Baldur’s Gate : on UX and RPGs
How UX can make newcomers’ life easier with complex game systems.
Crowdfundings can do miracles. After almost two decades of silence they brought back to life the isometric Computer Role Playing Game — or CRPG. And with the recent release in early access of two ambitious new games, Baldur’s Gate 3 and Solasta, both reclaiming Baldur’s Gate legacy and Dungeons & Dragons rules, new and old players couldn’t be more exited.
But CRPGs have the reputation of not being easy to grasp for newcomers. It was the case in 1998 when the first Baldur’s Gate was released and it still is an issue today with the brand new third episode. Mostly for the same “inherent vices” caused by the foundations upon which those games are built on.
Today I’d like to revisit Baldur’s Gate and what we could learn from it in order to improve the onboarding of games with complex systems.
Why do I take a 1998 game as subject ? Well, Baldur’s Gate 1 & 2 epitomize the isometric CRPG. Many of the problems more recent games show can be traced back to them. And they still make experienced players rant — sometimes for several hours.
But first, some ground rules :
- This series of articles about Baldur’s Gate and the onboarding phases of CRPGs is not a critique of a 1998 game with the UX standards of 2020. Firstly, because I will deal with the 2013 remaster, the “Enhanced Edition”, which brought some usefull improvements in terms of ergonomics and accessibility. Secondly, I am mostly motivated to see how things could be improved to make the player’s life easier without “dumbing down” the complexity of the game.
- Most of the pain points depicted here were identified by compiling lots of players’ questions and comments on dedicated videos, streams and forums pages. Their importance is backed and ranked by UX principles and psychology knowledge.
- Keep in mind that some of them are not inherently bad if they contribute to the kind of experience the developpers had in mind. Especially if they are the results of an educated choice informed by tests, observations and data. We’ll see which ones fell into that category.
- Finally, as everything with UX, the options of improvements suggested here, if implemented, should always be part of a cycle of observations, tests and iterations. Sadly, I’m not a modder and my programming skills lie more on data analysis than game developpement.
The problem with Baldur’s Gate and (most) CRPGs
The classical CRPG is perhaps, with Grand Strategy games, one of the most difficult genre to get into. Baldur’s Gate and lots of the CRPGs inspired by Advanced Dungeons & Dragons (AD&D), or other tabletop RPG mechanics, usally suffer from two kinds of problems.
One lies in the rules they use to run the game : here, the fairly complex Second Edition of AD&D. The other is the way these are applied into a video game genre that strongly differs from the usual way people expect video games to work.
As the most common option tabletop RPGs use the result of dice throws and how it relates to some stats to determine character’s actions success. Your own motor skills are much less involved than your knowledge and ability to pick and favor, at the very begining, the stats you’ll need to maximize your odds.
For 30 years computer RPGs made use of that system. On the other hand, even if the game he is playing is not a “button smasher”, the player would assume that only his own motor perfomances make the difference. Not the random results of dice rolls. And how could he be blamed ? The whole history of video games has shaped his mental model that way. CRPGs were among the few exceptions.
Moreover you usually won’t acquire new skills in most classical RPGs. The character build you set is almost definitive. As you play your charcacter gets better at the things he does. But he’ll start with almost all the mechanics available — your success in performing them still depends on your initial stats and dice throws.
So you need to know all the mechanics up front. Therefore tutorials are harder to design, since you can’t rely on distributed learnings to lessen first timers’ cognitive load.
Hence games like Baldur’s Gate not only propose you stats and randomness when you think of gameplay and manual control. They also assume you know how to take care of yourself. Or that you are willing to read — a lot ! Either within the game (dialogues and other diegetic pieces of information) or outside of it (thick manuals, D&D books, websites…).
The original Baldur’s Gate was released by BioWare in late 1998 in the US and in 1999 elsewhere. Like 28 others CRPGs before it used both the rules and one the settings of AD&D. At that time the second edition was already a decade old and a success — though the « satanic panic » simmered down its audience a bit.
That didn’t prevent Baldur’s Gate 1 & 2 to sell millions of copies in the late 90’s and early 2000’s. Why ? Because for CRPGs fans it was the perfect product-market fit : a new take on D&D CPRGs with fights in real time.
But things changed in the mid-2000s. The game industry shifted from this rather restrictive system to something more open and forgiving. Tending to appeal more and more to console players, psychomotor skills were introduced to new licences, turning pure stats into skill trees and keeping the “choices with consequences” aspect as the key part of the “RPG experience”. The recent and very successful Mass Effects (BioWare) and The Witcher (CD Projekt) series embody that trend.
Now there’s a comeback to more classical CRPGs — and finally a third Baldur’s Gate, twenty years after the last one ! But the problems they faced, UX wise, remain. Newcomers hung on in 1998 because there was no alternative but to sink or swim.
Once comfy at the deep end the peak-end rule would kick in : players don’t recall their whole experience. They mostly remember how they felt at its best and last moments.
Applied to Baldur’s Gate that means that the players who overcame its steep learning curve enjoyed the time they invested in the game. But now players are used to smoother forms of onboarding or want to discover the game by themselves.
So if one either wants a game or product to be used as intended, or for the user to reach at least one moment of enjoyment, one would better propose a good onboarding.
So let’s see how an UX approach can help new players handle their first encounter with CRPGs.
On onboardings and why it matters
Onboardings can make or brake game success — or any complex digital product. During that critical early phase the player gets to feel if (s)he is going to be confortable and proficient with the gameplay.
The onboarding can be very, very, short. Especially with free to play games, where the sunk costs fallacy won’t keep players “motivated”. Or if refund policies end up within a couple of hours.
If one would venture on dedicated YouTube videos and forum pages, one would assume the first journey into BG is full of insecurities and misunderstandings for the options available at your very first steps are vast and, most likely, irreversible.
Or as someone casually explained to a first time player during a Twitch stream “the game just throws numbers at you and hopes you know what they mean” :
The joys of character creation
So let’s sum up some of the issues of the very and CRUCIAL first step of the game, the character creation phase :
- Lots of texts, jargon — including two very important and non explained acronyms that are D&D specific (THAC0, AC) — stats and bullet points for most of the 8 steps of the character creation.
- 7 races that make sense for most of people — they are Fantasy canons — but without no prior experience of the game you can’t forsee if their key stats will favor the type of gameplay you would expect.
- 49 individual class kits to pick from the 11 main classes that are thrown at you without much visible order or classification — strangely the manual does sort them nicely. This might be deceiving since some of the differences between them are not as trivial as they might appear — if you didn’t read and thought that Mage and Sorcerer play the same since both deal with magic, be ready for a rude awakening.
- 24 proficiency options at your disposal and only a few points to spend. Hopefully they are easy to understand.
- Though paralysis analysis might occur at every stage of the character creation, it is especially predominant when dispatching ability points after dice rolls. And with spellcasters too, when the players is asked to choose few among many spells options to start the game.
- Lastly, if you want to go one or several steps back to make a change you’ll still have to redo them entirely. There is no way, at the last step, to review one or several steps individually. It’s an all or nothing journey.
To top it all off, if your cognitive load wasn’t on fire already, there are some affordance issues : due to a lack of visual cues you may not realize you can uncheck some options the game has set up for you (abilities, spells…) and reasign them as you please.
And if you do take the time to explore and read everything carefully before selecting your first spells, you would probably experience what the Fitts and Hicks laws are about — with a hint of aversion loss too.
So, if we apply all of these insights into the actual player flow, pointing all the issues that may arise, we end up with a chart like this :
How can we make the game to do more for the user ?
Quite an overwhelming first experience isn’t it ? Yes. But is it (that) bad (of a) design or (done) by design ? Of course the developpers of BioWare and then Beamdog were not sloppy with UX — though that term might be a bit anachronistic for the former, the interest isn’t. They just simply have developed the game with a particular user in mind.
I don’t know if in 1998 BioWare used personae, but they built the game for someone already familiar with D&D rules, who fancies CRPGs and is not afraid to read.
And when Beamdog released the Enhanced Edition, they listened their fans feedback, focusing on gameplay fixtures and modding facilities — though the big picture they had in mind were whittled by unexpected difficulties. That’s great if you’ve already been playing the game for a decade. Not much if you’ve started your journey these last years.
The example of a real tabletop D&D session
Since Baldur’s Gate is a retranscription of the kind of adventures you can expect from a D&D campaign, an interesting solution would be to meet and see tabletop D&D players. And observe how the Dungeon Master and experienced players introduce and help the new ones.
With the rise in popularity of D&D related streams, you can find some of these directly on YouTube, like this one from Geek & Sundry who also hosts the much acclaimed Critical Role show :
Here Jason Charles Miller explains the key elements of the 5th edition of D&D (5D&D). But the core principles still apply to the 2AD&D.
He does an excellent job at making everything as clear as possible for each one of the new players :
- he presents races, classes, sub-classes and kits in a way the player can pick accordingly to the kind of character he wants to play
- he embodies alignments by comparing them to famous fictional characters
- he makes the dice rolls and ability points less puzzling by advising the player on what to prioritize for the class/gameplay he chose.
- he depicts clearly the bonus stats attached to a race or class with the use of contextual examples.
With that in mind, the goal is now to create something as cooperative as what Jason Charles Miller did, in a organic way without patronizing the player. That will be the purpose of the next article of this series about UX and the onboarding of CRPGs.
