Some of you may have been following a hashtag that appeared on Twitter recently as #GuildBallFieldTest.
The initiative is a way of engaging with the community to push Guild Ball forward.
One of the first things that the Field test is aiming to look at is:
How to ‘Widen the appeal to new players whilst retaining the Guild Ball identity and appeal for existing players’
Guild Ball co-creator Mat Hart has recently posted this on the Guild Ball Supporters group on FaceBook:
“Why a Field test?” I hear many people ask…
“Guild Ball is perfect!” I hear people saying…
“Fix all the other stuff and it’ll be sunlit uplands for everyone!”…
But for every statement of “the rules were never the problem” I can point to the same, if not more, number of “it’s a full-time job to be good at” or “there’s too much to learn for a beginner” or “I don’t enjoy getting beaten every single time I play”.
As I said in the Adepticant keynote the other week, the Guild Ball ruleset IS in great shape right now…it’s a really stable platform with a lot of ‘knowns’. This makes it ideal for a solid look at what the next stage in its life-cycle should be. For a game to continue to be successful, it needs more new players coming in than old players moving on. As the curve gets steeper, the worse this gets. So we must address this if we are to go forward successfully.
We all agree it’d be great to have a new starter set, but what rules do we put in them so it doesn’t go out of date? And more importantly, so new players can discover this wonderful game and the depths it has to offer and not be confused or ill-equipped.
It’d also be awesome to have loads of new models coming out but with each model added to the game, we exponentially increase the development and testing cost, deficiencies in either of these endeavors are cruelly, and quickly, exposed by the game’s great balance and expert users right now.
For us to look forward, we need to carve out design AND development space, we need to smooth the balance curve to give us room to take the game forward, and we need new players picking the game up and having a great time!
So let’s mess up some unicorns, shall we?
But that being said, and before we go any further…
I want to stress that Guild Ball IS, and will REMAIN, a competitive game!
It was designed to be a game that rewards player skill and experience. The more you put in, the more you should get out was a clear stated design premise from the beginning. This doesn’t change as part of this process.
But model balance and a clean ruleset (that needs no changes) isn’t the final finish line, there is a subtle overall ecosystem that also needs attention. Let me explain…
Let’s talk about player-skill and why this makes Guild Ball so appealing to a lot of people. It is well known that I spend a lot of hours a week playing a MOBA…just one in particular, even though there are other MOBAs out there. I found the reward for putting in hard yards, learning the champions, their abilities and threat ranges, how they combo with other champions, how the item upgrades work, what is a good item set for a champion but also taking into consideration the enemy team composition and the state of the game in progress, what time to go for objectives, when to turtle and when to push etc. etc.
Every year, the MOBA I play ends the current season and we go into a bit of a free-for-all where the developers (let’s call them Rito) try out some wacky ideas and new elements that they think can keep the gameplay feeling fresh for the next season. And I hate it! Every single time it happens.
It means that my hard-won knowledge of the game is eroded and I feel like I have to start again. I don’t want to learn new item builds, I like my old ones. I don’t like the jungle timers changing ‘cos my old routes won’t work anymore and they always got me my level 2 gank more often than not. And what the hell is this new thing in the river now?
But the truth is I don’t actually start from zero each time, it happens. My practiced (earned) mechanics and game-knowledge allow me to update my knowledge base and I can start seeing my decision making catching up to the new meta. After a little bit of time, it feels fresh and exciting AND the challenge of learning new stuff is simply fun to do!
My love for MOBAs has bled into my design philosophy. I was playing a lot of MOBAs when I first dreamed the idea of Guild Ball and some of you old guard knew this when you saw some of the champions I liked to play creeping into the design space of the Guild Ball players…Stave for example?
So we can look at how Rito thinks about their game and use similar lines of thinking for our game of Guild Ball. Let’s talk about player skill for a bit, and to do so I need to define some terms…
If we assume that player skill is made up from these key elements…there’s probably more but these feel like the key areas to think about:
· Knowledge – knowing what models can do, their set-plays, individual or teammate combos, DPR output…all parsed against INF allocation and current game-state
· Decision Making – picking the right option at the right time and executing it properly
· Predicting – Alternatively could be called Empathy, this is reading your opponent, their team, play-style, game-state and second-guessing what they are trying to do…and then trying to counter it.
Win probability can be calculated from a combination of comparing player-skill and the results of the dice in the game. As you all know, Guild Ball has a low-variance RNG curve with higher than average (for a wargame) predictability. The dice mechanic flattens the curve so much, especially when going for the lower Playbook results and you only need one or two results to get what you want (which is probably why whiffing feels so much worse).
And so the higher your player-skill, the less you want dice results to be the deciding factor…this is why you will see comments such as “GB is perfect as it is, it doesn’t need any changes!” which I suspect translates to “I have worked hard to get my Knowledge, Decision Making and Predicting skills to a high level, please don’t undermine this by making the random factor more important!”. One of our development team plays to a very high level and once told me that he’d prefer it if Guild Ball didn’t use any dice at all…I won’t name names so don’t even ask me!
But then we can look at Godtear, a game in which we poured a lot of the lessons learned over the years. The dice are much closer to a traditional minis game curve but It still has hard tactical and strategic decisions that if you are better at making then over time you will win more games.
So to draw this line of thought to a conclusion…as player-skill increases, so does the gap between a new and experienced player. The bigger the gap, the longer (and harder) the journey is for the new player to get to a decent playing level. And whether we like it or not, that turns a lot of people off of a game because no-one likes to get smashed every single time they play with next to no hope of winning!
I’ll explain…
Let’s talk about win probability in detail. If you play regularly, you will probably be able to sort your regular opponents into categories that you either…
· …can’t beat
· …beat 1/10
· …beat around 5/10
· …beat 9/10
· …always beat
Right now, I’m not convinced there is enough impact from the variable elements in Guild Ball to allow the 1/10 or 9/10 win-rates to be statistically probable. This is why some players will argue over millimeters!
Against a given opponent, you will either win every time, lose every time or if you have an equal skill then it’ll be 50:50 between you both. In the 50:50 bracket, the determining factor in a single game of Guild Ball is usually a single dice roll or a single decision, which leads to a punishingly low level of tolerance inherent in the game. Therefore, if the deciding factor is not making a mistake, this leads to a low-risk style of play being more successful and being adopted across the competitive community. This then becomes the de-facto way to play the game because that’s what all the key influencers are talking about. So to my eyes at least, players currently seem to play to “not lose”, rather than “playing to win”.
So going back to the three key elements of player-skill for a minute. Ideally, I think the game would require an even and equal balance of all these items. Let’s steal a term from fighting games and call this the ‘neutral’ position of the game.
It must be noted that Neutral is a fluid position, it changes constantly as models get added and core rules get polished. Look at it in terms of relative value when calculating player-skill, when the Knowledge aspect continually grows and therefore the gap between new and experienced players also gets ever wider, this then forces the Decision Making to get smaller (in comparison to Knowledge) which in turn leads to less reliable Predicting so this aspect reduces too. In this form, the game meta then starts to hinge around key set-plays and patterns, which further increases the Knowledge requirement and we can easily see a vicious circle occurring.
Right now, think about it…how many pivotal moments in games are dominated by a fully-loaded captain or power-model? As opposed to good positioning of a team and solid all-round play? And then when you come up against this, if you know the set-play then you might be able to counter it, but if you don’t then you simply lose the game because a single mistake is all that’s needed.
Don’t get me wrong, I think it is fine for skill gaps to exist, this is why the game is so addictive as I mentioned briefly before. Every game gives you a tiny bit of experience that helps you get better and so your player-skill grows with practice, playing and studying the game. I think it is entirely reasonable for the more skilled player to win more often, but not to a 100% win rate level. The game inherently needs a more volatile game-state to allow a weaker player to compensate somehow for their (relative) weakness and steal a game or two. I’m not saying drive this down to 50:50 regardless of skill gap, that would be too random, but the 1/10 and 9/10 scenarios need to be brought into play, ideally, they would be closer to 7-8/10 and 2-3/10 correspondingly.
Reference to Magic the Gathering, arguably the world’s most successful competitive game in our hobby’s history…ever! Pro versus amateur, win rates are closer to 60:40 in favor of the pro. Pros will know they will win more games than not, but they go into each game with the excitement and tension of not knowing for certain…and we see the MtG circuit growing and growing after all this time. It is interesting to think about…
I think I, therefore, want to make the difference in skill between players a little more ‘ephemeral’ and a bit harder to define clearly. But to do this, we need a little more variance in the game.
But this is where it can all go wrong if we’re not careful. Undefined variance is not helpful, we need to look at the player skill attributes and deliberately source or create our variance from specific aspects of the game ecosystem, in order to bring the game closer to neutral.
Having looked at the game for some time, I think we can initially attack this from several different angles in pursuit of the neutral balance:
· Shrink the knowledge gap by lowering the volume of data points that need to be learned.
· Emphasise the execution of play by making board control and positioning more important.
· De-emphasise set-plays and patterns (which like Fools Mate in chess, will not work if you know about them but will work if you don’t).
· Emphasise the need for better Predicting skills to win.
Further, to provide a cushion and to encourage more open play, we need to find ways to increase the tolerance of mistakes and try to reduce games being decided on a single decision or dice roll if we can.
So despite the popular view right now is that we don’t need a Field Test, I believe there’s actually a lot of tinkering we must do for the health of the game state right now!
The knowledge aspect has got way too big and it is crushing all other aspects of the game. Much like Rito’s Close Season, the Guild Ball Field Test is a process for us to come together as a community and talk about the game, bounce some ideas around and see what seems fun and what doesn’t. In the background, I’ll be parsing the feedback/data and working out how each idea affects the game state itself and slowly forming a new season’s ruleset to take the game forward.
I’ll be mostly working on Guild Ball over the weekend (gotta find those dev hours from somewhere!) so I will be around and about. Then next week I’m expecting to publish the actual proper guidelines for the Field Test via the official SFG channels, this will set out how its all going to work and what we need (hope) from all of the community.
Phew! That was long! But hopefully gives you all some insight into the thinking behind what we’re looking at during this Field Test. I don’t really do FB anymore so any questions or useful comments, then find me on Twitter (@c4rp3r) instead of here, but please be nice, to me and to each other, we all want the same thing and there’s enough nastiness in the world already…
Exciting times to come!