Richard Garfield’s Designer Diary for King of Tokyo Duel ⭐
Iello first asked me for a two player version of King of Tokyo in 2020. It was an immediately appealing idea, there was no reason the formula of rolling dice with some combat and some victory points and a mess of special cards shouldn’t work well for two players, but it certainly was lackluster with King of Tokyo. A lot of King of Tokyo was designed around making combat work for 3 or more players in a nice way; a way where you didn’t have too much control over who you damaged and you had a lot of control over whether you put yourself in danger. With a 2 player game all that was wasted design; and didn’t really work well.
Dice Faces:
Deciding on the faces for the dice was fun. I knew I wanted energy, health, and damage. The remaining faces in King of Tokyo are 1,2,3; with which you score victory points. I decided I wanted a bit more structure to this part and settled on a destruction track and a fame track which acted as a tug of war. Getting one all the way to the end, or both decently in your favor were the two non-combat ways to win. This left one face, which I decided to dedicate to special monster dependent powers.
The Shop:
At first the shop worked like King of Tokyo, but I soon decided that I wanted to see more cards in the game than in King of Tokyo. In King of Tokyo I usually thought of players having 3 directions – combat, victory points, or energy; and I wanted players to be able to succeed without being forced to do any particular avenue. In particular, players can pretty much ignore the cards if they want to; and they can do fine that way. When players did that – generally the cards still impacted play through other players. The cards provided variety and interesting situations; but there wasn’t a reason that every player needed to participate in it. With 2 players the game felt kind of empty when neither got any cards.
To this end, I added a discount slot so that one of the cards was a good deal. This was fun, and added some exciting situations to the game – but it wasn’t enough. The one energy discount could be good but it wasn’t good enough to encourage all players to buy cards. If it were that good – the discount would dominate the game too much anyway, I realized you really couldn’t solve the problem with a discount. Giving players one energy on turns they didn’t buy something was a really solid answer to the problem. With this, players completely ignoring energy would eventually accumulate enough to make purchases. Also, it made the act of not purchasing something still feel like one was engaging with the card buying system – and opened up new ways to handle your energy. It was a viable and exciting strategy to make many purchases on one turn, in order to gain energy in the turns between your shopping sprees. Had I come up with the concept of gaining energy on turns you don’t make a purchase – I probably wouldn’t have added the discount slot as well – since it solved my problem. However, as I was already playing with a discount slot, I found that the system combined well with the energy accumulation, and that this played into the exciting multi-purchase turns that were cropping up from time to time. Usually I see games as having a complexity budget, and extra systems like this need to bring a lot to the table to make them worth adding to the game. The play value that the combined system was bringing was great, however, so I kept both systems in place.
Balance:
Getting the right relative strengths for the monsters was an issue. One of my goals with design is to not ‘overbalance’ the game – to get the balance right for a narrow group of good players rather than for a broader audience of players. My favorite illustration of this concept comes from a digital game I did many years ago (with Alexey Stankovich and Skaff Elias), Spectromancer. In Spectromancer there was a free class, the Cleric, which was widely dismissed as being underpowered. We expected players to believe it was underpowered, but we didn’t choose it because it was underpowered, we chose it because it was flexible and showed off a lot of what made Spectromancer fun. On the other side was the Necromancer which was widely believed to be broken. With games this comes up ALL the time; but this was the first time I had been in a position to actually check the statistics and they told an interesting story. Cleric WAS underpowered… for beginners. It actually had a 55% win rate for experts! Similarly, Necromancer was overpowered for beginners, it was about even for intermediates, and for experts it was a dog at about 45%.
A common design philosophy would have pushed us to make Cleric less powerful and Necromancer more powerful, so they were both balanced for top level players. It is easy to see that this approach will make the game worse for beginners and intermediates; they already thought that Cleric was underpowered and Necromancer overpowered – by doing that we would risk making the game really unfun for those players. Of course, balancing to their complaints risks making the experts situation worse – buffing Cleric so the beginners and intermediates were happy would make it the only viable class for experts. As it was; out of 6 classes, there were at least 2 that were favored for beginners, intermediates, and experts – and all 6 classes were favored for one of these sets of players; and I thought that we really couldn’t do better than that. And, it was pretty lucky since we didn’t intentionally design it for that.
Anyone who has played King of Tokyo duel will see why this is relevant. All the monsters were viewed through this lens. The one that got the most attention was Mechadragon, because the power was dead simple – the special faces were multiplied by the attacking faces; so 3 claws and 2 special faces were 6 attack! At first this seemed crazy, and our games sometimes ended in fast wipeouts – but the more we played and adapted our strategy to the opponents powers the less it seemed overpowered. In the end, it felt like the Necromancer from Spectromancer; it was favored for beginners, because their opponents didn’t know how to play around it; but not great for someone who knew the game well. Of course, it will win some games sometimes without contest, and that might loom large in ones memory. With a game like this I aim for the better player winning maybe 70% of the time, not 95% of the time.
Anyone who has played King of Tokyo duel will see why this is relevant. All the monsters were viewed through this lens. The one that got the most attention was Mechadragon, because the power was dead simple – the special faces were multiplied by the attacking faces; so 3 claws and 2 special faces were 6 attack! At first this seemed crazy, and our games sometimes ended in fast wipeouts – but the more we played and adapted our strategy to the opponents powers the less it seemed overpowered. In the end, it felt like the Necromancer from Spectromancer; it was favored for beginners, because their opponents didn’t know how to play around it; but not great for someone who knew the game well. Of course, it will win some games sometimes without contest, and that might loom large in ones memory. With a game like this I aim for the better player winning maybe 70% of the time, not 95% of the time.
Did we end up in the right place? We analyzed the success for different levels of players, and, like with Spectromancer, all the powers were favored for some audience, and none with a winning record that was by my estimation out of bounds. It is possible, of course, that with a broader community we might find that there are groups of players that we didn’t account for, and time will tell if we need to adapt something to the realities of play. Games are often tweaked to more casual modes, or more serious modes – as the community demands – sometimes independently of the publisher.
Speed:
It became apparent early that this game was subject sometimes to very fast resolution. I began to view a game as more like a hand of gin than a game of gin. You sometimes just win; but if you play the whole game those lucky hands cancel out – and you are left with a competition that has swinginess but isn’t, in the long run, dominated by that luck. The games variance in length became one of its charms for me; it was tense from the very start. When I play myself – though – I usually play best of 5 games. This usually gives me a pretty satisfying experience – but – the longer games are such that you can’t really make that the standard way to play. It would scare people off seeing ’90 minutes’ on the box. But if you like the game enough you will probably find that way of playing worth the time.
Buzz Tokens:
Buzz tokens were suggested by Max-Tobias Walter my excellent developer at Iello, he wanted to see more done with the tug of war tracks. The idea of building a sort of cityscape, with bridges and buildings was talked about. It was a fine idea, and added a lot to the game. My only complaint, in the end, was the monster icons on the Buzz tokens; I think the game would have been far better off if it had letters, names, or icons that were clear – like a silhouette of a tower. I am not blaming Iello, they allow me to review everything, so it is as much my responsibility as theirs.
Art & Graphic Design:
The art by Quentin Regnes and graphic design by Vincent Mougenot were outstanding. The look has a wonderful mix of excitement and playfulness that I look for in swingy games like this – and the treatment to look like a comic book really works well for me.
Thank you to Richard Garfield for this great Designer Diary, originally published on Board Game Geek.