Designer Diary: The Stealth and Strength of Ninjato

There's an old saying that success is 10 percent inspiration and 90 percent perspiration. For the game Ninjato this was absolutely true. Most of the 10 percent came quickly and easily during a phone call in June 2008. I was talking with my great friend Dan Schnake, co-designer of Ninjato. "Let's make a game with ninjas. Let's make it easy, too – and fast paced. It should finish up in 45 minutes. It'll be great! We'll have ninjas sneaking into houses!"

I'm sure many aspiring game designers have thoughts just like this. The problem with inspiration is the game itself always sounds fantastic – like it's almost in reach as soon as you finish the thought. Galactic Emperor was just like that. Ninjato on the other hand took nearly two years. The reality is game design is exceedingly hard to do well. I'm still searching for solid design rules to guide me and shorten the path (and time) to a great game design. So far, the only solid rule is to playtest repeatedly. During testing, it generally feels like wandering, so the process is uncomfortable. You do have to lay out objectives and goals for the game.

It's strange to look back through all our playtest and design notes. (I keep copious notes during design.) The game that is being published by Zev and Z-Man games is nothing at all like the versions before it. At one point, we had a version of the game with a map of a single temple. Another version had worker placement spots where you collected weapons and smoke bombs. We had a diluted share track akin to Chicago Express. We had a future of Japan track that was like a simplified Die Macher (as if that were possible). None of that found its way into the final design. In fact, the only thing that remained reasonably consistent was the concept of a game that played in about an hour and involved ninjas. Here are notes from over a hundred playtests:

Quote:

Playtest #24: There are huge swings in the auction if one person begins to bid it up. For example, if a player bids five actions, it means giving that to another player – who is then more at liberty to bid some large amount to get what they want. This isn't necessarily bad, but it does make for some swingy decision – especially so if you have a newbie who bids higher than he should and gives actions to an experienced player.

Auction mechanisms are incredibly hard to build. Later we ditched the auction mechanism completely, although we did get to something pretty fun. In this design, your currency for the auction was your actions. Pretty interesting really. Maybe we'll use it in another game.

Quote:

Playtest #36: There could easily be some complaints about luck, but the fact is: you can use Temple to scope out a house. However, you can't really act on it except in combination with House and Dojo. So it seems kinda worthless really. I mean – what do you do if you can see the cards? Decide what you need to take from the dojo. Else – there's *nothing* you can do – except avoid a stuffing by push your luck. I dunno. It just doesn't work really well.

We worked continually to reduce the amount of luck in the game. I often think in terms of what players would complain about or would suggest is the only reason for victory, and if they suggest it was all luck, the game isn't very strategic. Like all design elements, luck should be thoughtfully placed.

Early version of the game board

Quote:

Playtest #43: One tester said at the last round: "I can't win!" This lead to a big discussion about possible plays by all the players. In the end, it turned out he did have a few choices, but none of them would get him points necessarily (or at all). This resulted in a king-making feeling. Although not guaranteed because other players would have to cooperate, it sure felt that way.

An important aspect of playtesting is being able to sit by quietly and see what the players do. In this test (which was really enjoyed by the players), at least one person felt there was no way to win. That sets up things like kingmaking from boredom or spite. This version of the game didn't last much longer either – even though I was very encouraged by this one playtest. I find you have to temper even great playtests by taking careful note of every bit of problem experienced by the players.

Quote:

Playtest #52: The opening felt really bad and was generally a tone for the game. When a player has guard cards, they are going to go spend them on clan tokens. The only decision is which one – and there is a bit to think through – this is where the game feels a bit mathy. When you were low on cards, you pretty much went and got them – although there was a bit of decision making if you had (say) three pretty good cards – do you go ahead and take a house to get guard cards? That was kinda cool.

The essence of the game was being drawn out at this point. It's another important aspect of watching a game play: what worked well, what was enjoyed, what made the players laugh and joke and have fun? That's the stuff that needs to become the focus of the game. In Ninjato, we eventually knew that attacking the houses was the center of the game. Finding this type of focus early in design moves the rest of the design along faster.

Quote:

Playtest #107: Favors although intentionally gathered and plotted out, scored too high in the end. In percentage, it was something like 52/124 - 40% of one player's score. That player had three swords 5 x 5 (so 25 pts), two elites 3 x 5 (15 pts), one skill (3 pts) and two honor favors: 9 pts. Total is 52. It should be no more than 30%. So with about 70 prior to favor scores, they should have gotten about 30-35 pts. Which means I would need to reduce the total by about 20 pts.

Much of game design revolves around math and statistics. Dan and I routinely fall back on computer simulations to figure out how something will work over thousands of plays. Here we were trying to get the end game "surprise" scoring to balance out. I dislike a game where a final surprise score blows away everything else in the game. In my opinion, no single scoring lane should outweigh any other scoring lane. In Ninjato, we adjusted the number of cards and frequency of rumors until the very last moment in design. I think this part of Ninjato is thematically sneaky and fun – especially the "Dishonorable Opponents" set that scores best if only you have managed to collect them. That was Dan's design idea. Brilliant.

As you can see, the 90 percent perspiration was a willingness to play the game over and over, then throw away whatever didn't work. After about thirty versions of the rules, I came to a point of nuking the entire thing. I literally threw out everything that was even slightly out of place, leaving only the core items that made the game work and made it fun.

It was then that the idea of playing cards to defeat guards came into focus – and another touch of inspiration which was the result of solving a simple problem when using cards: How do you make all the cards valuable to the players? This was especially troubling when we had numeric values on the cards. If we made high value cards most important, the low cards just fill up your hand. If low cards are what win, what about the high value cards? How do you make all the cards worthwhile? That was when I was inspired to let the player choose how they would invade a house: either by stealth (making low cards valuable) or strength (making high cards valuable). But then what about the middle value cards? They wouldn't be valuable in either situation. The solution was to make the middle card a regular value – and also a modifier to both high and low cards. Now every card is worthwhile and invading houses was thematic, too.

The other very important concept came from more perspiration: reading historical non-fiction on medieval Japan. Ninjas have appeared in board games before, but rarely with any sense of history. After reading the book The World Turned Upside Down by Pierre-Francois Souyri, it was immediately clear to me that we needed to use the clan rivals and the rise of the Bushido code as a backdrop to the game. Early on I was convinced to have indirect conflict in the game. To me, this means each player has a plausible, intermediate reason to interfere with an opponent. It isn't direct – there's always a mediating mechanism. The clan rivals of 12th century Japan were thematically perfect for this. Players would support and oppose a clan rather than simply themselves. If that interfered with other players, it was plausible deniability. They weren't directly attacking an opponent – they were merely furthering their own goals. This theme framed the game and refined all the ideas behind it.

Another important aspect of the game design is production values. I've watched BGG for years now and one thing that creates buzz for a game is the illustration for the board and the box. I'm a believer in great artwork for a game and to get that, you have to reach pretty high. I cold called several great board game artists, but none were available (and all were very polite). Then I saw Drew Baker's art for Legend of the Five Rings. I met him at Gen Con in 2008 and he agreed to do the game – I think mainly because it was a challenge to do a board game. Everyone can see he was perfect for the job, and arguably he single handedly brought Ninjato into the spotlight.

I also knew we needed something more than just illustrations. I am a passable graphic artist, and although Galactic Emperor was all right, a professional can do so much more. Like many, I've been a huge fan of Peter Gifford's work with game guides on BGG. At first, we thought we'd just use him for a terrific game guide for Ninjato, but he agreed to do the entire game with us. I couldn't have been more pleased. Peter put in a crazy amount of hours on Ninjato.

Last but in no way least, the playtesters made the game. Some graciously moved with us through terrible versions. Some refused to ever play again (understandably!). Some made truly brutal criticisms (that proved to be invaluable). But what I'm most happy with is even after we were ready to go to publishing and were surprisingly picked up by Z-Man Games, we continued to refine the components and the rules. I'm a big believer in letting a game go when it is done. You can't keep making adjustments forever because at some point, you're just moving things around and no longer improving. But this time, we continued to respond to playtest feedback. Zev made the suggestion that collecting cards is rather boring. We responded with a very smart change to give a player more cards if he has fewer in his hand, thus encouraging players to try to play perfectly, to not have a card left after attacking a house. Since the game plays for seven fixed rounds, making the optimal move every time – a move with as little waste as possible – is not only thematically perfect, but creates a natural tension and story arc, something that Dan and I believe every game needs.

Ninjato is our second published board game. We have a few others already in the works - including a murder mystery game set in 1938 New York City, another from our computer RPG design experience applied to the epic fantasy board game genre, and a third game based on the thousand year history of Athens. I'm extremely excited about each of these. Some games literally jump to life from the aether; others are a long hard journey. Ninjato was certainly the latter. Simplicity is a most difficult task, but a hard journey can be very rewarding. Most of all, I hope you really have fun with the game.

Adam West

(For more from Adam West and Dan Schnake about the creation of Ninjato, you can check out this video on the BGG game page. —WEM)