GDC fast approaches, and I’m trying to make sure all the new stuff I’ve added actually works as designed for people coming to play in the IGF Pavilion, and that means playtest → tune → playtest → tune…

Last night the gang came over, ostensibly to playtest the entirety of the aforementioned stuff, but the damned playtesters wouldn’t stop playing The Balcony map.  Over and over again.  If you recall, The Balcony was almost meant as a throwaway silly experiment.  I was slamming the tuning knobs of map size and complexity from one side to the other with The Balcony and The Veranda just to see what would happen.  However, after the last playtest, I made a bunch of tuning changes to the Seduce Target and Contact Double Agent missions that have sort of inadvertently made The Balcony a somewhat balanced experience.

The Balcony, in Practice Spy mode.

The changes I made look like this:

• Seduce Target – You can now flirt with the target just about anywhere people congregate on the map, including at the windows, bookshelves, and statues, not just in conversations like last playtest.  There is now an analog “flirt bar” that fills up as you flirt1, instead of having a fixed number of required flirtations, with the distance the Spy is from the Target multiplied by the Action Test determining how much credit you get for each flirtation.  It’s tuned right now to require two perfect flirtation attempts while standing right next to the Target, three if they’re just okay flirts, and four and six flirtations, respectively, if you’re far away on the other side of the conversation circle.  Since this change is hot off the press, there is currently absolutely no animation tell if you flirt while not in a conversation…you don’t even make a move, the only tell is the behavioral one of being near the same person multiple times.  Eventually2 there will be conversations that can happen everywhere, and so flirting will at least require being in a conversation, even at the windows.
• Contact Double Agent – I made it so the map could specify the number of Suspected Double Agents visible to the Sniper, and this number could go all the way to zero, and then I set it to zero on The Balcony.  Normally, the Sniper can discount all of the Suspected Double Agents since they (currently) can’t be the Spy.  This is  a nice head start on eliminating suspects, which is a key part of Sniper gameplay.  Plus, since the Sniper knows one of the Suspected Double Agents is in fact the real Double Agent, when the Sniper hears the “banana bread” audio tell, he or she can look around at the various conversation groups and pay more attention to the ones with Suspected Double Agents in them.

These two changes took the map from a quick Spy deathtrap into a relatively more balanced but still quick SpyParty-lite experience.  In fact, since the Sniper can concentrate on everyone simultaneously, it becomes a fairly high level behavioral map.  Instead of looking for tells, the Sniper can really watch for the “burgeoning relationship” of Seduce Target, or somebody who appears to be going out of their way to walk by the Ambassador.  It’s also much less stressful for the Sniper than the big maps, where statue memorization and book tracking are important but taxing skills.

One thing I didn’t mention is I’ve codified the current game types into Call Your Shot, Subset, and Opportunistic Subset, where Call Your Shot means the Sniper knows which missions the Spy is going to attempt, Subset means the Sniper knows only the set of missions from which the Spy will attempt a smaller subset, yet the Spy must pick this subset before play, and Opportunistic Subset is like Subset but the Spy can pick the missions on the fly while playing based on opportunity.  Obviously these are in order of increasing difficulty for the Sniper.

On The Balcony, if you know the Spy is doing two Call Your Shot missions, one of which is a subtle mission like Bug Ambassador or Seduce Target, and you hear the “banana bread” tell of Contact Double Agent, you pretty much have to take a shot within 10 seconds or risk losing because chances are the Spy has already done the other mission, so it becomes an interesting almost-arcade version of SpyParty, but that is still about perception and deception.

In fact, after a couple hours of The Balcony, we switched to The Ballroom, and it felt all serious and stressful, way more than it did in the past.  I hope The Balcony hasn’t ruined the playtesters!  On the plus side, it’s great that I can lighten the intensity of the game on some maps while still retaining the core aesthetic of the game.

Ian and Lauren trade the controller under the table, so Rachel never knows who she's playing!

After playing The Ballroom a bit more, we talked about how it seemed like some missions worked better on some maps, and how general the missions would be.  From a production standpoint, I need to be able to share missions across maps, and that’s proving to be true, but there do seem to be affinities between them.  Time will tell how plug-n-play the missions are.  Some customization per map is great, doing all custom missions for every map is a non-starter, so hope for the best.  :)

I leave you with some SpyParty trash talking courtesy of Ian, Master Sniper:

Lauren, trying to learn from Ian while watching him play:  “Why did you highlight that girl?”

Ian:  “Because she was the Spy.”

1. Comments about how it should be a vertical bar that engorges as you flirt will be sent to /dev/null ;) [↩]
2. Where “eventually” is defined as hopefully by GDC. [↩]

Tonight we did the first playtest of the new Seduce Target mission and the new maps I’ve been working on for the GDC build.  I also realized I never wrote up the results of the Action Test playtest session, so I’ll stick those in here as well.

The usual SpyParty playtest suspects, sans Paul, who was crunching on DarkSpore.

The Action Test Mechanic

I playtested the Action Test mechanic back in January, and it worked surprisingly well.  Going into the playtest, there was a lot of concern all around that it would be distracting, make the game more about twitch, and dominate the behavioral gameplay.  Paul even wrote a detailed blog post titled Why I’m Skeptical About Active Reload in SpyParty right before he came over to play.

Action Test

Before playing we discussed all sorts of different aspects of the mechanic, including how often players should get each result (good, okay, and bad), how subtle the tells should be for the different results, and the like.  Being a good game designer (not to mention a highly competitive game player), Paul used the new Practice Mode I put in for the IGF to practice the Action Test over and over again, and he kept track of how he improved.

Once we started playtesting for real, a very interesting thing happened that no one expected, and was, in fact, sort of the opposite of what we expected:  the mere existence of the good result in the Action Test forced the Sniper to play the “behavioral game” more than watching for tells. In other words, because there was the possibility of the Spy performing missions in a way that was not immediately detectable, the Sniper had to lift up to watching for behavioral tells—like partygoers acting with too much intentionality—rather than just waiting to see a hard tell.  This was surprising, and yet totally awesome.  No one expected the addition of this twitch mechanic to make the game even more about behavior!

It also turns out the Action Test isn’t distracting for new players, for the most part.  It goes by in about 1.5 seconds, and if you’re not looking for it, you don’t really see it.  I will probably still unlock it after a tutorial just to keep the number of things new players need to worry about small at the beginning, but the worries about it dominating the Spy’s mental space were unfounded.

Also, the double randomization makes it so the Spy has to pay attention to the bar if he or she wants to opt in, rather than just learning the rhythm like in Gears of War, so that worked well too.

Overall, I was super happy with how things turned out.  The Action Test worked even better than I hoped on a bunch of different levels.  Any time you add something like this to a game that’s already working, you run the risk of messing things up, but in this case it really did make things better.

One thing that became clear was the Contact Double Agent mission is too hard for the Spy against an elite Sniper because it gives the Sniper so much information.  Even though it’s a “soft tell” and only eliminates or implicates suspects, reducing the field is such an advantage that it’s a death trap for the Spy.  The good result on the Action Test I implemented still requires the Spy to be in a conversation, and so it doesn’t actually help much.  I’m going to need to change it so the Spy can somehow contact the agent from outside the circle if he or she performs a good Action Test.

Edit: Paul, who has been slightly busy shipping DarkSpore, finally wrote his followup: Everything Went Better Than Expected.

The New Maps

Tonight we tested out the new Veranda and Balcony maps.  The Veranda has a lot of partygoers, three bookshelves, statues all over the place, and the Sniper can’t see the whole map from any single vantage point.  The Balcony only has a few characters and is completely visible from all angles.

The assumption was the Veranda would be very hard for the Sniper, and the Balcony would be very easy, and this turned out to be the case.

The Balcony level only has the Contact Double Agent, Bug Ambassador, and Seduce Target missions.  The Spy can choose to do only one or two of these missions, and it’s still very hard to win.  The hard tell for bugging the Ambassador, which is so difficult to see in the main Ballroom level that it’s almost a gimme for the Spy, sticks out about as much as carrying a sign that says “Shoot Me”.  The “banana bread” soft audio tell for contacting the Double Agent is basically impossible to get away with because the Sniper can immediately see who’s talking.  Having to flirt with the Target multiple times allows the Sniper to see the relationship building as plain as day.  That said, it was fun to play because it’s such a quick little map, it has very little of the long slow building stress of the regular game.  If you happen to get away with something as the Spy it feels great because it’s so impossible.  It’s almost like a little arcade minigame of SpyParty, so it’s a nice bit of variety.

The Veranda level is the opposite.  It’s so big and there is so much going on that it’s difficult for the Sniper to get any deductive traction going.  Even Ian was overwhelmed and wasn’t using the normal process of elimination and was just watching for a mistake.  We started with 20 partygoers (which doesn’t count the Security Guard and the Waiter) and eventually tuned it down to 17 during the playtest to take off some of the cognitive load.  It is also currently set for four Spy missions, but I’m going to increase it to five.  We think the walkable areas are also a little thin, because this map really highlights the crappy pathfinder I’m currently using.  Overall, the Veranda was fun and added more variety, but it still needs some tuning work before GDC.

The Seduce Target Mission

As I discuss at the bottom of the post about the Seduce Target mission, there are a lot of options on where to take this mission.  But, for this first playtest, I just did the simplest possible thing:  the Spy has to get into a conversation with the Target and flirt four times, each closer in the circle than the last, and one of the two must leave the conversation between flirtations.

Oh Brimsworth, I choose you!

The mission worked out well even in this simple state, and people loved the “fiction” behind it.  For the metrics-addicted, here were the three most popular pairings for the 43 games played, arranged as Spy ♥ Target:

♥

♥

♥

It turns out it’s really hard to get the same person into a conversation four times during the match, so we reduced it to three times about halfway through the night.  I think I’m also going to change it from an integer number of flirtations to a bar that fills up, and instead of having the progression of closeness that starts with just being in the same conversation (which is just a freebie for the Spy), I’m going to make it so the arc distance in the circle impacts how much you fill the bar each flirtation.  The closer you are, the more bar filling you get.  I’m going to tune it so it takes two very-close+good-result flirtations, and six not-close+normal-result flirtations.  So, you can ignore the closeness if you’re willing to basically follow the target around the entire time, or you can be right up next to them and try to seduce them quickly in two or three flirts.

Here are a few more metrics from tonight’s playtest.  I need to generate these straight to an HTML page at some point…

Spy ran out of time 2% (1)
Sniper shot civilian 30% (13)
Missions completed successfully 16% (7)
Sniper shot spy 49% (21)

The Spy and Sniper were mostly evenly matched tonight (Ian’s skill advantage was outweighed by the randomness of testing the new features).

Action test bad 4% (5)
Action test ignored 27% (31)
Action test good 16% (19)
Action test okay 53% (61)

The Action Test is working as designed:  you can ignore it and get an automatic okay, or you can opt-in, and usually get okay, occasionally (16%) get good, and sometimes (4%) screw yourself with a bad.

This last batch is the last Spy action completed before the Spy got shot in the games where the Sniper shot the Spy:

Remove microfilm from book 19% (4)
Bugged ambassador while walking 19% (4)
Double Agent contacted 14% (3)
Fake Banana Bread Uttered 5% (1)
30 seconds added to match 10% (2)
Transferred microfilm 14% (3)

These are a little skewed since the long string of quick Balcony matches we played at the end increased the number of deaths after bugging.  I realized tonight I wasn’t journaling the level played, so I’m glad I caught that before GDC!  This is why I love playtests!

I hope to have another mission implemented, fix all the issues uncovered tonight, and playtest again next week.

Just in time for Valentine’s Day, the newest SpyParty mission is Seduce Target.  I was having trouble figuring out the low level design for the Poison Drink mission, so I decided to put it on the back burner to simmer a bit more1, and do a couple different missions for the GDC IGF booth build.

One of my requirements for the missions I work on at this early stage is that they expand the envelope of mechanics explored by the existing missions.  I’m trying to get a feel for the different design variables in play with the missions, and how they impact the feel of the game and the difficulty for the Spy and Sniper.

There are limits to this methodology, but you can pseudo-analytically break down the design-space of the missions into the different characteristics each one uses.  So, for example, Swap Statue is a mission that deals with a “fixed set of known world locations with a hard tell”, Bug Ambassador deals with a “known character with a hard tell”, Contact Double Agent has an “unknown character from a small set of known characters with a soft tell”, etc.  It would be nice if you could just make a multi-dimensional grid of all these characteristics, look for an empty spot, and say, “ah, I need a mission that has an unknown character and a hard tell”, but sadly it doesn’t usually work out that clearly.  The dimensionality of the space is very high, the characteristics are not really as independent as you’d like, and it’s hard to visualize.  Plus, there’s just an aesthetic component to coming up with a game design that’s hard to systematize.  I often talk to other designers about this, and we all keep trying to decompose the design space like this, but a lot of times you just have to take a stab at it, perhaps driven by the fictional context, or just a gut feeling.

Seduce Target looks like it’s going to be an interesting mission.  Currently, the Spy chooses a Seduction Target that is unknown to the Sniper.  The Spy then needs to flirt with the Target four times on separate occasions to win the mission.  The first flirtation can be anywhere in the same conversation circle as the Target.  For the second flirtation, the Spy and the Target need to be on the same side of the conversation, and for the third and forth flirtations, the Spy needs to be right next to the Target.  The Spy must leave the conversation between each flirtation, so they need to be spaced out over the match.

The mission is hooked up to the Action Test as well.  The description above is what happens if you ignore the Action Test, or do an “okay” result.  If you hit the “good” result, it credits you for two flirtations, so you can skip ahead.  If you hit the “bad” result, you get no flirtation credit, and the Target immediately leaves the conversation.

There are no hard tells for this mission, the flirtation animation is just the normal conversation talk animation.  The only tell is the behavioral one of, “is that character constantly trying to get into a conversation with that other character?”

I’m going to playtest Seduce Target soon, so I’ll find out if it’s too hard for either player2 and it might change from the above description as I balance it.

When implementing all this, I was immediately struck by a few interesting thoughts:

• The question of authorship and an authorial voice is a fascinating one when it comes to interactive art and entertainment, because unlike a song or a book, good games abdicate authorship and put the player up on stage.  However, the game systems with which the player interacts have a different kind of authorial voice, because the rules the designer sets up constrain the player’s actions and give him or her feedback in certain ways.  So, for example, in SimCity, the rules for how taxation affects happiness are part of the authorial voice of that game, and have indirect but powerful influence on the player’s experience.  You could make an argument that these rules are a concrete expression of Will Wright’s philosophy of how cities work (or should or shouldn’t work).  The implementation of the Seduce Target mission started asking me questions about my model of how flirtation and seduction works.  For example, if the Target is talking, and the Spy interrupts him or her, should that offend the Target?  Miss Manners might say that was a no-no, but Neil Strauss might say it’s essential.  The code doesn’t care, but it wants me to decide.  Maybe I make it depend on who you’re flirting with…
• While I’m on the topic of SimCity and simulation rules, Will used to tell stories about how he’d get letters from hydrologists saying “your water model is awful, but your zoning model is great!” and letters from city planners saying “your zoning model is awful, but your water model is great!”  It reminds me of the old saw about how if you read something in the paper on a topic about which you have intimate knowledge, the article is almost always horribly inaccurate, but you don’t necessarily apply that thinking to articles on other topics, which is slightly terrifying.  I guess the analog for Seduce Target would be if some Casanova sends me a letter saying how bad my flirtation simulation is, yet the ballistics on the Sniper bullet are perfect!
• I love that the Spy player picks both the Spy character and the Target character, and the Sniper doesn’t know who the Target is, because it injects yomi into the mission, while also subtly commenting on diversity issues in games completely through player choice.  Because the Spy can be male, female, black, white, Asian, Hispanic, young, old, and any other categories I create characters in, and so can the Target, I can’t wait to see the metrics on the kinds of pairs people pick!  If you suspect your opponent has a specific opinion about homosexuality, do you have a male Spy seduce a male Target since they’d be less likely to suspect that, or would they be more likely?  Are Snipers less likely to suspect the young strapping Spy would try to seduce the old dowager, or vice versa?
• How far do I take this mission in terms of implementation?  This first version is pretty lightweight, where if you perform the flirts then it checks off the mission without any real behavioral changes on the part of the Target.  If I want to make it richer, do you need to talk the Target into meeting you out on the balcony, or (ESRB!) the bedroom?  Does the Target start approaching or avoiding you depending on how you’re doing?
• Even in tests, it’s kind of sad when the Sniper shoots the Target, sadder than if it’s just another random partygoer.  I’ll have to see this holds true in playtests, and if I can push on that a bit.
• The current design has the 4 required flirtation attempts, but I’m also thinking about playtesting a version where there’s a bar you have to fill up, and each flirtation attempt fills it by an amount proportional to how close you are in the conversation, times the Action Test result.  Although, the idea of having a bar that fills up as you flirt might be a bit too much…at least with the current UI the bar would be horizontal, not vertical!

1. I find plain old wall-clock time is a strict requirement for good game design; you can’t rush things, and it’s clear Poison Drink needs to stew a bit more in my subconscious. [↩]
2. Ian thinks it will be too hard for the Sniper, but he could use a little challenge! [↩]

The fellow on the right looks like a pretty nice guy, no?  Well, take my advice: don’t go to a party where he is lurking outside.

Meet Paul and Ian, my elite SpyParty playtesters. Both have played around 10 hours of SpyParty across multiple sessions.  They’ve played the game more than anybody else, including me if you don’t count my daily code testing1.  Paul, on the left, is a designer on DarkSpore, and he specializes in playing the Spy.  Ian, a designer on The Sims, is partial to playing Sniper.

When you’re trying to make a really deep player-skill game, you need to have people like Paul and Ian who will play your game a lot and get very good at it.  I’m really lucky to have these two friends who are happy to play my game so often and for so long2.  Given my depth-first, accessibility-later development philosophy, I am basically designing the game for these two guys.  If anybody else has a good time playing, that’s a great bonus, but right now, Paul and Ian are the most important SpyParty players in the world.  If I come to a decision where I have to choose between making the game better for newbies or better for Paul and Ian, I choose the latter.  It’s an audience of two.

For most of the development, everything was rosy, and Paul and Ian were evenly matched.  This was a huge relief, because SpyParty is a highly asymmetric game, so balancing the Spy versus the Sniper is a subtle and difficult challenge.  Every time I’d add a significant feature, I’d test with Paul and Ian, take a bunch of notes, go back and tune the feature if it was unfair (sometimes while they waited), and then we’d test again until it was balanced.  The fact that they’d played the same amount, and they’re both ultra hard-core gamers and very competitive, meant I could trust that if they were each winning about 50% of the time, I had a balanced mechanic.

The Good Old Days

Naturally, when I got the Action Test feature working, I called up Paul and Ian and the rest of our playtest gang and had them come over to try it out.  We ordered pizza, and everybody had a good time casually testing the new feature and discussing its pros and cons.  Finally, it was time for Paul and Ian to sit down and see what impact the Action Test had on the elite level game.

They hadn’t played each other for a few months because I’d been doing so much traveling, but last time we played they were at the same level, so I figured after a few warm up games they’d be back to form and the intensive balance testing could begin.

Except Ian kept shooting Paul almost instantaneously.  Over and over again.

What the heck was going on? Had I screwed something up with a recent feature? Did the Action Test actually do the opposite of what it was supposed to do? Had Ian planted a mirror behind Paul?

Spy Bloodbath

I decided to sit down and play the Spy.  I’m not really a fair person to play against, since I wrote the code, and I know exactly what percentage of the time people take books from the shelves, and how fast they sip their drinks.  These are huge advantages when I’m playing Sniper, but even for Spy, I can usually beat anybody.  However, even though I did a bit better than Paul, Ian managed to kill even me over and over again.

Okay, something was wrong.

Most people who hear about the game think the Sniper must have an advantage, but this isn’t actually true.  The metrics are pretty clear on this:  even for newbies, the Spy and Sniper are pretty evenly matched.  For example, in the recent playtests at NYU, the Snipers won 48% of the time to the Spy’s 52% over 188 games, and at an even more recent playtest at Valve, the Spy won 55% of the 104 games played.

However, at PAX, an observant fan in line noticed a trend…the Sniper was winning more and more as the show went on.  Day 1, the results were 47%/53% Spy/Sniper.  Day 2, the score was 35%/65%.  Finally, on the afternoon of day 3,  the issue was noticed and debugged, just in time for the tournament, but overall the day was 36%/64%.  What was going on was quite interesting and subtle: because there was a huge line with a long wait, everybody had plenty of time to read and digest the Four Page Instruction Manual and then watch a whole bunch of games before they got to play.  This, it turns out, is Sniper practice.  The Spy is a kinesthetic experience; you need to be holding the controller to get better at playing Spy.  But the Sniper is mostly about perception, so simply watching 10 or 20 games while you waited in line, seeing the various tells from different angles, and watching what worked and what didn’t for each side, gave the Snipers a huge advantage.  We fixed the problem by not giving verbal help to newbies playing Sniper, figuring they’d had enough practice while waiting in line, and only helping out the Spies.  This was enough to restore balance for the people playing the game for the first time.

However, did you happen to notice anything about all the pictures from the PAX days linked above?  Here, let me help you out with a couple examples:

Our pal Ian was one of two friends who helped out in the SpyParty booth the entire three days of PAX.  Without Ian and Tom doing the bulk of the work, I’m sure I would have died right there on the show floor.

Well, this means Ian got something like 25 hours of additional Sniper practice over the course of PAX by tutoring people and observing games the whole time.  Look at him in those pictures, sucking up information about how the NPCs move and how often they do and don’t do things.

This is the current best theory explaining why he’s winning all the time now. He’s an unstoppable machine because he’s played three times as many hours as the next best player.

This is good news and bad news.  The good news is if this theory is true, the game is working exactly as I’ve designed it.  He should be beating Paul (and even me) every game now, because he’s played so much more, even if his hand didn’t touch the controller more than a few times at PAX.  I’m designing the game to have an intense player-skill ramp, and this is what should happen if I’m successfully achieving that design.   The bad news is it’s hard to know if the theory is true without me or Paul practicing enough to get to the same level, which is a lot of practice!

So, I’m going to need to practice my own game, which seems a little cosmically unfair, but such is life when you’re striving to make a truly deep game, I suppose.

I’m also using this as an opportunity to fix as many of the subtle places where the NPCs and the Spy diverge due to lazy coding.  Ian noticed some of these during PAX, and I need to deprive him of that advantage.

An example of this kind of divergence is the Ambassador would only have his or her personal space violated by the Spy character. Ian observed that the Ambassador would run away sometimes, and deduced this was a tell.  Now the Ambassador will move away from any character that stands too near for too long.  This is subtle stuff!

I’ve also changed the way the conversation groups work.  Now NPCs look for open areas, and randomize within the entire open arc, rather than just aiming for a slightly random point near the center of the arc.  Also, there are no longer any hard coded maximum number of characters in a given conversation group, the NPCs look for an opening big enough, and if there is one, they can go for it.   Ian was asking me the other day about how many people would be in a conversation at once, so now there is no straight answer to the question.  Take that, Sniper!  There is a probability table that tells NPCs how likely they are to want to join a conversation given the number of people already there, but I’m not going to publish that,  sorry Ian.

One thing I’m not going to fix is how NPCs won’t go out of their way to be near the Ambassador, while the Spy needs to sometimes.  This is happening at the behavioral level where I want the game to live, so the onus is on the elite Spy to bide his or her time waiting for the Ambassador to be in a plausible position to walk near.

One other good thing to come out of this is an interesting modification of the “3 out of 4 missions” game mode.  When newbies start playing, they play with the Spy required to do 4 out of 4 missions; it’s too overwhelming for newbie Snipers if the Spy doesn’t have to do everything.  Once the Sniper starts winning a lot, usually around 10 or 20 games in, we switch to the Spy picking 3 out of 4 missions.  With this “subsetting”, the Sniper can’t camp any individual mission, and there is interesting yomi in which missions the Spy picked.  However, given Ian’s skill gap, I realized a good new mode would be “any 3 out of 4 missions”, where the Spy doesn’t have to pick the 3 up front, all 4 are available, and the Spy can be opportunistic during play.  This should be slightly harder for the Sniper because now the Spy has more options, yet still only has to do a subset of the missions to win.  I’ll need to playtest this to be sure, but I think it’ll work well at elite levels.

In the meantime, watch out for this guy:

1. …which you probably shouldn’t since it’s not very good practice [↩]
2. My pet theory about why Ian likes the game so much: during his day job, when a Sim misbehaves, it’s a bug and he has to go pester a programmer to fix it, and it’s a big pain;  while testing SpyParty, he gets to shoot the misbehaving Sim in the face and win for doing so. [↩]

One of the new things I’m doing for the GDC build is adding two additional maps.  Up until now, everybody who’s played SpyParty has played in the Ballroom, a box with 12 or so characters in it and windows on two sides.  The Ballroom is carefully tuned to be fair for both Sniper and Spy at all skill levels (well, at least from total newb to 10 hour elite hardcore, the current limit of my playtest depth).  All the screenshots you’ve seen of the game are of this map, although for pictures I often increase the number of characters walking around.  Here’s a typical view from the Sniper’s perspective of the Ballroom with the correct number of characters at the party:

The first new map I’m adding is called Balcony.  I haven’t playtested it yet, but I’m pretty sure it’s going to be a deathtrap for the Spy.  There are only 7 or 8 characters, and the space is tiny.  The Spy has to either bug the ambassador, or contact the double agent.  Neither is going to be easy, since a reasonable Sniper can trivially watch everybody simultaneously.  I dunno, maybe the Spy can hide behind that plant:

On the flip side, the Veranda level should be a Spy’s dream mission.  There are 22 characters walking around, and the Sniper can’t see the whole level from any single position.  The statues are spaced very far apart, as are the bookshelves, and there are conversations going on everywhere:

Here’s a bird’s eye view from a debug camera:

I can’t wait to playtest these.  When you’re tuning a game, it’s sometimes useful to slam the knobs to either extreme, so I’m interested to see how these maps feel to play.  Is the Balcony hopeless, or does it feel like a fun little diversion?  Does the Veranda feel like a hard Sniper challenge, or simply impossible?  What happens if you have two Snipers watching the Veranda simultaneously?

I’ll report back when I’ve got some data!