|Posted by Rowan Powell on February 18, 2019 at 3:40 PM||comments (0)|
So this particular bugbear has struck me again as of late, having had the time again to go out and experience more new things, which of course requires a degree of learning. I love learning, I'll often seek something out purely because it's novel and contains the potential for me to be in an environment where I'm learning and improving. That being said, there are poor ways to learn and directly - poor ways to teach.
Show, don't tell
While this is often true, naturally any rule has it's exceptions. However it's such a crutch to fall back into the easy way of teaching which is to simply dump the tutee with information. Particularly common in games, the endless deluge of dialogs to be clicked through wading through the assumption of absorbtion without any actual context or worse, programming or step by step tutorials that simply list a collection of tasks to be performed with the goal in mind, but without transferring any capacity for it to be repeated independently from the tutorial.
This can also be extremely true of narrative, to call out one bad example I recently encountered; Story of my Uncle would frequntly lock the player (Who typically has high movement with multiple options for traversing the environment) into very slow or static movement so that they could convey a cutscene-style dialog. I dropped the game on the spot. While a good story can add a huge amount to a game it should come from experiencing the world as you move through it. This is the reason I hold portal 1 as an extremely well delivered puzzle game with a story attached and portal 2 a decent story with a puzzle game attached. The priorities were clear as you move through the world, what we made is more important than you are. Your character has agency and drive and freedom. You do not.
Why why why
Another disturbingly common problem I run into is things being taught because, just do this step then this one then this. Ta da! You did the Thing. You might even be able to replicate that set of steps or with a little bit of thinking and domain knowledge perform some variants upon it. But they don't teach you why you do it that way or how to think through the world space in a way to achieve such conclusions on your own. This often leads to horribly stuck players, users or designers particularly because they have no frame of reference of how or where to go.
A learning curve must be a curve
The big, stompy elephant in the corner of the room. A frequent frequent failing of domain experts is not just to be incapable of un-abstracting their domain knowledge, but in fact forgetting such a curve of abstraction exists. Given tasks, obstactles and opponents that simply are too much of a gap in execution or lateral thinking to provide a decent way to progress in their experience. Machine learning, videogames and common sense all spout the same mantra - that you should take on something around your level. But clueless beginners and misguided experts often litter the road to mastery with massive potholes that slow down the process massively.
Learning is a skill, learning how to learn is another skill entirely and learning how to teach is one that extremely few have mastered.
|Posted by Rowan Powell on December 30, 2017 at 9:35 AM||comments (0)|
Main menus were a staple of games when I was growing up and in many cases still are, at least on the PC platform. But there's a trend that swept the mobile game up by storm that I failed to notice and in hindsight, may have contributed pretty heavily to a lack of popularity for my mobile titles. Main menus are dead.
As arcades once did, the new generation of pick up and play gaming has an emphais of getting players in and getting them hooked as soon as possible. In an environment where players aren't already invested (i.e. haven't put a reasonable chunk of change up front to gain access to the game) they're very sensitive to any dropped beats in reward and stimulation until they're hooked and invested in progressing further. Main menus fall straight into this catagory by requiring the player to spend time navigating to the gameplay and skipping past lots of options they simply don't care about at this point.
Pretty much every game I've ever released has followed the formula of;
And while that's okay, it does tend to slow down how the player engages with the game (I've used colours to indicate this). Morden games have started to strip out some of these components or merge them together. Games such as portal received high praise for essentially hiding the tutorial in gameplay (A feat Mario Bros achieved a long time ago and other developers seem to have trouble replicating), titles such as Ultimate Chicken Horse subvert the menu structure by avoiding having too much of a break in the gameplay, you can run and jump around the menu screen and in fact you have to in order to be able to play your new levels.
A game I played recently that particuarly gave me a kick in the ass about this new trend is Everwing, a mobile/facebook shmup that took my friend group by storm. Playing around in the game for a little bit made one thing very obvious - the game wanted you to spend as little time as possible doing anything other than playing, even though the main hooks of the game are actually about the leaderboard and grinding for new dragons. In particular the game is in the menu or 'paused' state with the gameplay essentially loaded in the background, you can just tap once at any time to instantly be controlling your character again and starting a new run.
This mechanic in particular I wanted to emulate in my new game, I'm skipping past all the usual menus I add, making the gameplay appear in the background right as you start and getting the player into the game as soon as possible.
|Posted by Rowan Powell on October 16, 2016 at 4:30 PM||comments (0)|
When working their way through a game, the player generally wants some form of payoff for their time and effort investment which can come in a range of forms depending on the game and the player;
Axe Of Kings primarily draws from Novelty, Options and to a small extent Resources. This is because the core engagement of Axe Of Kings is Mastery of the system - knowing how to beat the monsters, maximise the effectiveness of the abilities and 'beating' the game. So the game needs plenty of fresh concepts to master fighting against (Enemy mechanics) and using (Items, Abilities) and constraints to consider (Debuffs, terrain, resources).
The player also wants the reward to be roughly proportional to the time or effort invested into getting the reward (The classic example being getting common loot from a raid is rather poor design), but at the same time is trying to maximise reward for investment and will attempt to identify which elements of your game have the highest ratio. One issue with this, as roguelikes and MMOs have shown repeatedly, is that players will do content they really don't enjoy just because they have the highest ratio. So the trick is having the fun content being the rewarding content, or reducing the 'price' (Time/effort) for the less fun parts.
The other issue I've been working on is the reward structure overall; having a range of different rewards, sizes, durations and so on. Let's take a look at the game's reward curve from the start of development, just floors of the dungeon and enemies to kill with no particular mechanics to each enemy.
A standard engagement curve looks a bit like the image below, with cycling intensity building to a peak and then tailing off.
Whereas the curve for AxeOfKings near the start of development looked a lot more like this;
As you can see, it's fairly monotonous and after only a handful of enemies the game will get pretty dull. And so it was, players got bored very quickly and stopped playing, often before clearing a single floor. The issue is once they've seen the full set of rewards, there's not much to work towards and each reward is less meaningful. So let's take a look at what the player's theoretical reward curve looks like with varied enemy health, gem rewards based on that health and the ability to complete the dungeon.
Better, but not great. Players would now sit down and play maybe a floor (Lack of good introduction to the game contributed heavily to this issue), which is no where near enough to make any of the later-stage content worth building - if they're never going to get there. To help resolve this a bit better, I sat down and thought about the immediate, short, long and overall goals the player would have and broke those down into how they applied to my game.
* immediate - dealing with an individual monster or clearing a room
Anywhere that the above reward structure didn't feel varied or was simply lacking, I designed new elements to keep the player engaged at these frequency intervals. To compliment this, the reward structure;
Now the curve looks a bit more like this;
Much much better! There's a lot of content now to be egaged with and a good mix of frequency and intensity and the way the 'alerted' mechanic works means that the player gets frequent rises and dips of challenge, which keep adding up untill ~50% of the monsters are engaged and then it starts to taper off and monsters are cleaned up and the loot that couldn't be delt with during the fight can grab the player's attention again, but in a relaxed setting.
Notice how the gameplay doesn't just feed into progressing through the dungeon (Healing, improving stats and such) but also tries to focus on the two key draws of AxeOfKings - Mastery and Improvement. The game doesn't give much mastery reward if all the content is obvious from the get go, so having lots of novelty is important - new enemy mechanics, new objects and level structure are vital for the player to feel like they have something left to 'beat'. As for improvement, the game actually doesn't really let the player improve their stats (Which is very unusual for a roguelike!) but instead gives them strength through the gameplay options and the knowledge they get about enemies makes them easier to defeat (Knowledge is power!). Having items that are all consumable allows the player to keep building up the power of their character and having that extra strength at a key moment, but also the AbilityLevel system allows for a more long term feeling of progression.
Reward structure and pacing is absolutely vital for keeping players interested in playing the game to get to the next element of the game, but it's really worth understanding that rewards are not just in-game rewards of numerical changes. Particle effects, story, exploration, domination of an opponent, the feeling of elitism of achievement are all other valid approaches to rewarding a player. You need to sit down and understand what draws the player to your game and what the core desires your game feeds into.
|Posted by Rowan Powell on October 10, 2016 at 2:40 PM||comments (0)|
Sun & Sky was my first experience making a mobile game, outside of small prototypes and tech demos and as such I made quite a lot of mistakes, which is exactly what I wanted/expected!
The first thing I had to deal with what making marketing material for the Play Store. As much as I have confidence in my art and design, marketing material is something far outside my usual remit, so not sure I did a great job with the work I did for that, though it seems passable enough to get people from the app page to the download button.
Big mistake #1
I tested the app out quite a lot on my personal phone before releasing it, but I didn't think to expect any major differences on other phones! The resolution differences meant that the UI scaling was unplayable and turned off many users after politely trying to squint at the UI for the first few levels...
The other issue this created was that the intro to the game was horribly low resolution, which made the game look really low quality!
The solution turned out to be really simple - just a changed setting in the publishing window!
Big mistake #2
I did absolutely no marketing or promotion about the game, a common problem for developers - we get focused on the game and not on the business! Unfortunately this obviously meant I had very low downloads (18 as of this post, most of which were me encouraging people to download it face to face!), but this was very easily preventable. I could have;
I got a lot of feedback from people early on about difficulties understanding the controls, the story and watching where they stopped playing. This was really valuable and I took advantage of it as much as I could. I also had really good workflows and made a huge amount of progress on the game in a short amount of time and it will be extremely easy for me to add more content down the line. I feel like a built a pretty solid and enjoyable game, it lacks in polish though and a few cases of lack of effort on my part really hurt it's success.
|Posted by Rowan Powell on October 8, 2016 at 2:35 PM||comments (0)|
During early playtesting for Axe Of Kings, I was looking out for how the players engaged with the various elements of the gameplay and given that I had revised the level layout before (As seen in an earlier blog post) I thought the maps were in a good spot and I could focus on the other parts of the game during testing. As it turns out, this was not the case.
The levels were now much more compact and meaningful to navigate given the pace of the turns and interacting with the movement mechanics of enemies was much more meaningful due to the constrained movement, but it was the same set of interactions over and over again, even worse was that it was now hard to work out where you were in the dungeon!
So the dungeon generation needed to be revised again;
The pace of the dungeon is generally the same, though a little faster in open areas and slower in cramped areas, the interactions tend to follow the same pace, though a little faster overall due to some new additions in map objects that spawn. The number of novel situations is vastly improved and the terrain now plays a much bigger part in how the player chooses where to go. The solution I came up with are very simple and solve the main problems I was having, though I think it can be improved further, fortunately this is very easy for me to do with this new system as I can just add new room structures.
The game now generates some slightly different room structures but also will merge open or dougnut rooms together to further mix up the structure of the levels.
I've also added a few extra objects such as locked chests and floor traps that turn out to be useful for orientating the player in the level as well as mixing up the pace of the game a little. You can see all of the new structures in the image below and how each room now tends to have something fairly identifying.
If all else fails, the game now has a map in the menu, with icons to show which rooms contain at least one of the main party members.
|Posted by Rowan Powell on September 15, 2016 at 5:30 AM||comments (0)|
There's an interesting psychology in how players approach certain genres of game due to their core motivations for playing those games. Fast paced cathartic shooters don't tend to invite much of a story (Though some do try to offer something up) as the player isn't looking for a narrative, just a setting to enjoy the mechanics and power fantasy. Strategy games demand an explanation of the setting, but once it's established often leave the player and the mechanics to fill in what is 'happening', calculated destruction or tense wars of attrition are often decided by the player's skill than a predefined narrative. A dungeoncrawler or other rpg often provide a running context and *other people's* stories, but leave who you are fairly open to interpretation, which I feel fits perfectly into why I play these games - I want to explore being someone else and defining a character.
Notably, some games go about this from unusual angles that don't fit my 'theory'. The main example that comes to mind is Elder Scrolls: Skyrim. At first look, Skyrim has all the mechanics that screams 'RPG' and the self-defining blank slate that I love in RPGs. But playing through the game didn't give me that feeling AT ALL, which in hindsight is quite revealing of how the game is presented. The game opens with you being neck deep in someone else's story - you were adventuring somewhere else and have returned home, captured and now facing execution. Skyrim is an RPG no doubt, but you are playing *someone else's* role and given that most RPGs only tend to tell a story explicitly if they are not the player Skyrim states clearly to me that the game is not about me, quite opposite from the RPGs I normally play.
Thinking about this can be useful when deciding how much story you need to put into your game, though notable examples such as Dark Souls and Guild Wars 2 offer huge wealth of lore material around the game for those it interests, and many people clearly are. But I've actually found it's even more useful to use the reverse conclusion: rather than 'X genre needs Y amount of story to satisfy players', I've listened to how much story the players are expecting or wanting from my game and that has been helping me understand the implicit feeling and genre of my mechanics!
When I was working on Sun & Sky, I threw together a rough intro to the game and focused on the mechanics primarily and one of the comments I got back is 'I don't know why I am doing this/fighting this' and it made me pause. My game felt like a puzzle game. There were no long-term strategic choices to be made, the tactics were minimal and all the mechanics pointed towards 'put the pieces in the right place', which if my game had been more abstract may well have not needed any story at all, but I had chosen to represent the puzzles and zones with 'turrets', 'fungal' and 'planets'. As it turns out the game therefore felt that it should fit into the strategy genre, at least to some extent. An opening cinematic of sorts seemed to settle concerns about motivation (Fitting what I theorise about strategy games! If they had wanted more development of the story I would have rethought this again as clearly the players would be expecting to journey through the narrative of the campaign to reclaim the galaxy). A player also commented about wanting longer term choices and unlock options, again not the usual expectation of a puzzle or tactics game where the resources reset for each level as mine does - but perfectly fitting for a strategy game!
I have resolved this conflict between mechanical-genre and feel-genre somewhat with the addition of the intro and outro sequences and I'm planning to go back and add some long-term choices (Perhaps between which planet to save!) but it was an interesting insight into how players percieve the game.
|Posted by Rowan Powell on September 15, 2016 at 4:25 AM||comments (0)|
I worked on a small mobile game called Sun & Sky from mid July to the end of August over the summer, based off an old tactics based idea for a game that had been bouncing around my head for quite a long time now - reclaim the worlds from.
The core concept is inspired from a webgame I played as a teenager called Creeper World, essentially a never-ending flood of 'creep' pours from certain areas of the map and you need to hold it back while you push towards certain objectives located on each world, managing your infrastructure and energy consumption. I modified this idea somewhat, removing the pylons that you needed to move the energy around and instead focus on a smaller scale of engagement - dealing with a small handful of turrets and combating terrain issues as much as the enemy itself.
This means that the game ultimately is more of a puzzle game than a tactical one, despite my original designs and motivations. The 'enemy' is a fungal that spreads across the map from hives and the edges of it's domain, which means that it's 'attack' is fairly unsurprising and you don't need to react in most scenarios, given that you've planned properly - with the exception of more advanced management of power toggling your buildings. This means that for the most part the 'challenge' is finding how to lay out your buildings in a way that they don't get overwhelmed and can make the most efficient progress.
The enemy types then help to cement this idea of careful placement over all else; the rapidly growing clusters will quickly reclaim ground where any mistakes leave an opening in your lines, long-range spores will cause issues behind you if you don't have the ground appropriately covered, clusters will punish you for leaving a flank exposed by launching attacks that can quickly pick off nearby solar panels (The core resource generator in the game). I'm very happy with how these 'enemies' worked out to play against as I didn't design them to 'do' anything strategy-wise except provide a new set of challenges and the level design allowed me to capitalise on where those mechanics were the most interesting. The reason this makes me happy, rather than simply having designed a good mechanic, is that I often see people theory-craft new concepts and roles far too extensively, thinking about how each little scenario plays out. Not to say this is purely wrong, I theory craft all of my designs to some extent, it's important to think about *why* you're adding something to the game, but people often get caught up in details that don't matter or mechanics end up interacting in ways they don't expect.
The turrets and terrain are where the puzzle elements of the game really start to show through though. The game starts the player off with a basic Flamethrower (Short range but quickly clears any nearby tiles) which serves the player well until they meet water tiles. Water tiles cannot be built on and Flamethrowers are too short range to shoot across them, so the tutorial quickly offers a new solution - the longer range AntimatterBioCannon (ABC for short). The second level allows quick-thinking players to take advantage of the longer range to tackle the level much more efficiently, but the Flamethrower still has it's strengths available here. The third level really pushes home the play between the two buildings available to the player, some parts of the level are simply inaccessible to the player without grasping that ABCs can get you over gaps and Flamethrowers can quickly clear a 'landing' area.
My favourite mechanic by far actually crops up a little later into the game (Level 7, the first map on the third planet) which is the Dehydrator. This building has absolutely no combat functionality and drains energy alarmingly fast, but offers a solution to a problem posed from here on out - two or more water tiles are between you and the next island of dry land. ABCs can't shoot that far, but Dedhydrators can turn water tiles into dirt, allowing for buildings to be placed in range of the tiles you want to shoot. But by doing this, you create even more terrain for you to defend (A dehydrated tile is often quickly met with an incoming barrage of spores trying to infest it!).
The game also features a simple set of achievements for each level (Lose no buildings, use less than X resources, use less than Y buildings) which helps reinforce to the player the goals needed for playing the game well as well as provide a nice set of shiny gold medals for those who like to hunt down 100% completion of a game. They also unlock extra bonuses for the player if they get all the medals on a planet.
|Posted by Rowan Powell on April 21, 2016 at 6:30 AM||comments (0)|
I've been making a lot of progress on Axe Of Kings, adding lots of new content and getting players to try it out so I can try to gauge what's fun and what really should get an overhaul.
The original level designs for Axe Of Kings used the algorithm from the AI demo with some tweaks to allow it to make bigger levels, which gave some really interesting results and are enjoyable to look at ... but not very fun to play. The issues with this sort of level design stems from the fact that AoK is a turn based game, so moving down long corridors consumes a lot of actions, doesn't really give much interaction or reward for progress. You can see in the image below that the player would spend a lot of time moving around and not really doing much (Compounded by the issue of having 3 people in a party!).
The new levels are much, much more compact with small rooms being connected to each other by very short corridors, which keeps the mixups between having wider areas to kite back enemies but also choke points to take fights or cautiously move through. There are some problems in the larger levels of having the rooms all be a bit too similar (Which is less interesting due to the novelty rate, but also can end up with you getting a little lost, not ideal!).
The tree stumps and bushes also got removed, they ended up just being action-sinks which didn't really feel rewarding to destroy, they just got in the way of the interesting parts of the game, though pots take on a very similar role.
The pots in the current build block the player and enemy movement and can be destroyed (Much like the old tree stumps) but only take one hit to do so and give a big explosion of gems when they're destroyed which feels rewarding. They also have a chance to drop items which are very useful for your progression through the dungeon.
They key thing in the new builds is that there's always something directly nearby to engage with and gives the player a next task to follow which helps stop them finding a point to stop playing, the gem drops from the enemies, pots and chests help them feel rewarded for making progress through the dungeon. These (Along with some other factors i'll talk about later) have helped take the game from a stage where players would engage for maybe 30 seconds as they figured out the controls and got bored to now where the players will happily sit and play through many levels. The next thing for me to focus on is to give some more rare and interesting events in the levels, to keep them feeling fresh. I'm personally a fan of rougelike's odd hidden alternative victories such as making your way into a special section of the dungeon or bringing an item to a location, which helps for adding a bit of replayability.
|Posted by Rowan Powell on December 31, 2015 at 2:20 PM||comments (0)|
So something that I've always struggled with, simply because it requires a bit of perspective shift from my default state, is that what is fun to create and let lose as a programmer, isn't always fun to play with as a player.
This may seem obvious to a lot of people, but there's a subtely to it that's important, many times I've found that the very clever and useful bit of code I wrote into a game, the player simply would never notice without being told or just doesn't care about. I've written complex AI to challenge the player, which are not found exciting and intricate level generators which the player just breezes through.
The concept that becomes clearly more important (When you want your games to be more than an idle hobby), is the idea of code that interacts and talks to the user, rather than to itself. Tiny, simple code snippets that make elements of the UI bounce and shake, or tilesets to convey how a level will play out (more on this later!) have had a significantly bigger impact on the way users are engaging with my content.
I'm in a rather unusual position of being able to create acceptable art alongside being a professional programmer, which gives methe logical and pragmatic mindset of code to pull the puzzle pieces of a design together, but also the absolutely vital ability of an artist to realise the feel of a composition of components, beyond just the technical aspects behind it.
One of the most recent reminders of this has been my recent work on Puzzle Slide and AxeOf Kings. Puzzle slide was a very quick and dirty cobbling of some clever code for puzzle generation polished up with some cute art, informative UI and basic sound. And it was one of the most popular games I've ever made, reccieving far more praise than Anya'sQuest in it's behemoth glory, a heap of code, hard work and duct tape. My old art teacher would often encourage me to step away from trying to precisely line in the little details and just paint freely, where I was strongest and most uncomfortable. This is a lesson I need to bring to my work on games, to step back from the neat lines of code and add some broad strokes of properly engaging content to my games.
|Posted by Rowan Powell on September 1, 2015 at 5:45 AM||comments (0)|
So in my most recent project I decided to make a puzzle game, but rather than just hard-code a whole ton of puzzles I decided to make the computer attempt it for me, with mixed results.
The algorithm that I built for this is fairly simple and is based of simulating a player 'solving' a puzzle that doesn't exist and building a solution to that effectively random wandering.