Tuesday 26 March 2013


Assuming you count 7DRLs as roguelikes, I've just beaten my second roguelike ever: Michael Brough's 86856527. I suspect this is the game that Keith Burgun is trying to make with Auro - you should go play it (and ideally beat it) before reading any further as what I'm about to say involves spoilers. For the record, I got 21 points, using three programs.

I'm going to commit the cardinal sin of back seat redesigning a 7DRL - or at least part of it. The programs in the game are generally 'top tier' in the sense they are equally useful for their cost, although there are some programs which are clearly great in combination. I'm going to focus on the few programs which are 'second tier' in the sense they are missing one of either overall utility or interesting synergy: the four I feel fall into this category are ANTI_V, KILL_D, DCRPT and to a lesser extent BOMB - I'm still undecided about POLY although the suggestions below indirectly buff it.

The first three second tier abilities are all enemy specific, and are still useful in the sense that every enemy is a distinct challenge, but they don't have much synergy with anything else, and are still not powerful enough to be necessarily a win-button against their specific nemesis. The challenge of course is figuring out how to improve them while keeping close to the terse descriptions in the current game. My suggestions are ANTI_V should change to 'Damage all virus for $', KILL_D should change to 'Destroy nearest Daemon[s] for [two energy symbol squiggles]' and DCRPT should remain unchanged but end up with a buffed implementation.

Rather that using ANTI_V and KILL_D to directly credit the player, the floor grid under these affected enemies should get an additional credit or energy when the ability is used provided it hasn't already been siphoned. This allows the player to take advantage of the level entry spawns to increase the overall floor grid values, and gives POLY a purpose beyond helping destroy enemies.

DCRPT should be buffed so that while you have it active, you are given additional information about the play area. My initial thoughts were along the lines of having one block with ???? for the program/score and ? for the alert level on each level which DCRPT would reveal, until I realised that there is one other fact which is hidden which would be useful to know: the identify of spawning enemies. DCRPT should reveal what each enemy spawning is instead of the Transmission icon, giving you an extra turn to plan how to handle incoming enemy(s).

BOMB is incredibly useful given that it doesn't take time to use: but it has the downside of destroying adjacent walls which would otherwise channel enemies towards you. I'd like to see the floor exposed by bombing walls to get either energy or credits like other floor sections, to compensate for this and add a further level of strategy to bombs.

As for further program ideas - I'm sure there are plenty, but I like the sparsity of the current design. I am going to suggest one, borrowing an ability in Auro, because it seems to me that the movement abilities .WARP and .PUSH are the most useful and interesting part of the game, and to those, I'd like to add .FORK.

.FORK costs two (or three?) credits, and allows you to 'Make one move freely'. Since using a program takes no time, .FORK allows you to move without your enemies responding - but also make illegal moves such as moving towards an enemy (when you'd normally attack), move onto an enemy or wall, and move off the edge of the play area, warping from north to south or east to west edge or vice versa. Fork allows you to .DEBUG suicide in the finest Nethack tradition and it costs credits to fulfil the old adage that time is money and to try to prevent .EXCH/.SIPH+ abuse.


Darren Grey said...

Well done on your victory! Though I think the game is more about the score challenge than trying to directly win. After the first win all you care about is score.

Your Fork suggestion is pretty similar to what Push is used for really. Could be handy, but fairly niche difference.

Also all abilities take no time, not just Bomb. Apart from Wait of course. The demon/virus killers are still fairly weak though - if you have them they can come in handy, but they're not worth spending a siphon to get. I think the best way to improve them is to make them act on all enemies, not just a single type, but increase their cost.

Brog said...

Hey, thanks for this, I'm working on a more fleshed out / balanced version and this is helpful.

- you're underestimating .ANTI-V; if anything it's too strong (but I'm not planning to weaken it).

- I'd been thinking about adding resources to tiles as a way to improve .DCRPT but hadn't come up with anything fitting; it might work better for .KIL_D, yeah.

- I like both of your ideas for .DCRPT, I'll try them out.

- yeah, putting resources on tiles cleared by .BOMB had crossed my mind but then I didn't follow it up, I should.

- I have an ability very much like your .FORK in an unfinished game, I hadn't thought about applying it here. Sounds potentially really interesting, I'll try it.

Andrew Doull said...

Darren: When you catch up to my score let me know :)

Andrew Doull said...

Brog: Thanks back at you. Thinking about this more .LOOP is a better name for this ability as it hints at the non-Euclidean effect.

Looking forward to the new version.

Andrew Doull said...

Just tried a ANTI_V/POLY run and died on level 8 with 1 credit and 47 energy. Sure, I could have probably pushed my way to victory but it felt like a great combination foiled by a lack of resource.

Darren Grey said...

I'll beat your score when I stop being too greedy for points ;) I keep getting good runs on level 7/8 ruined by a tempted score 9 box.

And I do love the resource split in this game. It can be frustrating as hell, but it forces you to really pay attention.

Joseph said...

When can we expect to see Un86856527?

Andrew Doull said...

I'm getting a bit over that... I may have to start making original games.

Darren Grey said...

I should note here for posterity that I have won with 81 points. I can't allow Andrew to best me at a roguelike! ;)

Joseph said...

You really thinking of starting up your own project? That'd be very interesting.

Keith Burgun said...

Cool, good stuff. I would just note that this is really very different than AURO - it probably has roughly the same amount of similarity to AURO as Zaga-33 does. But, anyway, the reason I like this game more than Zaga-33 is that there's a bit more donkeyspace. I'd like it if there was even more, personally, but as it is, good stuff.