Friday, May 13, 2016

Tesla vs. Edison: Developer Diary, Part 2

Developer's Diary, Part 2: 

In part 1, I talked about how I got involved in Tesla vs. Edison: Powering Up!, some of Dirk's original design goals in the game, and how we prioritized the balance between "adding new things" and "providing a familiar experience". 

After making the change to HQ cards, Dirk and I felt like the expansion started to hum. We started focusing on balancing the inventors against each other, creating abilities that were both thematically appropriate and powerful enough to be interesting, and tweaking various things to get the game length and experience where we wanted it to be. 

If all of this sounds like "playtesting" to you, you're right! When Artana hired me one of my directives was to expand our network of playtesters, and TvE: Powering Up! gave me the first opportunity to do so. Fortunately, I'd previously worked with Plaid Hat Games, Level 99 Games, and other companies with top-notch playtesting programs, so I had a pretty good idea for how to grow a system that would allow for lots of folks to playtest, get access to files, and easily report their plays. 

We used Mailchimp (and later, Google Forms) to collect feedback, Google Drive and email to distribute files, and a private reddit subthread as a place for conversation, updates, and deeper reports about the game. Big shout out to Ben Harkins of Floodgate Games for suggesting reddit subthreads - they're completely private, free, and really useful. 

Playtesting is one of the things I like most about design and development. There's something special about having strangers ask to try your stuff, and it's just the bee's knee's when they really dig in and approach the game from new angles and find new problems. To all the folks who playtested Tesla vs. Edison: Powering Up!, thank you - you played a HUGE role in making this expansion as good as it is. We couldn't have done it without you. 

The Powering Up! expansion includes more than just HQ cards, of course - it's got an event deck that introduces random bonuses and penalties, auctions, and stock effects. Happily, these were nearly spot on from the start. Dirk designed the effects to be flavored for their real-life names. For instance, the Panic of 1890 closes the stock market during bookkeeping, which means players have one fewer chance to buy and trade stock during the game. This is a huge effect, as it narrows the gap between first and last, since the player with the most cash can't use it to get even further ahead. 

One thing Dirk wanted to address in the base game was that experienced players regularly "maxed out" their stock price at $77, since it's the upper limit of the board. This is something he felt strongly we should prevent, so we went to work discussing ways to incorporate "downward pressure" on all stock prices. 

In the base game, the only time a stock price would go down is when a company's stock was sold. While that was certainly "helpful" in solving this problem, it clearly wasn't enough, given how many reports we were seeing of maxed stock prices. The original design of the expansion introduced a depreciation step during bookkeeping, which simply pushed prices backward. We pretty quickly moved to incorporating that into the event deck, so we could vary the downward pressure and tie it to the theme of the cards. 

Now, when panic strikes the stock market, it affects your company in two ways - first, that you're limited in some way during play, and second, that your stock price immediately drops. It felt thematic and solved what we were looking for, and it stayed this way through most of playtesting, until we added some upward movement early in the game to make things a little less punishing in phase 1. All-in-all, this is one of my favorite parts of the game. It adds a subtle, dynamic pressure that removes a bit of the AP (analysis paralysis) from the game, as you can't perfectly predict what later rounds will look like. 

Another journey we took over the development process was figuring out how best to capture Madam C. J. Walker's dynamic business and personal success in her inventor ability. We went through a lot of iterations of this, mostly around "get something free when you take a certain kind of action." Early in playtesting she was by far the best inventor, averaging over 60 points per game when the other inventors were around 48-52. It was clear that basically awarding a free action - especially when it was centered around giving her more money - was very powerful, and we had to be careful with it. 

We settled on the ability she has now after a ton of testing and deliberation. It took a while to get it "right" and also "balanced", but I'm really proud of the work. Now you can use Madam Walker's company to do things other than pursue patents, since when you take propaganda actions you can then benefit from the patented work of other inventors. In the meantime, if you've developed your Office, you don't have to worry about paying those patent fees anyway, making you a formidable profiteer to be reckoned with! 

A huge part of this expansion is the AI decks, named after inventors Charles Babbage and Ada Lovelace. Dirk's idea for these decks was to have them take the same kinds of actions that human players take, and to do so in a way that felt like they were human players. So, nothing so complicated that it took minutes per turn of rulebook referencing and "if: then" statements. We wanted players to be able to pull these out of the box and start playing with them right away. 

These went to our playtesters who specifically requested they be included in solo playtesting, and they did a great job in helping us balance the actions and confirm the relative "difficulty levels" of each of the inventors. Not only are the Lovelace sides "hard mode" for each inventor, but the inventors themselves are intended to be ranked in difficulty, so that players could design an easier or harder AI experience based on their preference. I'll let you figure out which ones are harder than others, though :D

I'm really proud of how accessible these decks are. We talked through some more complicated solutions that would have been less "deterministic" and allowed for more nuanced play from the bots, but ultimately what we ended up with was something that won't slow down a game and will introduce dynamic challenges to solo or multiplayer play. I tested these a BUNCH, and I found it really interesting to work around how the AI behaved, even when I "knew" what was going to happen. It's a bit like playing against the bots on the Ticket to Ride app - part of playing against them is learning what they do and responding. Beating one isn't that hard once you know it's pattern, but playing with more than one at the same time is a real challenge! 

Coming up in part 3 - how we finalized the scoring system in Powering Up!, why the stock market has been moved to bookkeeping, late changes to the game based on playtesting with Dirk through Tabletop Simulator, and my favorite parts of the game.