How I playtest

You’ll hear the advice everywhere, “playtest your game as much as possible” but what does that actually involve? There’s pretty standard advice out there, “don’t just play with your friends”, “get as many people to test your game as possible”, “playtest with the sort of people who’ll play the game”, “ask the right questions”. All of this advice focuses around playtesting with other people. In my work as a designer, playtester, and developer I spend a lot of time playing games I’m working on, on my own. Yes some of that is solo, but only so much, the rest I’m playing the game as multiple players. I’m lucky to have some close friends who live nearby who are happy to playtest games, but I can’t take over every game night with prototypes and I have no regular access to playtesting specific groups. One friend in particular, David Ellis, is incredibly good for testing games with, and we have a great working relationship, which is pretty much co-design on some projects, but a lot of the testing falls on just me. So how do I do that? How do I make the most of that time? What about that method has helped me develop my own skills? You’re about to find out!

This is not the perfect solution

The methods I will go on to detail will not work with every game. Some games lend themselves to this method of testing better than others. Some mechanisms are impossible to replicate on your own, some require a lot of roleplaying (which I will go on to explain). If you’re reading this you can probably think of what games won’t work, social deduction, hidden roles, real time etc. However it does work for a lot of games you might not think it does. My personal tastes are for medium to heavy euro style games, that’s not all encompassing but it’ll do. These are also the games I’m often asked to work on. That’s about me knowing where my strengths are and getting to know the right people. Anyways, if you think your game design might benefit from this style of testing let’s get on with the how.

Examples

It’s not appropriate for me to quote from projects I’ve been involved in but at times I feel the need to provide examples. I’m going to use one of my own games; The Seven Dwarves, in these examples. It was my first design, and for a number of reasons I don’t think it’ll ever be published. It therefore will serve us well for an example piece of work. It’s a dice placement euro game for 1-7 players, core is Kingsburg like, for those of you familiar with that.

Initial stages – Just play it

Play the game. I’ll often start at 2 players and just play the game a few times to learn it, see what it’s about and get a feel for it. There’s no right number here, sometimes 1 play will be enough sometimes it’ll take a few and at different player counts. The aim to get a good overall understanding of what the game is about, how it plays, and more importantly how to win. If you have the option to play with other folks at this stage do so, but only if the game is playable. If it’s an early prototype don’t inflict it on others yet, but sometimes other people may spot things you didn’t notice. Across these plays just jot down some ‘big picture’ notes, if you think anything like me you’ll need to actively ignore going into too much detail.

So in The Seven Dwarves, I would have notes like this:

7 different player powers, 7 different powers ups, so 49 combinations

1 way to score points but in 3 different types, some in-game, some end-game

Track mechanism that provides upgrades, player order, and end game trigger

Some variable set-up

Dice placement, but there is some mitigation, no dice are ever wasted

7 resource types of varying values

Build the players

The next stage is outlining the strategies you want to test and how you are going to assign them to your dummy players. There are many different ways of doing this and it depends much on the design itself. 

In The Seven Dwarves there are player powers, and upgrades so all 49 combinations of those needed testing. Each player has a colour and each upgrade gets a number. From these we get RED 1-7 etc. This list produces 49 players we need to test, the number of tests you need depends on player count in each game.

There are also 3 different types of cards we will be building to score points, so there is another 7 players.

7?! Yes seven. Here is a key part of the process. In our example there are 3 types of cards we will score points from, so that gives us 3 strategies, I call these the “extremes”. The players will do all they possibly can to solely focus on one of those card types. Then there are 3 “50/50” strategies where the player now has 2 things to choose from. Finally an “all balanced” strategy where all the card types are considered equal.

So we now have our players, all 343 of them! Your game may be a lot simpler, it might not have player powers, or ways to upgrade or engine build, there may only be 1 way to score points. Or it may be more complex and less easy to work out. If it’s simpler, great! If it’s more complex we may need to figure out another way to outline our players and that’s why you need to play the game enough first. Write a list of strategies, it doesn’t have to be all-encompassing, but write a decent list of what players might have as a strategy in your game. Think outside the box too. We all have that one gaming friend who manages to come up with something really whacky, emulate or talk to them. One of my known traits when learning games is to ignore something, I’m going to digress briefly on that.

Ignore one thing 

Here’s how I’ll often play games for the first time, outside of design work. I’ll listen to the teach, or read the rulebook, or watch a playthrough and pick out one element of the game to ignore. Because of my personal preference this will often be combat, or bidding, but could be a certain worker spot or action, or a mini-game, or a phase. I will then play the game ignoring that thing as much as possible. Sometimes it quickly becomes apparent that I picked the wrong thing and I have to buckle and do it. This is not failure, I have identified that that thing is essential to gameplay, this discovery is really important. A lot of the time it will provide a huge insight into how that thing and all other things play out within a game. Sometimes it will make no difference whatsoever, that is a huge red flag for me. If it made no difference at all to the game experience why on earth is it there at all?!

The Seven Dwarves features a track that represents each Dwarf jostling to try and be Snow White’s favourite. It’s a great example of ignoring something. If ignored it, I’d never get the upgraded player powers, or get an extra dice, or score bonus points at the end of the game. On the flip side I would always be first player and would have 1 less resource to worry about so can focus more of my attention elsewhere. By playing the game ignoring this track I can clearly see how important it is from almost every angle. I find this a lot of fun and very insightful.

WARNING: If you are going to do this in a playtest someone else is running tell them first! It will often look like you are trying to break their game intentionally and that is not well received. Until people understand what you are doing, be polite so they don’t go onto the back foot.

Play the “extremes”

We should now have a list of strategies we categorise as the “extremes”. Ideally these would be within your player count. So if you have a 4 player game with 3 extreme strategies, we are good to go. As soon as the number of “extremes” exceeds your player count you are in for a much longer process. I’ll leave you to work that out, but it should be easy enough.

Play a game in which all the “extremes” play against one another. Here’s a top tip for doing this. Post-it notes. Write the strategy on a post-it note and put that in that player’s area. It may be that players have boards, or hands, or resources, or whatever, put the note with them. It will remind you how that player wants to play the game.

We identified our extremes as the 3 card types which are; armour, weapons, and jewellery. So we need a 3 player game where each player follows one of those extreme strategies, simple.

Record the results and look at the trends. Don’t forget to write about player experience as well as the scores. It is as important to find out if a player is having more fun than the others as well as if they are winning. Even more important, look at how the two relate to one another. What you do with this information, how you look at it, and what you go on to change or work on depends on your design brief.

What do I want my game to be?

I always write myself a design brief for my designs. When working for someone else I get them to tell me what they want. Does the game need to be perfectly balanced? Should player powers be exciting and varied? Should an all balanced strategy be better, worse, or equal to an extreme one? How much strategy do you want players to have? How adaptable can strategies be? How much tactics is there as well as or instead of strategy? Do I want the best player to win? How much scoring is open? How much scoring happens during the game and how much after? What should a winning score be? How much of a final score should each thing be? I could go on and on about this but you need this information in advance to know what you are testing for and why.

Play the “50/50s”

We have looked at the “extremes”, analysed the relationship between them, and probably made some changes to the game. Now we move onto the “50/50s”. This could not be any easier. Remember those post-it notes you put in a player area? Move them half a place around the table so each one sits between 2 players, simple. Now play some games where each player balances 2 extreme strategies, perhaps here you’ll need to add some other type of note to help make decisions, maybe one player wants in-game points and another end game points, for example. This will start to expand your testing into more varied strategies. This phase will take longer as there is more variance. Study these results (don’t forget player experience is of equal if not higher value than the raw data) both against themselves and the extreme tests. Go back to your brief and work out what you want to do with these results. Don’t forget that small changes can have huge knock-ons. “The player with the most cows always wins so lets just make cows worth less VP”. “Oh crap, now no-one wants to buy cows and the supply chain has broken down completely”. If in doubt go back to your brief and your theme. These will inform your decision making.

Play the “all balanced”

If possible add this player into your “extremes” playtest as an additional player, once you feel you’ve ironed out a lot of the wrinkles. This is the best test for this. Now more than ever you need to have a clear idea in mind of what you are trying to achieve in terms of balance. I repeat; player experience is as important, if not more so, than the numbers. 

I wanted balance in The Seven Dwarves. You can imagine the time I spent trying all 49 player setups and that’s something specific to this game as it had the 7 player powers and 7 player upgrades. The strategy testing was as important. Specialising in only 1 card type is suboptimal as you end up with too much waste, perfectly balanced is also suboptimal as you’ll be chasing too much. So here’s the key point for me in this design, not everything is balanced but there is no right or wrong answer. Tactics (the decisions you make during the game) outway any strategy, the player who plays the best will win most of the time. However a real strength is that players have a lot of fun having done one thing better than the other, “you may have won but I got way more jewellery than you and that’s cool!”. In this way the game filled the brief perfectly.

The hard work

Be prepared to play the game a lot!

Remember we identified 343 players that needed testing, and yes I tested them all!

Record every playtest and always apply your brief to the results. Make changes when you need to and want to and keep testing. Sometimes you’ll need to go back and test using a player group again, sometimes you can move on and not worry about it. 

When you are ready or able to playtest with other people let them play however they want to. You’ve tried every option you can think of already, so give them freedom to do what they want. This testing method should get your game into a really polished state, playing with other real people will get you all sorts of other information.

Yes I spend a lot of time playing games like this but it appears to be working and unsurprisingly has led into designing and developing solo modes for games, which I will cover in the next blog.

Thanks for taking the time to read this and I hope it’s helped in some way. Check out all I’m doing board game wise at www.facebook.com/DavidDigbyBoardGames