So, we made a trailer last week, which may have inspired beta player jecat to make this awesome SpyParty parody of the famous Dead Island Trailer:

Ye Olde

That long image there to the left is a picture of the current spyparty.com homepage as I write this. As you read this, there will be a new fancy doesn’t-look-like-a-default-WordPress-install homepage. Welcome to the 21st century, spyparty.com!

Plus, to go along with the new homepage design, we made the first SpyParty trailer! More on that below.

The development blog is staying the same old design for a while,1 but I wanted people who came to the site looking for information about or to play SpyParty to be able to find it easily, which was not the case with the old homepage.

So, I went with a design like most indie game websites, with the name, a trailer, a short description, some screenshots, and a “buy button”, all near the top of the page. Of course, since I am…well, let’s just say I’m never at a loss for words…you can keep scrolling and get more information about the game.  I wanted to highlight all the various ways to follow the game and get more info, talk about how cool the community is, call out some of the great quotes from the press, acknowledge that I have no idea when the game will be done but that I will keep making it cooler, and finally talk about John and Keith a bit. Oh, and proudly display the game’s laurels, as required by Indie Game Law.

I wanted to show some of the new art on the page, but also mix in the old art since it’s the art you actually play right now.

Also, there’s finally a link to the beta site login for existing players to find the forums or re-download the game, which has been requested a million times here and elsewhere, except I probably hid the link too well so no one will actually see it. I also think I might make the blog link a little more obvious. I’ve got a way to put a headline near the top of the page for important news, but I also might make a section for the latest headlines from the blog, and maybe even a list of live streams currently playing, kind of like the streams page.

One important goal of the website was to make it fully responsive,2 since apparently these mobile telephones and tablets are a thing now.  I got a bunch of help from the totally awesome Chris Remo with how to implement my design in HTML and CSS and make it use CSS media queries to work on small-ish devices.  You can resize your desktop browser window to see all the dynamic stuff happening as the width shrinks, I’m pretty happy with how it came out.

One thing I’m not happy with is how it appears to be basically impossible to write a javascript lightbox that doesn’t suck.  If you click a screenshot on the homepage (or most places on the blog), the lightbox will pop up.  I’ve used a heavily modified version of Lokesh Dhakar‘s lightbox script for a while, and given my effort to make the page work well on phones, I tried to update the lightbox to do touch interactions seamlessly. Well, as far as I can tell, it is basically impossible to implement a lightbox that does not allow the user to scroll the lightbox off the screen, allows native pinch zoom in the browser (not emulated crappily by javascript), and allows swipe to change images when at the edge or one finger scroll if zoomed in, and is cross browser and degrades gracefully.  Basically, I want a lightbox that acts like a native photo gallery app, but it does not appear to be possible.  I decided native pinch zoom was the most important feature because I’m always frustrated when a mobile site disables it when there are images on the page, so I prioritized that and ended up allowing the lightbox to scroll off the view, which is bad, but was the least worst of all the tradeoffs I found.

The Trailer

As part of this homepage redesign, we finally made a trailer for the game!  Keith did a great job with it. We’ve had trailers for specific features before, like the New Art Trailer, and the Replays Trailer, but never for the game as a whole.  Here it is:

I wanted this trailer to give the flavor of the game, but not explain too much. It should have a bit of a narrative arc and tell a light story, but nothing too distracting. In the end, I think it might be a bit confusing for complete strangers to SpyParty, but I hope people can figure it out from the video and the short description, and hopefully the trailer makes people who are just hearing about the game want to learn more.  I also worry that the trailer implies there’s a fixed character who’s the Spy, but it wasn’t clear how to correct that without a lot of exposition that would have messed up the flow.

You’ll notice this trailer is almost completely the old art and is completely in-game footage. This is intentional, because the last thing I want is somebody watching some super polished pre-rendered trailer and then loading the game and feeling misled. Of course, this means the Spy pours a drink down her face in one shot because the runtime code for taking sips from a drink isn’t aligned with the mouth properly in the old art. Oops.

The one place where we diverged from the game was having music in the trailer. The music was a relatively late addition; we were originally going only have the crowd walla, but in the end it felt necessary and I think it really helps tie the whole thing together.

We’ll be making new trailers every so often now that there’s a good place for them right there on the shiny new homepage, so as the new art comes in expect the new trailers to reflect that. I’m also thinking about inviting fans to make trailers to see what people come up with; I really liked it when Minecraft showed fan-made trailers on their homepage.

Also, here’s to the little details:

Bugs and Comments

Please post comments below if you have trouble with the new page on your preferred internet viewing device. It’s impossible to test all the different ways to browse the web, so I’m hoping my limited stochastic sampling found most of the big stuff, but I’m sure I missed some things. If you want to report a bug on the page, mention your browser and device and what you saw when things went wrong.

Thanks!

1. mostly due to priorities and time constraints, not any particular love of the old design [↩]
2. which, the last time I made a web page, was called liquid, and also get off my lawn [↩]

I was playing a birthday match1 today with jecat, and after one of the games where I was the Spy and thought I was highlit the entire game but in fact was lowlit, we got to talking about often this happens, even in elite games. In fact, the Spy often thinks the Sniper knows (or strongly suspects) who he or she is, and so Spies often find themselves playing conservatively even though they know from experience that when they’re Sniper, it’s very hard for them to figure out who’s the Spy; you usually feel pretty snowed under with information.

I got to thinking about it, and this Spy world-view, where you always think you’re the center of attention, even though there’s an entire party going on around you filled with people doing their party things, seems really narcissistic, or egocentric, or solipsistic…I’m not sure which, maybe all three? Fighting that assumption that you’ve already been made and forging ahead is one of the core biases to overcome as the Spy. You see it at all levels of play from newbies to elites, which I find very interesting.

This is all related to one of my core aesthetic goals of the game, namely forcing consequential decisions with partial information, which I’ve always thought of as primarily occurring on the Sniper side, but it really does happen on both sides. As the Spy, you have to decide to do something (accomplish missions), and then what to do (which mission), even though you don’t really know if you’re a suspect or not. How you go about this “doing” is obviously very important, but just the act of overcoming your confirmation bias and starting at all is something you can feel when playing.2 This is one of the parts of the game that I’m most happy with, because I think this concept of embracing uncertainty and making decisions even when you’re not sure is very au courant, in that most big decisions in modern life are made with only partial information, but you still have to make them to the best of your ability.

Here’s an edited clip of the stream with parts of this game and the discussion after it:

If you want to see the whole stream, it’s here.

In the stream chat, varanas had a nice quote on this topic:

Sometimes when I think I’m done for I’ll just start taking crazy risks because I think I have nothing to lose. I think it’s best to assume that if you’re not dead then there’s still everything to play for.

Given how often the person who thinks they’re doomed ends up winning, I have to agree.

1. usually round numbered games like 1000, 5000, but in this case, 12 games around his 2222th game, since we missed his 2000 [↩]
2. This idea of games making these thought processes explicit is covered nicely in Frank Lantz’s GDC talk, Hearts and Minds. [↩]

Two years ago checker invited internet strangers into his game and I was lucky enough to snatch an invite from the very first group. I had been following the game closely before I got the invite, and that is obviously still the case today. Over the years I have observed a lot of changes in the game itself and in the “meta” surrounding the game.

Last year I made a video that highlighted my first year playing SpyParty.  This year I asked myself a simple question:  how have the meta and balance changes evolved over my career as a player? Here is my personal recollection of how everything went down. First and foremost let me set the stage for SpyParty v. 1919 (April 17th, 2012)…

SpyParty v1983

Maps:
Balcony
Ballroom
Veranda

Missions:
Bug the Ambassador
Contact Double Agent
Transfer Microflim
Swap Statue
Inspect Statues
Seduce Target

Balanced Game Type:
Pick 3 of 4

Game Type Evolution

Way before the beginning of the closed beta there were two elite players named Ian and Paul. They felt that Pick 3 of 4 was the balanced game type for its time. Pick 3 of 4 means the Sniper sees 4 missions but the Spy only needs to complete 3 to win.  The Spy picked the 4 visible missions and the 3 to accomplish before the game started. By contrast, Any 3 of 4 means the Sniper still sees all 4 missions, but the Spy can do any of the 4 as long as they get 3 done. 

The Pick 3/4 preference was passed down to the newly recruited beta testers as the standard game type. Over time, the new closed beta players began figuring out how the game works, and they soon surpassed Ian and Paul in game time played. With more people came more feedback, and one day it was decided to try out Pick 4/5 and see how that felt. This was an experiment to see how people liked or didn’t like it. Over the course of 24 hours, Pick 4/5 became the new standard of balanced game play. Needless to say people loved it! Pick 3/4 never saw the light of day again.

There were a lot of discussions about balance between the Spy and Sniper. Some statistics checker has been collecting showed Snipers were a bit favored and we were encouraged to try the game type Any 4/5. This was a slight Spy buff because with Any, the Spy could be more opportunistic and take what the party gave them. This new game type was overwhelmingly popular and once again the balanced game type changed. From that day forward, Pick was less popular than Any.

By this time there were tons of strategies, tutorials, and guides posted for both the Sniper and the Spy. People were just starting to break into the quadruple digits of games played, and the Spy win percentage vs Sniper win percentage was still pretty close, but it could be even closer. A Sniper strategy that started to show results and popularity was the idea of camping the hard tell missions. At the time there were still 6 missions, which meant on any given balanced map and game type you were still required to do at least 1 hard tell mission. This allowed the Sniper to watch two missions and catch the Spy with reasonable success. The Spies received a buff once again and tried Any 4/6, requiring Snipers to watch an additional mission. This was met with welcoming arms and Any 4/5 died in favor of Any 4/6. 

People seem to enjoy playing with a slight bias toward Snipers winning. It’s no fun if it’s too hard as a Sniper, while winning against stacked odds as a Spy makes you feel good. However, can’t the reverse be said? Why don’t people feel just as good for winning as a Sniper when the odds are stacked against them? Why can’t it be no fun if it’s too hard for the Spy? 

I can not  shed any light on that but what I do know is people continue to play slightly favored Sniper games as the balanced game type. There’s obviously something going on here if we’ve been doing it for awhile now. We started off with a slight Spy bias and now we are at a slight Sniper bias with Any 4/6. However that does not mean the win rate for either one strays far from 50/50; it’s actually really close.

The Story of the Courtyards, Panopticon, and Map Tweaks

Courtyard 1

Courtyard 1+2
The first new map to be introduced to the SpyParty beta community was Courtyard. Everyone rejoiced with little butterflies in their stomachs as they got a new medium sized map to play on. It featured a high vantage point for the Sniper and a whopping 16 guests. There were 3 statues in the center area and no bookshelves. It was the first map that allowed a complete 360 degree view of the entire map.

This map posed a particular problem for the Spy–at the time Any 4/5 was considered balanced. There was a max of 5 missions on Courtyard and 2 of those involved statues, therefore, the Spy must visit at least one statue to complete all of their missions.

If you were to highlight anyone that went to statues you’d have a great chance of highlighting the Spy. Not only did you know the Spy has to visit statues, but the AI didn’t visit the statues very much at all to begin with. If the Spy chose to inspect, they had to visit at least two statues and two visits is a death sentence to any Spy. 

Courtyard 2

Thus, Courtyard 2 came into existence. Courtyard 2 provided more cover for the statues and the Spy no longer had to visit them more than once to get inspects done. There was an additional statue added to make 2 pairs of statues. This helped alleviate the problems Spy had with statues and there was a field of view change for the Snipers. 

In an effort to try and keep Courtyard 1 semi-balanced, as an experiment checker tweaked the settings so the AI wanted to visit the statues on that map.

Spies were also having a bit of a hard time completing missions on these two maps. The start time for them was originally 2:30 but seeing as the spies needed a slight buff, 30 seconds was added to both maps and they now reside at 3:00.

Panopticon
This leads us to our lovely and ginormous map called Panopticon. It featured a whopping 3 bookshelves, 6 statues, 20 guests, and enough window space to time add for days. That’s not the only catch here–the Sniper was on a fixed point and could not move around. The only thing they could do was look around and it was impossible for the Sniper to see the entire party at any given time.

Even though new maps are fun and experimental, that doesn’t mean the older maps didn’t get a little loving either.

Veranda
Veranda used to feature 6 single statues. Getting inspect done on this map was a nightmare. To help with this, it was decided to make the front 4 statues be paired together, 2 on each corner. This has since changed to 4 pairs to give a small buff to the Spy. The front section of the map, where Damon resides, used to be more narrow. It was widened in an effort to reduce so many collisions when trying to cross from one end to the other.

Veranda Now…

…and Previously

Ballroom
This map has only received minor changes, for example the small crossbars on the Ballroom windows used to completely block the laser, but now only the big pillars will block it.

Inside Gallery

Gallery
The newest map came in the form of an art gallery. This long and narrow map aims to provide tons of visual occlusion for the Sniper. The moment this map hit, the testers loved it, even more so than when Panopticon came out. That’s not to say Panopticon is a bad map but there’s something about gallery that settles well with the current testers. Perhaps it’s how colorful the map is compared to the others or the occlusion itself providing a unique challenge for the Snipers.

Missions Are Ever Changing

Missions have also been touched time and time again. 

Inspect Statues
To inspect 3 statues on ballroom, you had to visit 3 different statues. This is why it was popular to figure out how many times people have gone to statues. This strategy was a very easy way for the sniper to be purely deductive and still maintain a high win percent. 

After thinking on the subject, the idea was to add action tests to speed up the inspects to allow Spies to inspect statues that were near by. If you wanted to get all 3 inspects done, all you had to do was go to the middle statue and both neighboring statues would be available, assuming they weren’t being held by other guests. Each consecutive inspect takes less time as well.  Of course, this made visits to the middle statues slightly more suspicious than to the edge statues!

Swap Statue
Originally when you got a green statue swap, it would instantly change the statue shape in your hands (actually, see the comments below – checker). This was different from the white test because white tests would fade into the next statue instead of instantly changing. Against any experience Sniper this small difference is irrelevant. The solution was to have the green swap be handled by an NPC. This meant the Sniper had to know who was at statues and pay close attention to how the statue swaps (it’s slightly faster if the NPC does the swap) to know if the Spy is the character swapping the statue, or if it was the previous character.

Purloin the Guest List
Perhaps this was the inspiration for the purloin. When purloining the guest list was first introduced, a green test would give you a drink and the next NPC to take a drink would make the list fade quickly. A white test would give you a drink and the fading was longer. To make the mission feel more different from Swap, the green purloin was changed to refuse a drink, which meant the chain of drink refusers could grow quite long, making it more difficult for the Sniper to track.

Miscellaneous

Virginia “Orange Dress” Vulpes

Orange Dress
You may have heard some rumblings and jokes on streams about a character named Virginia Vulpes or more commonly “Orange Dress”. Who is Orange Dress and what’s with the attention? Let’s start from the beginning.  

Way back in the beginning, there was a beta player named skrymir. He was on a mission to make “Purple Dress”, Anna Klàvsky, the most famous of all SpyParty characters. This is the earliest example of deliberate mind gaming that I can remember. Often times skrymir would pick Purple Dress or bring her to the party with him. Every map had Purple Dress and that alone could drive some Snipers mad. 

Anna “Purple Dress” Klàvsky

After a while playing with the Purple Dress meme, an alternative game type was invented by players called MindParty. For this game mode you play on Balcony, the smallest map. The Spy has to manually pick their character and cast members. They were not allowed to take control from the AI at the beginning of the game, and they must say one sentence. It was typically about a character and it could be true or false. The Sniper had 30 seconds to shoot someone. 

This is where we arrive to Orange Dress. There are two beta testers by the names of canadianbacon and scallions. It is unclear of who originally started it, but whenever they played MindParty they would always play mind games around Orange Dress. This gave the game mode the alternative name of Orange Dress Party. The Orange Dress meme has spilled over into other testers’s games and has since become known throughout the community.

Time Adds
Enough about specific characters though, let’s talk about adding time! Adding time at the windows used to only net 30 seconds of additional time for the Spy. Players felt that this wasn’t worth the risk if you take in consideration the time it takes to walk to the window, do your time add, and walk away. Now time adds grant 45 seconds worth of time to any Spy willing to risk it.

Four Levels of Suspicion

Highlighting And Lowlighting
Next up is the ever so important Sniper tool–Highlighting and Lowlighting. Ever feel like if you had just one more level of highlights or lowlights you would win more games? Two level highlights are a thing of the past. You used to be able to mark people suspicious, and then mark them once more for an even great suspicion rating. Likewise, you used to be able to mark people as less suspicious and least suspicious.  This was changed because it felt too deductive. You could reliably keep track of who had visited statues and how many visits they’d made by using the levels as a counter. At this time, statues were still a problem area for spies. Imagine at the beginning of the game you lowlight everyone. This gives you the ability to mark up to 3 statue visits per person and you still had a lowlight to reduce the people you don’t suspect.  Losing a level of highlights and lowlights didn’t come without some resistance from the community though. However, after playing with it for awhile, the community seems to regard it as a good move by checker. 

In Conclusion

What does the future bring us? That’s a good question my friend. Maybe checker will cease production on SpyParty and become a florist instead. Maybe he continues to work hard on SpyParty and it single handily creates world peace, solves world hunger, and deflects oncoming asteroids so Bruce Willis doesn’t have to sacrifice his life for the greater good of humanity while making checker gazillions in the meantime. 

Realistically it’s probably somewhere in between those two scenarios but whatever the case, here’s to one more year.

The new map, Gallery, is out in the latest build! The design goal for this map was to make it long and narrow with a lot of dynamic occlusion caused by the partygoers.

Obey local ordinances!

It seemed to work pretty well, we played it during the release notes stream, where I did two different “birthday games”, the first set with scientist for his 5k birthday, and then smash10101 for his 1k birthday.

I made a few subtle behavior changes for this new map:

• I made paintings more attractive on this map, and to do that I had to slightly change their probabilities on Ballroom. I think it won’t be noticeable, but we’ll see.
• Due to restrictive local ordinances, open containers of alcohol cannot be consumed or served on the sidewalk. Some people flout these ordinances, but most partygoers are responsible citizens.
• The conversations are a little less crowded now. People will still stand close to each other, but they won’t all jam into a tiny conversation circle, so you have to be a little more aware of how the NPCs behave around crowded conversations. I had to be careful with this code so I didn’t break the delicate balance of Balcony, but I think I was successful. I’m sure the beta players will tell me if I screwed it up!

Also in this build, I worked around a bug in Wineskin so MacOS beta testers should have their replays directories updated properly now.  

If you’re a beta tester already, come try out the new map, it’ll be there after you auto-update. If you’re not, what are you waiting for?

Here’s the video, and if you want notification of videos like this, you can subscribe to the SpyParty YouTube channel: