Recently I went all nerdrage on an Australian tournament for having bad missions. I'm not going to get dragged into the 'point' of that specific tournament but a tournament is by definition is "a series of contests between competitors for an overall prize." We all agree each contest should be as fair as possible and not advantage either side to the best of the creator's abilities? Good. Let's then look at how we can do this because world-wide there are huge issues with making missions which are fair to both sides (even Games Workshop is pretty bad at it; hi planetstrike).
When designing missions for tournaments you need to keep a lot of things in mind. The over-arching goal we've just touched upon which is obviously making sure neither side has a significant benefit. The problem then becomes the hundreds of variables you have to include. There are something like 15 different armies in 40k and whilst some of them are really hard to make competitive lists for, you can't just exclude them. People are still going to play them and disadvantaging them further isn't giving you a proper tournament. You then have to consider the basic 40k rules (currently 5th edition) or whatever other system you are designing the missions for. Do. Not. Stray. From. These. Tiny deviations are going to be necessary, you are making something new afterall, but you must ensure you stick as close to the core ruleset as possible. Make sure your missions emphasise what the rulebook is emphasising and not making up whole new systems. This is what a lot of missions in tournaments don't do and you end up with some crazy stuff that reeks of older editions or benefits certain armies or army types (i.e. 4th edition 40k, advantage for foot lists).
Before we get started on mission design there is one more important thing you must consider: randomness. Randomness is bad. The game has built-in randomness to it (D6 system, random game length, etc.) and doesn't need TO's adding in more. We are all after competing and nothing sucks worse than a random dice roll you have no control over swinging the game. Everything else, whilst based on dice, you and your opponent can stop. You also know random game length kicks in at Turn 5 so you must prepare for the game to end then whilst also going on, etc.
Deployment
With these broad strokes in mind let's break down the process of mission design. The first step is obviously deployment and I believe this is one of the major reasons why a lot of tournaments try to make new missions. The rulebook only gives us three deployment types which can apparently get boring to people. Making new ones is therefore an attractive prospect to potential TOs but unfortunately it is very difficult to do well. Let's look at some common features of the three deployments of Dawn of War, Pitched Battle and Spearhead.
All of them have a set minimum distance to your opponent and this works out between 18" and 24". This is where a lot of 40k gaming eventually happens and is commonly referred to as the 'midfield.' When designing a deployment for 40k you're going to have to include this midfield corridor between the two opposing armies but you have to manage where that corridor is to balance the gain shooting armies have. For example, if you had a 24" corridor running vertically down the centre of the board with the deployment zones being the short table edges, shooting armies are going to find it a lot easier to not only out range your opponent but also block them in midfield as there is less flanking room. This is why Spearhead uses corners. It clearly gives an advantage to shooting armies (corner to corner distance being 86-87") but with the long table edge being the reserve line and a great ability to flank, other armies are not up shit creek. At the same time Dawn of War can advantage more aggressive armies as one can advance further upfield before a shooting opponent can bring their full firepower to bear. These differences however are minor and when made-up across a tournament using equal number of deployments, aren't going to significantly alter the standings (and singular games with different deployments don't hand out such major advantages which means you can get knocked out based purely on the deployment type).
To therefore create a new deployment then you must consider what the current deployments are doing and attempt to replicate them. The obvious goal here is to not advantage any army with a single deployment or create three new deployments which mildly benefit shooting armies, are neutral and mildly benefit aggressive armies giving you an even spread across three games. You also have to consider what special rules are being enforced. In 5th edition everything is allowed in each mission (deep-strike, scouts, infiltrate, reserves, random game length, etc.) so deviating from that is a bad idea. However, where do reserves come in and how does this affect the set up? Are the Night Fighting rules enforced and for how long?
Winning Conditions
Once you have deployment set out you then need to look at winning conditions. I'm a big fan of the tiered system, especially in tournaments. Why? It absolves draws and levels the playing field a lot more, particularly with Kill Points. If you army is poor at a particular mission compared to the army across it (i.e. a foot list against a MSU mech list in an objective based game) you are secure in the knowledge you can tie on the objectives and push the winning condition 'down' the tiers. I've explained this more fully here. What's great about this system as well, is with each new deployment one can create, you can get at least three more missions out of as you rotate each winning condition through it.
With this in mind I would keep the tiered system but if you don't, you need a way to either address draws in your tournament draw or in how your mission is setup. From there you need to ensure your victory conditions meld well with the rules of the system. In 5th edition there is extreme emphasis placed on troops, midfield and mobility. You don't want to stray too far from these and whether it's something as simple as getting into your opponent's deployment zone, make sure it fits within the concept of 5th edition as much as possible. You must also make sure these winning conditions aren't hugely different across armies. Whilst the tiered system if used, can minimise the advantage any single winning condition gives an opponent, you don't want that advantage to be hugely magnified. For example, killing an opponents HQ hugely benefits an opponent who can hide/forego/has a cheap HQ such as an Imperial Guard Commander compared to an army with a large/expensive/important HQ such as a Tyranid Hive Tyrant. There are very different degrees in how each HQ affects their army and how easy they are to kill. The Tyrant may be much tougher but is a lot harder to hide and it's a lot more important to the Tyranid army that he affects the tabletop. Equalling rewarding someone for killing a Hive Tyrant compared to a IG Commander is folly.
Play-testing
No matter how sound or awesome your theory and logic may be, make sure every single mission is playtested over and over. You'd want hundreds if not thousands of playtests with these missions to make sure there isn't some quirky matchup you haven't thought of that has some huge disparity. Grab your other organisers and anyone who is interested in playing at the tournament and get them to test them as well. Make sure you use lots of different armies and army types to ensure the missions work for everyone and don't be afraid to go back to the drawing board or make changes as necessary.
End Result
What you should get in the end is a well-balanced mission which doesn't advantage any single type of army or codex with winning conditions/secondaries which are generally equally achievable for any army. It's going to be impossible to design missions which are equal across every army and list arch-type but as long as these are balanced across all the missions and not so extreme that any one mission is a huge advantage/disadvantage depending upon match-up, you'll have created a balanced tournament from a mission standpoint.
