New Project – Reverie
Oct 30th, 2009 by AlexWeldon

Having recently decided to make a split and treat game art as my job and game development as a hobby, I’ve started a new project, more experimental than previous ones. The working title is Reverie, and I’d call it a roguelike, though it has less to do with the Big Four (Nethack, Crawl, ADOM and *band) than they do with one another. For one thing, it’s intended to be much shorter, beatable in a single sitting. For another, pretty much all the numbers will be hidden from the player – the game will be much more narrative and less about min-maxing than other roguelikes, and many things about it will be intentionally mysterious to the player.

That being the case, I don’t want to say too much about it, but because I love procedural generation so much, I wanted to share some examples of the maps that are generated for the huge forest in which most of the game will take place. My top priorities were to have lots of connectivity to encourage exploration, and a very organic feel, unlike the blocky dungeons that are most familiar to players of roguelikes.

Anyway, here are three maps (the lower one of which has been magnified 2x). Black pixels are impenetrable trees, light green is open grass, the two darker shades of green represent different densities of brush, and the yellow lines are dirt paths.


PC/Mac Game Review – Telepath Psy Arena 2
Oct 18th, 2009 by AlexWeldon

Gameplay: 5/10 Graphics: 4/10 Sound: 7/10 Originality: 7/10 Overall: 5/10


Available for download direct from Sinister Design.

When I saw this game mentioned on Rock Paper Shotgun, I went straight to the creator’s website and bought it, without even trying the demo. This is an indicator of how much I loved the concept. It’s a turn-based, top-down, tactical battle game. Turn-based games in general are a sadly-neglected genre these days, especially turn-based strategy games. Even within the genre, Telepath Psy Arena 2 is unusual in that it does away with randomness – attacks always hit, and always deal a fixed amount of damage. Characters are persistent from one mission to the next, and death is permanent. Although it’s easy enough to purchase a replacement for a slain ally, all the money invested in training the character is lost, and it can take a while to get the replacement’s power up to par with the rest of the team.

There’s a lot to love in that concept, and it seems like a recipe for success. Unfortunately, the game does almost everything else wrong.

For one thing, the game is riddled with balance issues. The player’s characters have much higher movement than similar characters on the enemy side, so there’s little challenge in outmaneuvering the enemy. In particular, the Assassin character type can move 9 squares per turn, which is greater than the size of the map in its smaller direction, and quite early on, acquires the Leap ability, which extends the character’s movement range as well as allowing it to pass over friends, enemies and some obstacles. Combined with its ability to deal massive damage when striking from behind, investing a heap of money into an Assassin’s strength attribute (boosting attack damage) allows an insta-kill of almost any one enemy per turn, anywhere on the map.

The creator would probably argue that the Assassin’s offensive capability is balanced by its relatively weak defense, and that moving it far from the rest of the player characters to kill a distant enemy is likely to result in the Assassin being killed, but this leads us to my next major complaint about the game, which is that it’s infinitely grindable.

Any battle can be replayed multiple times. I’m not sure whether there’s a limit to how many times, since the game does seem to keep track of how many times you’ve beaten a given battle. It’s irrelevant in any case, as the game also allows you to fight battles against random assortments of enemies of whatever difficulty level you like. This sounds like a nice feature, but the problem is that it means that, rather than forming a better strategy in order to get past a difficult level, the player need merely spend an hour raking in cash by beating up on easier foes, and buy his way to victory by upgrading his characters.

The reason for the inclusion of this feature is obvious – if the player had to progress through a series of ever more difficult battles, without being able to engage in optional ones to earn money, then there would inevitably be a moment of Pyrrhic victory, in which the player’s team would be left so crippled that the subsequent battle would be unwinnable. This feature is not a good solution, however – aside from the above problem, it also means that character death punishes the player, not by having fewer assets at his disposal for the next battle, but by forcing him into a long series of boring, repetitive, easy battles in order to purchase and train a replacement. This is not good game design.

Technically speaking, the game is amateurishly programmed. There is little animation or special effects, and nothing in the gameplay should require much processor power, and yet the game crawls. Part of the blame probably lies with Adobe, as the game is written in AIR, and I’ve found that ActionScript in general runs poorly on Mac. Nonetheless, I’ve seen plenty of more elaborate ActionScript games that have run smoothly, so I’m positive that a more talented programmer would have been able to make this game perform well.

All of this could be forgiven, if the game’s most important aspect had been given more attention. I am referring to the AI.

If you make a game single-player, turn-based and luck free, you’re committing yourself to writing a strong AI opponent. Simply pitting the player against ever-stronger, but equally stupid opposition is fine for an action game, but it feels cheap in the context of a strategy game. Imagine purchasing a chess game, and discovering that increasing the difficulty simply gave the computer more queens, rather than having it make better moves.

Without having seen the game’s code, I nonetheless have a pretty good idea of what sort of decision-making it employs: first, it checks if it can kill any of the player’s characters. If so, it always does so. If not, then it attacks the one with the fewest hit points remaining. If it can’t attack anyone, then it simply moves one square in a more-or-less random direction. Enemy characters with healing abilities always default to using those, rather than attacking, unless there are no injured allies in range. It never makes any attempt to keep its characters’ backs covered, either, which simply increases the power of the Assassin character. This extremely predictable, simplistic behaviour on the part of the opponents makes the battles rather tedious.

I wouldn’t recommend that anyone buy this game. It would certainly be worth a download if it were a freeware title, which is what it feels like, but it isn’t executed in a professional enough manner to be worth the price tag, low as it is ($12.99). Those craving a challenging, turn-based strategy experience would be better off checking out Battle for Wesnoth, which is actually free, and open source. Meanwhile, game developers thinking about making such a title can learn two important things from Telepath Psy Arena 2 – firstly, they can learn from its mistakes, and secondly, my immediate and unhesitating purchase of it should confirm that there is a market for such games, and that it’s woefully under-served at the moment.

»  Substance: WordPress   »  Style: Ahren Ahimsa
© All posts (c) Alex Weldon as of date posted