The fact that the new system is a purely optional didn't offer me much comfort. Isn't this how scope creep happens? I did not want Root to become the sort of game where players were faced with a blizzard of toggles when they opened their box. I don't want it to be an endless sandbox where no two groups are ever playing the same game. I came of age as a designer during the days when every group played Twilight Imperium: Third Edition with such a varied set of house rules that they could hardly be said to be playing the same game. I didn't want that for Root.
Those are two very big strikes against the new system. More than once during the development of the hirelings, I wondered if the whole effort was a mistake. It seemed reckless to mess with the game in such a fundamental way. Today, I want to talk about why I think it was worth it.
Originally, my role in this expansion was limited. It was very much Patrick's expansion to design. The only thing I wanted to do was revise how Root did setup. There were a couple reasons for this. First, I thought it was important. Root was not originally designed to be expanded so robustly. Mostly the system has adapted well to the demands we've put on it, but that's not true for setup. With every faction added to the game, the setup rules had to be stretched and bent. I knew I wanted to fix this and try to structure the rules so that they support the game for as long as it was needed. This didn't seem like a huge job, and it was one I was excited to do. After Oath, I was tired! I badly wanted a small job like this so that I could enjoy a creative cool-down.
I built the new setup rules on a simple premise: I wanted to combine the game's setup wit ha faction draft that would be used to quickly generate a scenario that the players could play. For the game's draft, I started with the existing Plus-One draft, pioneered by fans over on the Woodland Warriors (especially Bott_bott, Guerric S, Justin K., and MarcustheCat) as well as Matt and Hunter over at the Space Cats and Peace Turtles Podcast. The hope was to use our huge testing resources to take all of that good work and build it into the core rules.
While providing support for competitive Root was a nice benefit, the biggest reason to build the setup rules was simple: I wanted to help Root players play more Root. We've all been at a game night where the first 30 or 40 minutes is spent deciding what game will be played. A big modular game like Root further compounds this because even once you've decided to play Root, you've still got to decide what you mean exactly by saying you're going to “play Root.” Which factions? Which maps? I hoped that the setup cards and draft could serve as a kind of scenario generator that players could use to quickly decide what content to use.
In general the efforts to redesign the game's setup went very smoothly. I was happy it preserved the variety of match ups offered by the reach system, and even happier that it also offered a huge amount of flexibility to players that could really upset the game's meta.
But, as I was expanding the system so that it worked at all player counts, I started running into problems at the lower counts. These weren't new problems, but the setup system was bringing them into sharp relief.
Root as a Two-Player Game
Root can be played as a two-player game with no adjustments to the game's design. If I have one design commonality across everything I've worked on, maybe that's it. I hate rules that adjust for player count. I don't mind a few small alterations, especially when it comes to setup, but I don't want to ever design a game where the action menu changes radically depending on player count (I see you, Uwe).
The trouble with Root as a two-player game is the trouble of any multiplayer wargame at two. With more than two-players, combat is never zero sum in terms of the game's broader ecosystem. When I fight you, I am also likely helping a third party who isn't participating in the combat. That's not a bug, it's a feature. I have to decide if the attrition I'm going to suffer for fighting you is worth it. After all, next turn I might have to deal with that third party. At the same time, I know that if I come out of the fight badly hurt, that third party will probably be picking on the other player, not me. This isn't the case with two players. The ecosystem of Root can't re-balance itself properly. This means that there's nothing to stop a player in a strong position from ruining a player in a weaker position.
The new setup system actually made this worse! Because it gave players a lot more setup flexibility, players could find themselves in very, very strong starting positions (and also very, very dangerous starting positions). This wasn't a huge problem in multiplayer games, because players might take a gambit on a tricky starting position and hope that they can manage the game's ecosystem. With two players, there was nothing to stop your opponent.
This dynamic actually exists at every player count, it's just worse with two. I usually think about this in terms of herd control. With more players, the herd of players who are not you gets larger, which means that it's easier for the ecosystem to self-correct. At lower player counts, that herd is smaller, which means that knock-out blows are easier to accomplish.
I knew Patrick had been thinking about the two-player problem for awhile and he had planned to tackle it in the expansion, potentially with a special scenario or map. But, the more I worked on the setup, the more I didn't like this approach. I didn't want the two-player game to be quarantined off from the rest of the design. The idea of a smaller map is an interesting one (and something we might still explore), but I didn't think it would solve the problems with the two-player game. Ideally, I wanted a solution that would scale up through the player counts and potentially be worked in to new setup system organically.
So I asked Patrick very nicely if I might commandeer part of the expansion, and he was kind enough to relent.
How Do You Solve a Problem Like Root's Two-Player?
I started by just setting up the game's basic two-player scenario: Eyrie vs. Marquise. I have played this scenario a lot. Back when we were doing testing, Jake, our former ops director, and I used to play a match almost every morning for weeks. The balance is interesting. The scenario favors the Eyrie, but if the Cats use their early marches well they can build a strong enough line to hold them at bay. With a few bird cards drawn for extra actions over the course of the game and a well-timed ambush, the match can sometimes get very close.
But sometimes it's a knock out. If the Eyrie hits the Cat hard and fast, the game might be over by turn three. Another way of thinking about this is that the Cats couldn't make more than one mistake and win. The Eyrie had a lot more leeway. Still, it was viable enough to be included in the game and to this day I get messages from folks who like the two-player game.
Next I set up the Eyrie against the Duchy. This is a very odd match up because so little of the board is used. Here I think the Duchy is the clear favorite because they can so easily get around the Eyrie's forces. The board is also so empty that so much of the pressure that makes both factions interesting just isn't present. And, like the previous match, early knockouts are common.
These early cases gave me a few general design goals. First, the board was simply too empty. There's a reason why the Cats improve pretty much any game they are in and that reason has to do with their setup. Their thin line of warriors throughout the woods provide a baseline friction and texture to the map that just isn't there when the map's empty. I needed the two-player game to not be so empty.
The second design goal had to do with addressing those early knockouts. A player in a weak position needed some viable path to bounce back. This is one reason why I didn't like approaching this problem with a smaller map. The issue was not that the map was too big. The issue was that the game's ecosystem was too empty. There wasn't enough shrubbery to hide behind and regroup.
Hirelings Enter the Battle
As often happens, the solution for this problem didn't come from any one particular place. Rather, it was the extension of thinking about a few different things at once. Months ago, the company's accounting manager, Ted Caya, had asked me if I could design any small Root products. He asked about this because the Vagabond Pack and Exiles and Partisans deck continued to sell really well and were clearly resonating with our audience. This was an interesting problem. Ideally we'd just design single faction packs, but, because Root factions have big player boards, they can't be easily packaged as a $15 product.
I had also been thinking a lot about proxy powers in wargame design. In early drafts of Oath, the 6 suits actually had their own armies that players had to win support of. I really liked the idea of the players fighting over minor powers. And, of course, some of my favorite wargames (The Napoleonic Wars) and adventure games (Magic Realm) contain minor factions. They add so much color to a big game and make the world seem considerably less empty.
Then there was Undaunted: Normandy. Over the summer, we ran a skeleton crew at the Leder Games offices while we finished Oath. During lunch breaks, developer Nick Brachmann and myself played a lot of Undaunted. I'm not a huge player of World War 2 games, but I was compelled by the quality of the design. The various wargame concepts of supply, morale, command, and attrition were smoothly folded into a modern design aesthetic.
One of my favorite things about the game was how it tackled initiative. Basically, players would bid part of their hands to seize the initiative and determine which player went first. The player holding the marker would win ties, but winning wouldn't give that player anything since they already had turn order. But, if the player without it won it, they would get a double turn and could perform stunning maneuvers. Though some of the game's other systems were flashier, I think this is the best piece of design work in the game.
While we played, Nick and I talked a lot about the importance of tempo in two-player games. I thought about the concept of “sente” in Go or what it feels like to be on the defensive or on the attack when playing Netrunner or in Starcraft. Some kind of explicit tempo management seemed absolutely critical to the design of a two-player game.
When I started working on the two-player game, I knew some kind of tempo control needed to be in the design. My first thought was that it needed to be something that the players would bid on. But I didn't have anything they could bid. The development of the Corvid's during Underworld had shown very clearly how dangerous it was to mess with player hands and the game's card economy. If I wanted any kind of bid like Undaunted, I needed a medium of exchange.
As usual, there were so many possible ways of addressing this issue I feel like my design process never starts with a white page. Rather it starts with a very full page of options, and each day I just try to erase a little more until all that's left is the thing that needs done. Design isn't about generating ideas out of thin air. It's being backed into a corner by the demands of the problem.
So, I started by considering the design's demands. First, I knew I needed to think about the changes to the two-player game fundamentally as a catch-up mechanism. They needed to provide players in a weak position with a path forward and provide players in a strong position with meaningful challenges that would replicate some of the feeling of a larger game. Already I knew that tempo control offered one possible solution to this problem, but that solution likely required some kind of currency that could be spent on managing it. I hated the idea that I would spend all of the rules complexity of introducing a currency on only a single thing, and so quickly it became clear that players should have other things to spend their catch-up currency on. Here the idea of minor factions or hirelings jumped out as a clear solution. Furthermore, if I could keep their physical profile small, they might even provide a nice addition to the Root-line, allowing player who just want to beef-up their two player game an inexpensive option while also making the operations team happy.
I also quickly realized that the hirelings could be modeled on existing Root factions. This had a number of advantages. It gave each hireling design a clear mechanical and thematic touchstone. It also allowed us to reuse colors for any new pieces if we kept them mutually exclusive—that is, you can't use the Marquise-flavored hireling if you were playing with the Marquise. This might sound restrictive, but it actually opened up huge areas of the design that were previously locked. For instance, factions can't really mess with the discard pile because that's an area where the Lizard Cult works and Josh Yearsley (the game's editor) and I don't want to have to write rules for that interaction. But, if we built a hireling that interacted with the discard pile, we could just color it yellow. Heck, we could even have several different yellow hirelings so long as there was a rule that make it clear that there could only be one of each color in the game. This also gave Kyle permission to do more art that would riff on the existing faction designs he had built as well as new meeple shapes with the same colors.
All of this was terribly exciting. I just had to build a design that worked and we'd be off to the races.
The Design That Did Not Work
My first attempt was both a catastrophic failure and a near miss. I designed a handful of hirelings (then called minor factions) and a very basic currency generation system as well as a post-turn phase where players could act using the minor factions they controlled and spend their currency, called influence, on acquiring new minor factions.
Influence was generated two ways. First, you got influence depending on how far away you were from winning. Second, you gained influence by doing bad in battle. If you were a defender and a “0” was rolled, you gained an influence. This was to introduce a little noise into the influence system. Critically, I didn't want players to be able to game the system too hard for strategic advantage. A little positional dance was fine, but the system had to be noisy enough that you'd never quite know when you'd get your influence windfall.
Influence was spent bidding on various minor factions. The economy was largely zero sum, and players would often steal influence from each other once the starting influence pool was exhausted. If you took control of something, the old bid would be returned to the previous holder of the card. I took this system from Oath, which was used to measure the control of the banners and worked well in that space.
Initial tests were promising. Nick and I played the game a lot in the two-player format and it seemed to work well. We tested it a bit with three players two, cutting down the number of hirelings by 1 to adjust for the player count. This is where the problems started to appear. At that player count, there was more influence in the game but less things to bid on. There were other problems too. The turn order bidding was becoming more powerful with more players and more disruptive to the flow of the game. That was further compounded by the complexity of the new minor faction phase at the end of each player's turn.
To fix the problem with the number of minor factions in play, I created the demoted side of each hireling. Basically, hireling had two sides: a regular size and a demoted side that would be used in higher player count games. The demoted side generally didn't feature any piece and was instead just a special power that could be passed around the table. Imagine a crafted card that you could bid on and fight over.
But the other problems remained. Generally it was all working. Strategically, the new material was offering some compelling options. But the new phase was really disrupting some of the game's natural flow. I wasn't sure what to do, so I decided to keep the design in a holding pattern and widened our testing circle to include some dedicated Root players.
The first reports were not encouraging. While folks liked the new setup rules and enjoyed the powers the hirelings introduced, in practice they were too clunky and scaled horribly. This was confirmed in internal tests. Patrick was really unhappy with their direction, and I didn't blame him. Even though the hirelings were making the game more compelling, the cost in flow was far too steep for even the most dedicated players.
The core of the problem was two-fold. First, the influence system and end-of-turn hireling phase was clunky and flow-breaking. There was just too much to think about. Second, the initiative bid was straight broken at any count outside of two-player. But, despite that, I still felt like the hirelings were offering something really special to the game. If I could fix those two problems, there was a lot to gain.
I first tried tackling the problem of influence. I set up a game of Root in my office and rapidly worked through several scenarios. I loved what the granularity that the influence bids offered to the game in theory, but often it was clear that the granularity just wasn't needed. Basically there were three archetypes that a hireling fell into: they were uncontrolled, they were controlled weakly, or they were controlled securely. It really didn't need to be any more expressive than that.
So, I got rid of having the bids stored on the hirelings and instead went to a system were players just paid a cost to move a hireling from one position to the next. A weakly controlled hireling went in front of your board. A securely controlled hireling went to the table edge close to where you were sitting. This change went a long way in making them easier to think about, but the influence accrual, especially the fact that some of it it was generated during battles, was still too clunky.
The reason I had put it into battles in the first place was that it introduced noise into the system and that noise was really, really important in creating uncertainty in the two-player contest. I wanted the hirelings to be reliable enough to encourage players to take risks, but not reliable enough that a big gambit couldn't come back and bite you. With this in mind I scrapped the influence generation wholesale and rebuilt it entirely around that demand. If risk and uncertainty was the whole point, why not make it central to the system?
This lead to the creation of the influence die.
Each face has two elements. First, it has a number of circles. That's how much influence might be gained if you roll that face. The second element is the victory point threshold. The influence on a dice face is only gained IF you are behind by at least as many points as indicated by the threshold. So, you roll the “7+” face, you only gain 3 influence if you are behind by at least 7 points.
I kept the cost system simple. Any movement of a hireling cost 1 influence. Hirelings began the game in an uncontrolled zone. To shift one to your control cost 1 influence. To shift a hireling you controlled to secured cost 1 influence. You could also move the hirelings that other players controlled. So, if I had 2 influence to spend, I could unsecure your hirelings and then spend another influence moving it to my control. The only exception to this was the initiative card, which cost three influence to shift.
The dice also had some nice fringe benefits. Because they were rolled at the end of the turn, the passing of the dice could serve as a nice mnemonic that integrated more cleanly into the games flow. And, I could now easily scale how much influence was entering the game. In a two-player game, players would roll two dice and use both results. With three players, you'd only use the result that would give you more influence. In games with 4 or more players, you just roll one die.
I also went through the hirelings and reorganized the powers so that they didn't occur in some special phase. Here I was glad to have a lot of experience creating power templates from Oath. All of the powers basically fell into three or four different groups, and a big icon could go a long way in helping a player visually parse their actions for the turn.
Finally, I reworked the initiative card so that it had a different power at higher player counts. Because shifting the start player was simply too powerful and disruptive in all games but the two-player, it changes it's ability to a "treaty" which can be used to stop two players from fighting each other at higher player counts. This allowed some small-scope tempo management at higher counts being as disruptive.
When this version went out to testers the response was very positive and mirrored the encouraging feedback I had received internally. There was a lot of work to do still, but the new hirelings integrated smoothly with the setup draft and really fleshed out the two and three player games. They also worked pretty well in bigger games of Root, especially for experienced players who wanted to spice up their meta. What if the Corvid Conspiracy had access to the Marquise's field hospitals? What if the Eyrie could use the Woodland Alliance's guerrilla tactics? Maybe the Marquise could benefit from having a the help of a state-sponsored Lizard faith.
At the same time, I recognize that these sorts of questions may be less exciting to players who find Root already too much to deal with. We've tried to structure things so that players who back our Kickstarter will have access to lots of different ways to play and provide players who want to go deep into this new system with a lot of options for exploring it.
I can't wait to show you everything we have in store.
- Cole Wehrle
Find the original post on Board Game Geek.