The Black Orifice

Tabletop gaming resources and events from grumpy old games designers, Ben Redmond and Nigel McClelland

Some Thoughts on Games Design

Being a teacher, the summer holiday can sometimes leave me with time to think about stuff that I maybe wouldn’t get to think about if I had a Proper Job, especially towards the end of the holidays when my wife is back at work and I’m stuck at home with the kids and no money left to go out and do something interesting. This summer I had some thoughts mulling around about games design, which I thought might be interesting enough to make into a blog post.

I had spent a bit of time at the start of the summer starting to put together a game of my own. I had the idea that Etherscope would make a great skirmish game, emphasising the cinematic Matrix-style combat of the setting. I got as far as writing a first draft of the rules and making some of the components I’d need to play the game. However, there was something about the game that I wasn’t entirely happy or comfortable with, but I wasn’t sure what. That led me to reflect on the games I play, or have played, and break them down into the mechanical components that make up the game. This blog post is a write up of that process.

Objectives

One of the key differences between games is how they handle objectives. In many games, a primary objective is just to kill the opposing force. In fact, in most Warhammer Fantasy and X-Wing games I have played it is the only objective. However, there are other objectives used in many games, often involving being the only player with models close to a specified position or area on the game board. In some games there is an adapted version of the kill mechanic, where you are required to target certain units or models to win the objective. Warhammer 8th edition’s Blood and Glory mission, where you have to target a number of banner carrying units or the army’s general, or (although I’ve never played it) the assassination victory condition in Warmachine, are good examples of this type of objective. Guild Ball and Blood Bowl stand out amongst the games I have played for having an unusually complex objective: scoring a goal. In mechanical terms, the goal objective involves claiming position of a special objective marker, moving that marker to an area of the play area, and, in Guild Ball’s case at least, performing a special action to score points. Campaign play in Imperial Assault can sometimes, depending on the mission, have similar objectives involving moving objectives between locations and performing special actions to meet the mission objectives, and not always requiring both in the same mission. In Guild Ball and Blood Bowl the goal objective is made doubly interesting by the fact that both players can score with the same objective marker, making the fight for control of said objective marker (the ball) an important part of the game.

This, then, leaves us with five different objective styles:

  • Kill the Enemy: The staple of most miniatures wargames, this objective is always important, if indirectly so (if you kill all the enemy they can’t score themselves). I don’t know if I’d like a game that didn’t have this as an option, but in many ways it can be seen as the vanilla option.
  • Targeted Kill: Where the mechanics used for taking out an opponent’s models or units are used to target specific models or units, adding an element of protecting these models to the game play. This style of objective adds to the variety of the game play without requiring completely new mechanics to handle it.
  • Static Location Objectives: These objectives are common in skirmish-based wargames, where you must try to manoeuvre your forces into a position of dominance in a given area. In my mind, these are the least interesting of mission objectives. Using such objectives often reduces variety in list building and can often lead to long range, static shooting trade off games where models just try to hold one objective and shoot the enemy off another. Further, if there are several such locations to claim, it can lead to some models being left out of a game after claiming an objective if the opponent decides not to attempt to claim that particular objective. The point of these types of objectives, from a game design point of view, is to presumably reward players for actively moving their models around for position. It is ironic, then, that this can often lead to more static game play after some objectives have been claimed.
  • Mobile Objectives: These objectives involve models retrieving a marker from a particular location or enemy model, and then moving its location. Such objectives can make for complex and stimulating game play, requiring players to position models to both claim the objective, defend it whilst in transit, and get it to where it needs to go.
  • Action objectives: This type of objective covers anything that involves you completing some sort of special action to score the objective. The biggest effect that these types of objective are likely to have on the game is in list design. Having the right specialists to complete the required task is an important factor in list design. Whilst this can in some circumstances have a similar negative effect to variety in list building as mentioned for static location objectives, with the right mix of different actions needed, it can start to have the reverse effect of adding to list variety as there becomes a variety of ways to achieve the objective. This objective type works best if used in conjunction with another type of objective. Goal scoring in Guild Ball is a combination of mobile and action objectives, for example.

I’d like to think that the best games will have the biggest variety of objective styles used, however, a game I really enjoy is X-Wing, which (realistically) only uses the kill objective. And given that X-Wing is now the biggest selling miniatures game, others must agree. Even Guild Ball, my particular favourite, only uses one objective beyond the kill objective, although that objective is sufficiently complex to inspire deep strategic gameplay. It all leaves me somewhat puzzled as to what a winning formula is for games design. One thing that I think both X-Wing and (probably more so) Guild Ball do very well is keep the game alive and competitive for the entire duration. In Guild Ball, control of the ball will always give you the ability to score a big chunk of victory points, and the fact that your “killed” players get to come back at the end of a turn keeps you in the game even after taking a kicking. In X-Wing I think it’s more a case of the narrow firing arcs and fixed-manoeuvre movement mechanics. If you go behind early on you can fly yourself into positions where you can attack your opponent without being attacked back, to try and help you recover the deficit.

It seems, then, that the key to objectives in a game is to ensure that they add depth to the play, but also keep players competing against each other, and in with a shout of winning, for the full length of the game.

Turn Sequence

On initial inspection it’s easy to see two different styles of turn sequence: the “you-go-I-go” of Warhammer, and the alternating activations of Guild Ball. My opinion on these two methods is that alternating activations is intrinsically better, but not always appropriate. It allows you to react to your opponent’s actions and keeps you involved for much of the game rather than spending large periods where you are merely spectating. Alternating activations require you to have a similar number of models to activate so that one player doesn’t gain the upper hand by being able to lock their opponent out for too many unanswerable actions. How a game handles this problem can vary, but there will inevitably come the point where alternating activations becomes unworkable, and at that point the You-Go-I-Go method becomes a necessary evil.

Of the games I play, Guild Ball handles alternating activations the best. By having a set number of players on your team it keeps the balance between each players activates close, allowing players to return to play quickly after being taken out maintains the number of activations throughout the game, and its influence mechanic provides a chance for a team to still perform at full potential, even if they feel they need to run and hide some of their players for fear of being taken out.

However, X-Wing provides a third way, perhaps closest to alternating activations, but where models are activated in a set order, as determined by a score for each model. This way you can have a squadron of 7 ships compete against a squadron of 2 ships without having to revert to a complete you-go-I-go style of play. Mechanically, I think this is the real genius of X-Wing’s game design, however, it only works because of a few specific peculiarities of X-Wing as a miniatures game. Firstly, unlike most miniatures games, X-Wing has only one way of attacking the enemy. Most miniatures games offer a mixture or ranged and melee attacks, but in X-Wing you can only attack at range. Secondly, the combination of fixed manoeuvres for movement, the fact that you have to choose your manoeuvre for each ship in secret before performing any moves, and the narrow firing arcs, add important limitations to what you can achieve within a turn, so that play becomes less reactive than a typical alternating activations game. Despite that, the game manages to create high levels of tension and excitement because of the movement dials – the fact that you have to try and second guess your opponent and think several turns in advance to maximise your positioning.

This fixed-order method is an interesting mechanic, and one that I borrowed for use in my Etherscope miniatures game, but I’m not sure it necessarily works without all of the restrictions in place that X-Wing uses. I like the fact that you have to commit to much of your actions for the turn before you get to see anything of what the opponent has done, and a similar idea can be seen in Guild Ball with the distribution of influence in the maintenance phase, but this provides much more flexibility than X-Wing’s dial setting and leaves you with reactive options unheard of in X-Wing.

Probability

Practically every miniatures game involves some element of random chance. Most often this comes in the form of a dice roll, but can also come from card draw or a bidding from a limited resource (the Game of Thrones board game has several different bidding mechanics that make the game rich and strategic). With dice mechanics there are several things that you can play with that will change the range of possible outcomes: changing the number of dice rolled, or the number of sides on the dice, combining scores or binary results or binary dice pools, varying a target number, whether modifiers are decided before or after the dice are rolled. It seems that the vast majority of games use a binary dice pool mechanic for their dice rolls, where a pool of dice is rolled and each dice is either a success or a failure depending on whether it is below the target number or not. Even X-Wing in it’s unusual way, with positive and negative dice and pictograms for success and failure, is essentially a binary dice pool mechanic.

There’s several good reasons why a binary dice pool mechanic is idea for miniatures gaming. Firstly, it means you’re using the binomial distribution, which means your results will follow a nice bell-curve pattern, with extremes of good and bad rolls being much harder to achieve than the middle ones. Secondly, they are easier to calculate the final outcome, as you’re counting successes rather than adding up different numbers. I’m going to compare X-Wing and Guild Ball again, mainly because they’re the two games I play most often. In these two games, dice pools are usually around the same number of dice, with the majority of rolls involving between 4 and 8 dice (X-Wing may appear less, but remember that the full outcome is determined by a combination of both red and green dice). This seems to be a sweet spot for game design – it gives you a properly curved distribution and doesn’t require you to roll more dice than you can easily fit in your hands.

The use of cards drawn from a shuffled deck also has an interesting effect on probability. After a card has been drawn, the chances of drawing that card again go down, whilst the chances of drawing a different card go up. CCG and LCG players will tell you that you should work towards building as small a deck as you can get away with, to maximise the chance of the best cards being drawn during the game. In X-Wing we can see this effect in operation with the damage deck. Once you’ve drawn a particular critical hit, your chances of drawing it again are reduced, hence maximising the variety of critical hits you are likely to experience in a given game and making gameplay more interesting. I think the use of cards is an area of games design that miniatures games could explore more, and certainly something I would be keen to include in my Etherscope game.

List Building and Balance

List building is an important part of any miniatures game. Given the nature of the internet age I might even go so far as to say it is quite possibly the most important part of a miniatures game. People like to go online and share their list ideas with other who play the game. Whilst this can have a negative effect on the community in terms of reducing list variety, it is crucial in bringing players into the game. A miniatures game needs to give player options and choices to make. Deciding what models to take and what options work well in combination with each other can be as fun a part of the game as playing the game itself. However, the goal of providing options runs directly counter to the goal of having a balanced game. Chess, for example, is a perfectly balanced game, but it doesn’t generate discussion as to whether black nights are better than white bishops.

For the majority of games, list building comes down to a matter of points being spent on choosing your force for the game. In my opinion, this method is not without its problems. Getting the points score for any given unit right is a tricky task, and potentially one that changes with each new release for the game. This is particularly difficult if a model can be included multiple times in the list. Those who follow FFG’s Imperial Assault will know that 2015’s World Championships was dominated by a single list. The list, made up of 4 royal guardsmen and 4 imperial officers was the only list that stood a chance of winning because the designers had got the points-to-power of the cards spectacularly wrong. The list was especially strong because the models abilities combined particularly well and multiplied the effectiveness of what were already probably undercosted models.

Having unique named models that you are only allowed one of provides some limiting of the potential imbalance for a particular model, although as Warhammer’s End Times proved, a single super-powerful model, such as Karl Franz or Malekith, can have a dramatic effect on the game if they are so powerful as to be almost unstoppable. X-Wing does a good job of balancing the power to cost of its bigger named pilots against the option to have multiple generic pilots. There are successful lists that use one or the other, or a mixture of the two. This is achieved by restricting the amount of power increase a named pilot gets over a generic of their particular ship.

I don’t think I can talk about balance in X-Wing here without going into a bit more detail, as it is a game that leaves me with very mixed feelings over its game balance. Essentially there are a number of strong lists that are pretty well balanced against each other, and within many of those lists there is good room for tweaking and personalisation. On the flipside, there is far too many of the models in X-Wing that are unplayable if you want to compete. The designers do a good job of trying to bring things back into balance, but often this just provides a gradient for power creep. For example, one list archetype that is often referred to as the epitome of game balance in X-Wing is the TIE swarm, yet the TIE swarm of today is very different to the TIE swarms that were being played at the top tables of Worlds in 2012 and 2013. I’m not sure a TIE swarm of Howlrunner, 5 Academy Pilots and one other named TIE fighter will compete against todays’ Black Squadron Crack Shot swarm. There is no way to say that that is anything other than power creep. However, it’s well managed power creep, and their focus on fixing the underpowered (or overcosted) ships is truly great. I played Beastmen in Warhammer for years, and whilst Games Workshop felt the need to bring out a free supplement to nerf Daemons down to some sort of more reasonable power level, they had no great desire to give a boost to the weaker armies. All in all, X-Wing is a great game and very well balanced, provided you know what are the strong choices.

I have been waxing on to people recently about Guild Ball being the most balanced game there is, and an important part of that is the fact that the team is made up of 6 unique named players. So even if Shank is just a little bit better than all the other Butchers players, you can’t have 4 of him, so you’re going to have to think about who else is going into your team. Guild Ball is also fairly ground breaking in that it doesn’t have points values. You play a game with a set number of players and each is balanced to the same power level. You have one captain, who is notably better than the rest, and one mascot who is a lot worse than a standard player, but the other four are all unique named and equal in power balance. I’m not saying the balance between each player is perfect - Some players are weaker than others, some are stronger - but the with every player being a unique individual that you can only field once in your team, any small imbalance in power is quickly mitigated by the other players you can choose from. Where you can argue that Guild Ball falls short is the variety available when writing your list. You have a choice from 2 captains and 2 mascots, and then you can choose 4 players from the 10 or so available. It’s not quite the variety of lists you can have in X-Wing, where lists can range from 2 to 8 ships. Guild ball is more like a game of X-Wing where you have to build lists with 3 named pilots all upgraded to exactly 33 points. The extra objective adds an additional level of variety to the list, so you’re not just looking for the hardest-hitting combos, you’ve got to factor in getting the ball and scoring goals (or at least preventing the enemy from scoring). However, the Guild Ball does not suffer from lack of variety in tournament play. The captain makes such a big difference to the way a team plays that you’re going to be looking at, essentially, 17 different basic archetypes you’re likely to face. Once you get past the Obulus hype that seems to be going on in certain quarters of the US, there’s not many captains that aren’t competitively viable in Season 2. Looking at the data, we can see that all teams are fairly evenly balanced on a number of different measures. Butchers perhaps stand out as topping the table on a number of different criteria, but the difference between teams is very little, and all teams are capable of winning events. Fishermen seem a little low on this score, but I suspect that is because they are many player’s first team and not necessarily the team they stick with once they start to play the game at a higher level - not because they’re a bad guild, but because once you get to grips with the game you want to try out different things. Even if we feel we can rule out a few captain options for whatever reasons, I think you’re still looking at pushing fifteen different competitive team “archetypes” you’re likely to face. The competitive X-Wing meta would dream of having so many tier-1 lists.

It is also important to realise that balance in a wargame is an ever-changing thing. With each new release comes new balancing issues, and new opportunities. Just as X-Wing "fixes" things with new ships each wave, Guild Ball’s Season mechanic allows for new models, entire new factions as well as refinements of the game rules. With Season 3 around the corner, we have already heard some hints of changes to a number of existing players, helping to bring the oft-overlooked player back to the fore, and presumably to make the ever-presents less of an obvious choice. Not only that, but the Season release system means that you only need to pick up the latest version of the rules to be able to get going. There’s no having to dip back into a history of previous releases to understand all the rules, and new rules tend to get more streamlined rather than bloated as the game is expanded.

Game balance is a multi-faced thing. There is an inherent tightrope for any game designer to walk between variety and balance. Most games use a system of points to allow you the options to build your list for the game, but there is an inherent flaw with this mechanism, as there will always be certain models that are more powerful than their cost would indicate and using them. Further as the "meta" shifts, models will gain or lose effectiveness as their counters and targets wax and wane in their prevalence the gaming community. I love how Guild Ball scraps points values, and I love how having a set number of models makes alternating activations work better than nearly any other game I have encountered that uses that mechanic. But balance is always going to shift in a game, and game designers need to make sure they don’t write themselves into a hole and give themselves the opportunity to address issues and rebalance the game as it develops and grows.

A Recipe for a "Perfect" Game?

A bit of a presumptive title for this segment (although there is a question mark at the end), but if I’m to be serious about developing my Etherscope miniatures game, even just to a state for me to play myself, I need to start putting all this analysis together into some sort of plan for what I think makes a perfect game.

  • Objectives: I think with objectives there is a fine balance to be struck between variety of objective and simplicity of gameplay. My concept for my Etherscope game involves a gang of hackers going up against each other in a ‘Scope domain, trying track down or protect crucial datafiles. As all combat in the ‘Scope is non-lethal, being able to track down an enemy ganger’s physical location so you can exert some real-world pressure is also an important part of the game. Hence I will have two different types of objectives: grabbing datafiles and tagging enemy avatars when you force them to jack out of the scope. I also like the idea of secret objectives, in theory, so one of the things I want to playtest is making these objectives somehow secret from your opponent.
  • Turn Sequence: If the game warrants it, some form of alternating activations is the better way to go. Given that my Etherscope game is going to be a small-scale skirmish game, I think the scope for alternating activations is good, but the idea of having a set number of models like in Guild Ball sounds too much like an artificial straightjacket for the story behind the game. I really like X-Wing’s activation system, but to use it I think it will need some sort of equivalent of the dial system. I’m not sure if it will work, but I think a system where each model is given an order that either restricts what they can do, or gives them an advantage for doing some things would make this system workable. Similarly the split phases system that works so well in X-Wing wouldn’t work too well for a skirmish game where there is a greater variety of actions a model can take. Hence the activation order will need to allow some system of delaying or interrupting other models actions to allow the "better" models more ability to respond. I think I have mechanisms for how all this can work, but it’s perhaps a bit too complicated to go over briefly here, but I think I can make X-Wing style activation work for me.
  • Probability: When it comes to dice rolling, I’m a big fan of binary dice pools with variable target numbers. This provides lots of opportunities for different ways to modify the dice, and a nice bell curve of results. I also like using d8s rather than d6s – the maths of halves, quarters and eighths appeals to the computer scientist in me and adds a bit more variety to target numbers over a d6. I used them heavily for a roleplaying game I wrote a few years ago now Midgard: Viking Legends and have a bag of a hundred of them, so I wouldn’t be short for playtesting.
  • List Building: I think I want to follow guild ball’s example and try and get rid of points, but I need a bit more flexibility than Captains, Players and Mascots. Having one significant important model fits, and allows that big impact on gameplay variety by adding a single model. Therefore each gang will have a leader, and it’s the leader who will determine the size of gang and range of models that they can bring. I can then have a few different classes of model (I’m thinking Lieutenants, Specialists, Enforcers and Gremlins at the moment, but that may change) and they must build their team from a certain number of models of each type. I am going to have each model as a unique, named character (like Guild Ball) as I think that is one of the best ways of managing balance.
  • Unique Selling Point: If a game is going to be successful it needs to have something that makes it stand out from the crowd. So far All I have really done to make my game is pick and choose elements from other games that I think make for the best play. All I will be creating is an experiment in game mechanics. Guild Ball has its football theme, X-Wing is star wars and flying spaceships. Whilst I would say Etherscope has an incredibly rich and detailed setting background (well, I would, wouldn't I?), it needs something else to make it special. I have an idea for what this will be, but for now I will keep it to myself.
  • This has been a fairly long and rambling article. It's helped me get my head around a few things, and I think I might be able to put together a game I'd be happier with. There's probably much more I could have written, and as I put this together piecemeal over the summer holidays, and then finished it off in a trickle over the last few weeks its probably a bit disjointed, but I hope it provides some interesting insights and discussion points, if nothing else.