Jump to content
Sign in to follow this  
Vorpal Sword

Want to help me write my dissertation on X-wing? Here's how.

Recommended Posts

TL;DR: I need archived championship game footage! The newer the better, but anything is worthwhile.

 

***

 

I have, through circumstances beyond my control, had to start the process of my dissertation all over again. That's the bad news.

 

The good news is that my advisor has strongly urged me to work on something I actually like, rather than something that will seem like work. So I'm writing my dissertation on X-wing, and specifically on putting together a comprehensive system for estimating point costs. Basically, I'm attacking MathWing from a new direction, and in a larger sense, articulating a need to adopt quantitative modeling in the design of tabletop games (using the language and context of product evaluation, which is part of the more general field of evaluation, which is the work I do).

 

Part of that is working with a programmer (hi, Geordan!) to build an X-wing "shooting gallery"--a program that, given a particular game state for two ships, will simulate each attacking the other as many times as you want. That will let me estimate all sorts of things and do some interesting statistics, which will result in (1) a rigorous, objective, and easy-to-understand estimate for the cost of ships and upgrades, and (2) some really pretty pictures. 

 

But in order to do that, I need to make some assumptions about what the game state looks like. More precisely, I need to know proportions and distributions: how often do ships have target lock available to spend on an attack? What about focus, or evade? How often does a ship engage at Range 3, and how often at Range 1? 

 

I can mine a lot of data from the tournament-results thread, but it will be even better to have actual game footage. So please help me find footage from championship matches! I'm looking for anything after the cut in a major tournament (particularly Regionals, Nationals, and Worlds); more recent is better, but I intend to go back to 2014.

 

I've found a lot already, but I'm sure there's more I'm missing, and I haven't found any kind of organized archive of links. So if you have a feed I can look at, please let me know!

Share this post


Link to post
Share on other sites

Im not sure I understand where you are going with this.  But I am always willing to help if I can.  I will check my pictures.  However, I am pretty sure most of them are of players, cards or setup before the first turn.

Can you mine youtube videos for the data you need?  Most (all?) of the Regionals here in California had at least one table with recording equipment (for Youtube or a stream).

Share this post


Link to post
Share on other sites

I'm writing my dissertation this summer and all I can say is ... I'm jealous.

 

I'm not.  He had to start over.  That sounds like hell, and not a minor hell, actual hell.  Actual hell is starting your dissertation over.

 

 

Also, here's 3 games from the KC regionals a year ago, so it's sort of local to you:

 

 

I'm sure youtube has a lot more.  

Share this post


Link to post
Share on other sites

Im not sure I understand where you are going with this.  But I am always willing to help if I can.  I will check my pictures.  However, I am pretty sure most of them are of players, cards or setup before the first turn.

Can you mine youtube videos for the data you need?  Most (all?) of the Regionals here in California had at least one table with recording equipment (for Youtube or a stream).

 

I'm looking for full videos of gameplay, so I can collect information on the state of each ship at the start of each combat round.

 

Yes, YouTube is absolutely the kind of information I need. (In fact, still images aren't very useful.) The problem is that searching "youtube x-wing" is drinking from a firehose! It'll be much easier to hear from (for instance) someone who attended a particular Regionals event and knows where the archived video stream lives.

Share this post


Link to post
Share on other sites

But in order to do that, I need to make some assumptions about what the game state looks like. More precisely, I need to know proportions and distributions: how often do ships have target lock available to spend on an attack? What about focus, or evade? How often does a ship engage at Range 3, and how often at Range 1?

 

A question to which the answer is neither easy to elucidate nor static nor universal across all ships or even all players.

 

You'd be better served trying to find an approach where you don't need to make assumptions about these things.

Share this post


Link to post
Share on other sites

 

I'm writing my dissertation this summer and all I can say is ... I'm jealous.

 

I'm not.  He had to start over.  That sounds like hell, and not a minor hell, actual hell.  Actual hell is starting your dissertation over.

 

So did I ...

Share this post


Link to post
Share on other sites

Have you checked the list of video channels in the stickied thread of useful resources? It should narrow the list of channels, at the very least.

 

Have you thought of using Vassal games that were logged and uploaded to Challonge? The latest Reddit tournament is here, A search yields a lot more, though they may not all have logs. The good news is that going through a log is much, much faster than trying to fast forward through a video.

 

I gave up halfway through my dissertation based on how hard it was to acquire data for my specific topic, I feel your pain!

Share this post


Link to post
Share on other sites

Are you looking at Vassal matches as well? I know there's been several vassal tournaments on reddit (and the current league, where you might be able to use the deep core and possibly core worlds in place of the cut at a tournament). There's also things like the team covenant aces and team covenant open tournaments.

Share this post


Link to post
Share on other sites

 

But in order to do that, I need to make some assumptions about what the game state looks like. More precisely, I need to know proportions and distributions: how often do ships have target lock available to spend on an attack? What about focus, or evade? How often does a ship engage at Range 3, and how often at Range 1?

 

A question to which the answer is neither easy to elucidate nor static nor universal across all ships or even all players.

 

You'd be better served trying to find an approach where you don't need to make assumptions about these things.

 

 

You're right, and I would--if such an approach existed. But it's impossible to simulate any aspect of gameplay without answers to those questions, so any approach that involves gameplay has to ask them.

 

And I can't imagine how to estimate costs for X-wing components without reference to actual gameplay. (Or rather I can, and that's how we ended up with Wave 1.) So while I'm willing to admit it's a potential failure of imagination, I'm stuck. 

 

The good news is that I'm not planning to say "here's what typical play looks like." Or at least I'm not just going to say that. It'll be more like "here's the average performance across 1,000,000 rounds of gameplay. Here's the subset of rounds 'played' against ships with 3 Agility and multiple defensive tokens. Here's the subset of rounds played against 0-1 Agility..." And so on.

 

 

Have you checked the list of video channels in the stickied thread of useful resources? It should narrow the list of channels, at the very least.

 

Have you thought of using Vassal games that were logged and uploaded to Challonge? The latest Reddit tournament is here, A search yields a lot more, though they may not all have logs. The good news is that going through a log is much, much faster than trying to fast forward through a video.

 

 

Yes, that's where I started. I'm trying to cast a wider net now.

 

And yes, I considered using Vassal logs. The problem is that the structure of tournaments/leagues differs, sometimes substantially, from typical OP tournaments. That creates some fuzziness around the selection criteria, which I don't like. There are also differences (in my opinion/experience) between Vassal play and tabletop play--not major differences, but a lot of them. I think I'll be better off excluding them, if I can.

Share this post


Link to post
Share on other sites

 

 

 

I'm writing my dissertation this summer and all I can say is ... I'm jealous.

 

I'm not.  He had to start over.  That sounds like hell, and not a minor hell, actual hell.  Actual hell is starting your dissertation over.

 

So did I ...

 

 

I'm very, very sorry.

 

 

I have mixed feelings. I really kind of loathed my old topic, and I like my new one. And while I had a lot of resources set aside, and a lot of time invested, those are sunk costs--I'm not going to get any further ahead by being distressed about them, and it turns out writing a dissertation on X-wing is actually pretty cool. I like drawing weird connections, and there aren't many weirder ones than between an early 20th-century experimental psychologist and an early 21st-century tabletop war game.

Share this post


Link to post
Share on other sites

I actually wondered what you found to be unacceptable based on Mjs work. Or perhaps where your dissertation will increase understanding.

Ive also wondered if Xwing comes into an inherently flawed system that cannot avoid creating systemic imbalances due to its design.

What happens as player flying skill approaches perfect play?

Does the fact that PS and pilot ability cannot be broken away from each other make it impossible to have useful mid tier stuff?

Share this post


Link to post
Share on other sites

I can't get the link at the moment, but StoneForge Studios just uploaded as much of the Denver Regional as they had to youtube.

Edited by jonnyd

Share this post


Link to post
Share on other sites

 

 

 

 

I'm writing my dissertation this summer and all I can say is ... I'm jealous.

 

I'm not.  He had to start over.  That sounds like hell, and not a minor hell, actual hell.  Actual hell is starting your dissertation over.

 

So did I ...

 

 

I'm very, very sorry.

 

 

I have mixed feelings. I really kind of loathed my old topic, and I like my new one. And while I had a lot of resources set aside, and a lot of time invested, those are sunk costs--I'm not going to get any further ahead by being distressed about them, and it turns out writing a dissertation on X-wing is actually pretty cool. I like drawing weird connections, and there aren't many weirder ones than between an early 20th-century experimental psychologist and an early 21st-century tabletop war game.

 

 

I guess that's a difference between your field and the physical sciences.  If I had to redo my dissertation, it would be because there was a fatal flaw found in the experiments I conducted for the past 3-4+ years of my life.  The data I used in my dissertation (which in my branch of the physical sciences we called a thesis, I don't know why it's called one thing or another, but I'm keeping terms the same) was all my own, aside from discussions of previous work in the field.  

Share this post


Link to post
Share on other sites

If you don't mind my asking...

  • what is your field of study?
  • what was your old dissertation?
  • why was your old dissertation canned?
  • have you done a literature survey of the new topic?

 

I actually wondered what you found to be unacceptable based on Mjs work. Or perhaps where your dissertation will increase understanding.

 

He is approaching it in a different direction. Rather than model the underlying analytical fundamentals he's taking an empirical approach by brute-forcing combat scenarios. An analytical model and an empirical model when both done correctly must, by definition, converge to the same conclusions.

 

The public model that I have is fundamentally a blender approach of putting ships in a shooting gallery which is representative of the entire meta. The results can and will be different in particular, specific matchups.

 

 

Ive also wondered if Xwing comes into an inherently flawed system that cannot avoid creating systemic imbalances due to its design.
What happens as player flying skill approaches perfect play?
Does the fact that PS and pilot ability cannot be broken away from each other make it impossible to have useful mid tier stuff?

 

No, the system is actually extremely elegant. The problem is that balancing the game competitively requires a significant amount of math and game theory that FFG does not possess.

Edited by MajorJuggler

Share this post


Link to post
Share on other sites

If you don't mind my asking...

  • what is your field of study?
  • what was your old dissertation?
  • why was your old dissertation canned?
  • have you done a literature survey of the new topic?

My field of study is evaluation. Technically it's the systematic determination of the merit, worth, or significance of an object of inquiry; more colloquially, it's the application of social science methods to solve problems and answer questions posed by stakeholders. (Here is an example of my work; the German in the footer is not our fault.)

Evaluation is distinguished from social science research in a handful of ways, about which any group of three evaluators will have at least four different opinions, but in my view it stems from the aim of evaluation work, which is not simply to accumulate knowledge but to recommend (or not recommend) particular courses of action as appropriate or inappropriate. (Here's a paper I particularly liked and refer back to often as the foundation of literature reviews.)

That can get pretty tangled when it comes to, say, education programs or the effect of affirmative action on diversity in hiring practices. But it's fairly straightforward when it comes to product evaluation: it's fairly easy to identify criteria of merit (the things a product ought to do well) and usually equally easy to identify standards (how well it ought to do those things).

My old dissertation was methodological: "Using Success-Case Method as a Rapid Selection Method for Qualitative and Mixed-Method Evaluation Studies". It just just what it says on the tin, even if you're not familiar with the success-case method (and I can't think of a reason you should bother to become familiar with it, which perhaps says something about my attitude toward the topic). It was canned because I had a personality conflict with one of my prospective committee members, which drove some fundamental disagreements about what I should be covering within that topic and how I should be covering it.

The lit survey for the new topic has been skinny in one direction and fat in another. There's a great deal of information about product evaluation in general, as well as its application in particular fields. There's a fair amount of evaluation of user interfaces in computer game design, as well. But there's no evidence at all that anyone is even thinking of evaluation of games, let alone using repeatable, reasonably objective quantitative methods to improve mechanisms within board games.

 

I actually wondered what you found to be unacceptable based on Mjs work. Or perhaps where your dissertation will increase understanding.

 

He is approaching it in a different direction. Rather than model the underlying analytical fundamentals he's taking an empirical approach by brute-forcing combat scenarios. An analytical model and an empirical model when both done correctly must, by definition, converge to the same conclusions.

It's not exactly that I disagree or find it unacceptable. I'm addressing several different problems:

(1) Simulation and (MJ's) analytical approaches are very different methods. As MJ says, they ought to come to generally the same conclusions, which means one can be a good test for the other. And since it's my dissertation rather than MJ's, I'm using the one I feel more comfortable with.

(2) MJ makes a big assumption about the functional form driving prices in X-wing that I've never been quite satisfied with. I can't think of a better way to test it.

(3) I'll be re-deriving the assumptions he uses about the metagame, in approximately the same way he did but using a larger sample of games.

(4) Having done the shooting-gallery work, I'll be using a very different set of mathematical techniques to build them into estimates for ships' point costs. If I do it right (and MJ also did it right, which again isn't something I'm specifically questioning), we should get roughly the same results.

 

(5) MJ hasn't chosen to make his current work public. Again, that's not a strike against him, but it means there's a gap. And I don't want to simply replicate his analysis.

Edited by Vorpal Sword

Share this post


Link to post
Share on other sites

Ok.  Since you can use Youtube videos.

Hoth Open:
https://www.youtube.com/channel/UCT0_zqao2b2kJBe-bmF_0og

 

Mountain View Regionals (California) was filmed by these guys and should be up there soonish.  It is my understanding they also have recordings of Phoenix Regionals:
https://www.youtube.com/channel/UCprKWlwH0IZ7kKhEaZjTSZg

San Diego Regionals (California) was streamed here:



Yavin:
https://www.twitch.tv/firstearth

Madrid Regionals:
www.twitch.tv/firstearth

I will ask around and post more if I can find it.  Good luck with your dissertation.
 

Share this post


Link to post
Share on other sites

First, let me say that I feel for you having to start over with your dissertation. The same thing happened to be me after three years of research and experiments (someone left my research group and published parts of my ideas). Luckily, I found a new topic and all worked out in the end - it just took considerably more time than was initially planned.

Since you mentioned usability evaluation of games (which is overlapping my field of study), I am happy to point you to the relevant literature if you have any specific questions there. Also, as part of the work in my former team, a colleague built a hybrid computerized tabletop version of the game (using a multitouch table and custom built bases for the ships that can be detected by the table with their precise location and angle). I can get you in touch if you are interested; maybe they have recoded data of played games available - it would be much easier to process than videos, I believe.

Lastly, if languages other than English are no problem, there are a number of German YouTube channels featuring games from different European regional championships (e.g., Michael Geringer, "Flying Pink Unicorns of Death, Doom and Destruction", "Pirates of Tattooine", "Team Hooters"). It seems to me, however, that the meta and the play style differs considerably between X-Wing in continental Europe and the US/UK/Australia. This may a factor you have to take into account in your analysis.

Share this post


Link to post
Share on other sites

Sympathies on starting over. I did that in my Ph.D. and essentially lost two years. But getting away from a project and a committee you don't like to one that you do and now you can start with the benefit of hindsight can be a huge relief. And I'm just about done (formatting my dissertation for submission after edits) - so it's possible to scrub and start from scratch!

 

I once did a project based on warhammer for a stats class but doing it for your whole dissertation is cool! I'm sure this sort of thing is potentially applicable to a lot of things. I'm interested in your approach - I would tend to approach it like MJ based on probability and working up. 

 

While I'm a biologist, I did a chunk of my Ph.D. on probability theory and modeling of biological systems. You and MJ are both welcome to PM me if you'd like some independent eyes on any given problem, I have a soft spot for mathhammer.. 

X-wing regional at Raleigh NC: https://www.twitch.tv/atomicempire

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  

×
×
  • Create New...