On Making Game-Bots Considerate

So. Ranting is fun. Sharing ideas is great. Since starting dev work on PMDiscord, I’ve gathered more things I’d like to share than I have the time to write — and with the advent of PMDiscord getting a Patreon page, I’ve now got an even bigger incentive to start. Though it’s my first step into game design in several years, I was pretty surprised at what it’s given me insight on – not just regarding how games and text games are played, but also about Discord games in particular. It’s perhaps the most unique thing I could talk about, so allow me to share one topic of what I’ve found.

For all its limited features at release, PMDiscord managed to check a lot of boxes I hoped it would. Players caught onto the game’s mechanics sooner than expected, the battle system was received well, and it felt like there was just enough depth to enjoy it in short bursts. But when a large handful of players were caught grinding on the game for hours at a time, it got me to consider what this implied — from the bot’s perspective, rather than the persons playing it.
“So, your game’s addictive? That’s not so bad, right?” Well, yes and no. While compelling and engrossing games are good, ‘addictive’ games usually aren’t a positive thing — and it certainly isn’t a target for PMDiscord, which is both a game and a Discord bot. This distinction is important, I feel, and I’ll get back to that in detail later.

Either which way, it soon struck me that PMDiscord was missing the mark for what I intended for it to be. Players tended to get caught playing ‘one last battle’, and then do the next ‘one last battle’ right after that. Some admitted to skipping sleep so that they could chase that last team of Pidgeys out of the server for good. One particular case had a player asking to be banned from the game until they had finished their exam studies (to which I had to oblige). What was meant to be an accompaniment for a Discord server, had taken center-stage in a way that I wasn’t completely happy with. Now, sure, had PMDiscord been some kind of classic MMO game, having this much draw and appeal would have been called a success. Popular mobile games live by this rule, using addictive elements like gambling and false time-investment to keep their profits high. Of course, this isn’t the kind of trend that I want PMDiscord to follow after. But looking closer into it, and looking at the failsafe I already had in place, I began to think: Was there anything that made PMDiscord, and other Discord-based text games, different? What gave them so much draw, without all the colorful animations and ease of use that your everyday mobile game has? And how much of that even mattered?

PMDiscord does feature the usual culprits for compulsive gaming: characters that level up, along with randomly-generated events, are two of the main staples when it comes to keeping a player’s attention. But, I actually think that Discord games have another slight advantage in this area, depending on how you think about it. Unlike most games, a Discord bot can alert you to play outside of hours when you’re actively gaming or using your phone. It’s always there, even when you just want to chat with friends or check a few announcements – and a message notification on a channel would be all it needs to entice users into playing again. Technically, the bot wouldn’t be doing anything wrong. It’s up to the player how much time they want to sink into the game, and they could always block it if it’s distracting them, right? That’s fair to say, but as we know (and as PMDiscord has unfortunately proven), things don’t always work out that way. So with that knowledge, I started looking into something I would call ‘considerate practice’.

There’s a small caveat to consider here, before I go on further. At present, ‘considerate practices’ go against everything that can make a modern game more profitable, which is maybe a little unfortunate. Someone looking to make more revenue and nothing else, would have no reason to implement (or consider, heh) something like this. But, I wanted to explore the idea of making a Discord bot more enjoyable by making it ask for less of a person’s time — essentially making it act closer to most Discord bots’ good practices, and less like a constantly-active videogame. And since PMDiscord isn’t made to earn any direct revenue, it had nothing to lose from the little experiment I was about to do. The expectation was that, if the bot shared some of the players’ responsibility, choosing not to distract or beg for attention more than it needs to, it could still compliment a server like any other bot would, without being a hassle because of constant messages or endless grinding sessions.

While looking for ways to go about adjusting this, I started at the top. Level-up systems are the tried-and-tested reward system: it gets used in everything nowadays, and it doubles as a convenient skinner box to keep players invested. For PMDiscord, this reward is very necessary to provide a sense of achievement, and is almost required in some form thanks to the games it’s based upon. However, the grinding behavior that it encourages is something that I felt I needed to trim down. Even before I considered this a problem, I had tested out a few less-standard design choices to try and limit how much players would be given incentive to grind more often. The main choice that was made was to make level-ups in PMDiscord an incentive, but not a requirement. A level 5 PKMN in PMDiscord still has a fighting chance against a level 50 PKMN, differing by only 5 stat points. There is substantial improvement given for leveling up, but it is not so critical that lower-leveled players ever need to be separated from higher level ones. I had watched very few cases where someone would grind for the reason of catching up to other players, and so I considered this to be a small success.

This didn’t stop a large amount of players from getting caught in grinding marathons, however, and so later on a second change was added to help keep this in check. After every 25 minutes of enemy spawns, the bot would stop spawning enemies for 10 minutes. For how simple it was, it was surprisingly effective in practice, with only a small few complaints given in feedback. While it meant there would not always be a challenge available on a server, the wait would never be too long. If someone was playing as many ‘one last battles’ as they could, the game would give them a window to stop after 15-30 minutes of playtime. I don’t consider this to be a perfect solution, as being able to earn rewards from the game ‘on demand’ would always be preferred from a player perspective, and this is something I plan to address later. But as far as respecting the player’s time goes, I’m mostly happy with the result, and would maybe encourage anyone making a similar game to experiment with irregular spawning/event patterns like this.

If there’s anything that I’ve personally taken away from this, it’s that a great deal of ‘meta’ elements can drastically change how players interact with games, and they really shouldn’t be overlooked. I’d dare to say that it’s just as important as the game itself. But also, I learned that the power of level-up systems really can’t be understated, and adding incentives via random chance works far better in practice than I imagined it would. It’s a bit scary, but I don’t think much can be done about that until the majority of us become really savvy about it. While ‘addictive’ mechanics _can_ be beneficial in some ways, I do think there should be some consideration spent when implementing them. There’s been a few cases of other Discord game-bots relying on addictive feedback loops and constant message prompts, and even paid loot-boxes in one case — which is fine by them, but it’s certainly stretching further than what I would consider doing myself.

I see a lot of potential in the medium, and possibly the chance of text games or MUDs making a resurgence with popular platforms like Discord powering them. Honestly, a story-driven Discord RPG would be incredible to see down the line. But going overboard with intrusive and addictive designs could perhaps harm the medium in the long run, and in the worst case scenario, give it the same reputation that free-to-play mobile games have. Is there a need for more bots to be more considerate? Is it necessary, viable or even feasible in some cases? Will any of this become a problem in the days to come, or will it become a standard? It’s too early to say just yet, but that’s all the more reason to explore and tinker with the ideas we currently have.






Leave a Reply

Your email address will not be published. Required fields are marked *