Jon and Dirk discuss a few games they’ve played lately which make good first impressions but ultimately fall short: ‘Don’t Starve’, ‘Shogun’ and ‘Wiz-War’. They then transition over to talking about what they’ve been working on, including AI character personality and AI design in ‘At the Gates’, getting into the meat of development with ‘Futbol Strategy’ and switching developers for ‘War Stories’.
We covered some interesting games in this episode, all of which I’ve really enjoyed, but also ones I have serious concerns about. I’ll go over each of them briefly:
Wiz-War
This game can be forgiven if only because it’s a very old design, and the fact that it still holds up today is a testament to how well-designed it actually is. There are some interesting things going in with the environment, as the board is somewhat randomized and players have the ability to alter it with spells.
The two big problems I have with Wiz-War are its failure to fully capitalize on the theme, and the fairly bland map pieces. You might be able to do interesting things with it, but the board pieces themselves are nothing more than a generic underground labyrinth.
Shogun (the board game)
The best way to describe Shogun is “Risk with an interesting economic system.” It’s probably my favorite of the three games we covered in this show, but its incorporation of victory points makes me cry inside. I’m also not a big fan of how it ties victory and strategy together in a more general sense: laying low is usually the recipe for success, and like most such games it is often decided before players have made any decisions.
Don’t Starve
I was really, really enamored with this game for a week or two, but my interest rapidly dissipated once I got past the 15-hour mark. Given how much fun I was having, it took me some time to identify why this happened.
The issue with Don’t Starve is that it doesn’t force players to adapt enough. The game has a decent amount of variety, but your strategy from game to game tends to become fairly rote. There are new opportunities each time you play, but no real reason to leave your comfort zone. A lesson I’ve come to learn is that content is often wasted when you don’t force players to experience it.
– Jon
Jon Shafer on Design 
Tend to agree about Don’t Starve.
I found that in comparison to the traditional roguelike, I ended basically pursuing the same strategy every time I tried to play, hoping I didn’t make some silly error three hours in and die.
The odd issue being that Don’t Starve simply wasn’t random enough to elicit something different each time you play. It felt like each death was a sentence to a repeat a 90’s era JRPG cutscene again and hope I didn’t die to the boss *this* time.
Most roguelikes introduce enough randomness in the early game that it can totally alter your approach to the game. In Nethack, finding an altar on the first floor can change your entire game; or finding a lucky artifact pull. The game is hard enough that if you get lucky, it’s by no means a guarantee of victory.
Agree completely. It all goes back to the very first article I ever wrote (seems I chose a good topic to get the ball rolling with!).
Sure, you always need a wide variety of interesting options, but if there’s nothing pushing you – very specific goals, a time limit, dangerous circumstances, limited resources – then all you do is play around a bit with whatever catches your eye, until eventually your natural proclivity to explore just for the sake of it wears thin and you get bored.
For some people that’s plenty, but I only really enjoy games when I’m able to get really deep into them. I know I’m not alone there.
– Jon
This is the same line between Minecraft and Skyrim.
Minecraft’s engagement is predicated on the player constructing his own goals, and the game produces the mechanics and avenues for the player to do so and defines tasks to accomplish to pursue player set goals. But a player who never defines an objective for himself in Minecraft will find himself dreadfully bored, dreadfully soon.
Skyrim gives a world in which to engage goals provided to the player (although players still retain the freedom to self-engage). Then when the player pursues these goals, invariably self-dictated goals emerge along the way — mastering a skill, crafting ebony armor, etc.
Most roguelikes grant you a version of the Skyrim formula, only in a less expansive world. They provide a world that must be explored on account of being randomized and add flux in the scenario mixture to give hooks to the player for setting his own goals in pursuit of a larger objective (get to the other side). Then they introduce risk through combat and uncertainty to drive the player to adapt in pursuit of goals. Skyrim doesn’t generally require the player to adapt in pursuit of self-set goals like this, and replaces that impetus with storyline, quest structure, and crafting or gear based progression trees.
If we had a “Minecraft – Skyrim” spectrum — Minecraft a -1.0 to Skyrim at 1.0 you could probably place a lot of these games on that spectrum somewhere and argue their case.
Terraria in the -0.4 area — possessing many of the freedom driven goals of Minecraft, but with an underlying mechanical progress structure through gated crafting trees that require boss kills and certain resource harvests to proceed.
Elona is a semi-roguelike, much more in the 0.8 – 0.9 region — It plays like many roguelikes, but discards permadeath in favor of more Minecraft elements and very open-ended progression. It’s a canvas of systems where a player writing his own narrative gets devoured, and one looking to the game for narrative impetus flounders and grows bored.
Don’t Starve wants to be in the -0.8 area, but it lacks enough non-linear progression systems to deliver there, and hangs onto permadeath in a design model where it may not really belong. Now if Don’t Starve killed your character — but not what he had built into the world, and delivered a much deeper crafting progression tree I could see the game suddenly becoming a great deal more engaging.
We could make this problem clearer by imagining a version of Minecraft where your world is destroyed when you die. And who the hell would want to play that?
I think Don’t Starve is extremely boring as a game. It’s more of a toy actually (albeit not a really great one either). You run around this (mildly random) open world, discover what’s there, build some things, then it’s basically done. But then it has permadeath and a score and all this “gamey” stuff, that doesn’t really fit. It’s like Minecraft’s permadeath mode. It feels like it actually wasn’t designed to be a game, but they wanted it to be a game (as an afterthought, really)…
It’s strange how the gaming industry would think of the “hardcoreness” (the challenge) and the randomization of roguelikes as “bad game design” today. Challenge is a pretty core element to games, that e.g. seperates them from toys. And any single player game needs randomization, else it becomes a puzzle about figuring out the right path and not about learning and making tough decisions inside a system. I think the reason those central GAME features (yes, that’s what they are, roguelikes did not invent them, it’s just that roguelikes are kind of a “bastion” of real game design today) are somewhat despised is, that most games of today are just bad and so are the designer’s “best practices”, which mostly evolve around manipulating the players to “like” something, that basically has little to no gameplay(!) value.Their job is to figure out how to steer the “player” as smoothly as possible through their “game” while constantly keeping him “happy” with flashy scheduled rewards.
Funnily enough, the biggest problems I have with some roguelikes actually are exactly some of these “best practices” (e.g. loot skinner-boxes, sacrificing tightness in design for content and inherent complexity etc.).
Therefore, the best roguelikes to me are really the more focused and stripped down ones, e.g. Shiren The Wanderer, Zaga-33, 86856527, Brogue, 100 Rogues (which sadly is ruined by a game-breaking bug introduced witth the latest patch at the moment). Or even Desktop Dungeons (albeit I think the alpha version did a lot of things better than what is to become the commercial release).
It’s also kind of funny how we call anything including permadeath and/or random generation a “roguelike”. Spelunky is really not “like rogue” at all. Spelunky is just simply a platformer, but the first one to somehow get it right, the first one trying to really be a GAME, that’s randomized and that you can actually lose, and not an execution/memorization puzzle.
Don’t Starve is a great initial concept, but there does seem to be something important missing.
The best games are those which find a way to hit that elusive middle ground of “structured freedom.” Where Don’t Starve falls short with me is that lack of structure you’ve pointed out – no real goals, and little that forces you to adapt from game to game.
You’re right that “roguelike” probably isn’t the best way to describe a game like Spenlunky, but genre names are often subverted into something completely different from their original meaning. Paradox’s titles might be real-time strategy games, but they’re not RTSes. Because permadeath and world randomization are so rare feature in all true roguelikes the two are now forever associated. A good example of why I avoid semantic arguments over definitions/categorization that, ultimately, doesn’t really matter!
– Jon
I agree, GENRES don’t matter and, if anything, even hurt creative thinking and prevent design from the ground up.
However, I do think it is very useful to think about what defines games and what seperates them from other kinds of interactive systems. It’s useful to draw the same line between Spelunky and Super Mario as between Chess and Sudoku. Just that the former is rarely ever drawn by anyone talking about “games”, which has become kind of a catch-all term for basically any kind of digital interactive entertainment. And I think that’s actually harmful.
This is like the video game version of Sherlock Holmes website, the Science of deduction.