.uk/game/2016/05/30/television-canada-dialects-talks-esports/, but here's the URR-specific bit, about dialects and whatnot.

Dialect System

In light of all the excellent comments last week, I have some idea of how dialect learning will now take place in the game. The overwhelming answer seemed to be “some combination of 1 and 3”, and I agree that this is the best solution. If you’re talking to someone about a topic and they make mention of a particular aspect of that topic, then you’ve heard someone say that particular thing in that dialect, and that’ll get added to your list. Alternatively, you’ll be able to buy books or ask someone to specifically tell you about extensively about a particular topic (for a price), both of which will add appropriate information to your database for that dialect. There were also five really interesting ideas raised which I hadn’t thought about, and here they are, and my thoughts on them:

General books? It was suggested that perhaps one could build up a reserve of “points” which are acquired through “general books”, perhaps, which allows you to talk in a dialect for quite a while without needing expertise, but once you exhausts that knowledge of the “basics” of the language, you then pursue more specialist information. I quite like this idea for the basics, and perhaps a system where you can choose where to focus this basic learning? It needs a little more thought, but I think it has promise. Verdict: Probably.

Event learning? The idea was put forward that your character can learn dialect information by attending particular events. I think this could be a fantastic middle-ground between learning language bits just by hanging around in a city, and learning from items – if you go to an arena you’ll learn how to talk about combat, going to a religious festival will let you overhear info about that religion, etc. I’ll have to think about how exactly to put this in to prevent farming/tedium, but I really like it. Verdict: almost certainly.

Fast forward feature? One comment suggested that you can choose to spend time in a city just learning the dialect as another means of acquisition. I think if I implement this you will not be able to pick up language just by hanging around in a city normally, but you will be able to choose to click a button to spend X time in a city talking to people, and at the end of that time your dialect proficiency will go up; spending time as a resource, therefore, which is always a good option in a roguelike. Verdict: Probably.

Learning basics helps you learn advanced? It was suggested that you learn advanced parts of a language faster the more of the basics you know, but I’m not sure how well that would line up with learnings in “chunks” through books, events, etc, but the concept makes a lot of sense. I’ll think more about how to add this. Verdict: Maybe.

Mastery? The idea came up that when you reach over a certain % of a dialect, you are automatically given the small parts you don’t have – so when you have 90%, for instance, everything is automatically filled in, since your player character can “work out” how everything else would be said. This makes a lot of gameplay and thematic sense, and I like it. Verdict: Definitely.

Also, I realize that I should make explicit the difference I’m intending here between “languages” and “dialects”! All the major civilizations in the game world will “speak English”, but different dialects of it which produce different phrasings, different ways of speaking, etc. By contrast, smaller tribal civilizations will speak their own languages, which will need their own systems I haven’t quite decided on yet…

[Greenbane]

Still following this intently! Keeps getting better.

I'll add a question to the mix: why must all major civilizations readily speak the "common tongue"? Might be interesting to have different languages across the world. It seems like a bit of a waste to ponder and implement all these learning methods only to apply them to minor dialects. Well, unless those dialects aren't really minor, and thoroughly colour the variant of the common tongue a given nation speaks.

[Retropunch]

Still following this intently! Keeps getting better.

I'll add a question to the mix: why must all major civilizations readily speak the "common tongue"? Might be interesting to have different languages across the world. It seems like a bit of a waste to ponder and implement all these learning methods only to apply them to minor dialects. Well, unless those dialects aren't really minor, and thoroughly colour the variant of the common tongue a given nation speaks.

Completely agree - I'd be disappointed if there aren't a number of main languages, especially as language is so tied into culture. Dialects are good and all, but there are very few dialects (at least in reality) that you'd have to actually learn from a book or a teacher.

I'd suggest that there be a relatively small number of major languages, and the player would know a number of them to start with (depending on their house). On top of that you'd have a lot of smaller ones for tribes and the like. Whilst it might be hard work at this junction, I don't think you need to go too far into making the other languages 'other' though - a simple substitution for English words would do and you could always improve on that if you wanted to later.

[Ultima Ratio Regum]

Still following this intently! Keeps getting better.

I'll add a question to the mix: why must all major civilizations readily speak the "common tongue"? Might be interesting to have different languages across the world. It seems like a bit of a waste to ponder and implement all these learning methods only to apply them to minor dialects. Well, unless those dialects aren't really minor, and thoroughly colour the variant of the common tongue a given nation speaks.

Completely agree - I'd be disappointed if there aren't a number of main languages, especially as language is so tied into culture. Dialects are good and all, but there are very few dialects (at least in reality) that you'd have to actually learn from a book or a teacher.

I'd suggest that there be a relatively small number of major languages, and the player would know a number of them to start with (depending on their house). On top of that you'd have a lot of smaller ones for tribes and the like. Whilst it might be hard work at this junction, I don't think you need to go too far into making the other languages 'other' though - a simple substitution for English words would do and you could always improve on that if you wanted to later.

Thanks! . Hmm, basically the reason was that I haven't yet found a way to learn a *language* that wouldn't be tedious, grindy, annoying, etc. If I do find one, though, I do quite like the "small number of main languages" idea - that could be quite a nice balance between giving every nation its own language with the potential gameplay difficulties there, and the greater realism of having linguistic blocks across the world. I'll think about it!

----

Remember that technical stuff I talked about last time, and the dialect learning system we discussed the week before? This is going to be one of those rare URRpdates which is actually semi-technical, and I’m going to talk a little bit about how I’m going to implement and store the player’s knowledge of each dialect.

Dialects vs Languages

Firstly, as mentioned last week, I’m talking currently only about dialects – which for the sake of URR I’m currently defining as “a way of speaking English” – rather than languages, which in the URR context I’m defining as a totally different set of characters being used for speech, which might be unintelligible to the player character. The latter will be used for tribal nations (once I figure out a way to implement it that isn’t totally infuriating), but for 0.8 and all immediate future releases, we’ll be focusing on dialects rather than entire languages. It was suggested on URR’s Bay12 thread this week that languages could actually play a small role in the “main” civilizations of the world, i.e. the feudal and nomadic ones, by having 3/4/5 languages across the world as well as dialects, so that you have a lot of nations you can visit that speak your language (which would offset the potential annoyance of having exploration curtailed by language problems) but there are still multiple languages even in the large nations of the URR world. I’m pondering this option and haven’t reached any conclusions yet, but it’s definitely an interesting option.

Technical Specifics

So, I am faced with a bunch of interesting challenges here. I see these four in particular, and I have four solutions I’m currently implementing, with the target of having a working model by this time next week, and some nice screenshots to show off!

Firstly, a system is required that can both keep track of standard “sentence meanings” – such as [Standard Greeting] – and however many custom sentence meanings might be required within a particular game world, such as [Ask about the painting “Sunrise over the Empire of Nurnek”], which will obviously be unique to that particular playthrough. These should all be given to the player in the same way and at the same time, so that the player starts with just a bunch of standard conversation options, and then as time goes by and they learn more about that generated world, they’ll gain a massive range of other conversation options. My intention is that upon world generation, the player will be given a range of sets of the sort greeting_questions = [], farewell_questions = [], and so forth. These will then be populated with a standard set, but they can also be added to manually by the game as a particular playthrough proceeds; so encountering a work of art will append a new possible sentence about that work of art. In turn, each of these additional sentences will be part of an archetype, e.g. “ask about painting”, “ask about sculpture”, etc, and a “tag” on each will keep track of what each sentence means, so that when the player learns how to say [Ask about painting “X”], every sentence of that sort becomes ask-able. A quick mock-up of how this should look, if the player had clicked on the “Past Life” option:



Secondly, the game needs to track what dialects the player character can and cannot say each thing in. This has two obvious options – either the game tracks sentences and lists for each one the appropriate dialects the player knows, or lists the dialects and tracks the sentences that can be said. It became immediately clear that it should track the sentences and then record valid dialects for each sentence, rather than the other way around – if I tracked the dialects, then each dialect would quickly fill up a massive list of valid sentences which would take a long (“long” by roguelike standards) time for the game to sort through each time. Therefore, the game tracks all sentences, and in each sentences has a list of numbers, which relate to each civilization whose dialect the player can say that sentence within.

Thirdly, a method for storing dialects themselves is needed. I’m not currently completely sure about the correct method to do this, but it’s one of the things I’ll be thinking about this coming week, and then implementing the week after. If you look at the URR slides below, you’ll get a good impression of the variables that need to be recorded, and those need to be put into each civilization somehow. More on this later!

Fourthly, we need a method for sorting through certain conversation options that might come with a particularly large volume of viable options. For example, regardless of your dialect, the “Greeting” menu might only have a few options – “Warm Greeting”, “Greeting”, “Terse Greeting”, etc. By contrast, the “art” menu will end up including every piece of artwork you ever see, and if you reach the end of the game, and you’ve investigated a lot of galleries, mansions, cathedrals, castles… it’s going to be the pretty colossal list, and there probably needs to be some way for the player to immediately sort through them. Maybe options to sort by civilization, or start typing in a word and it tries to match that up? So once you know a particular piece of art called “Sunrise over the Great Southern Ocean”, you can type in “Sunri” and it’ll immediately offer you that option. There could be other methods, but I think some combination of those might be the best.

My intention for this week, therefore, is to populate all the option menus with an appropriate set of options, and to start figuring out how exactly to store dialects and therefore generate sentences. More soon! And here’s another mockup I put together for this week’s presentation:



Lastly, if you want the slides from my recent talks in Canada, head here! http://www.ultimaratioregum.co.uk/game/2016/06/05/dialects-the-technical-side/