Strategy Guide and Tech Talk

Posted by (twitter: @RustyBotGames)
August 22nd, 2011 12:55 pm

The compo is over, my game is finished and the rating starts. For anyone trying out my game (if you haven’t yet: HERE), I will explain some strategies to survive, just in case you have trouble with it. Additionally, as I tried to get some advice from my wife for setting up the level generator yesterday evening she wasn’t really enthusiastic, to say the least. So for anyone more interested in the tech behind I will explain, why the level generator currently doesn’t work that good for providing difficulty.

The game mechanic basically works with to values that decrease as long as you are digging tunnels (and even when passing existing ones). The first value is food/hunger. You can find three different kind of food which will give you 5 to 15 points back. Food, if there is any, is only revealed if you enter a cage. By that, it is often necessary to leave the direct path and search adjacent rooms if you are low on food. This strategy also has a second benefit related to the second value, hope.

Hope is generated by spotting new cages around you. So progressing the level is the key to prevent death by starvation.  The hope mechanics leads to another point. Often it is better to move to the center of the level because chances for revealing cages is better in free field than at the level borders. This seems contradictory to the aim of getting to the upmost row or the rightmost column which are the only places to escape the level later on. Remeber that you can see 2 spaces horizontally and vertically and one space diagonally (not if you are digging). So if you already are two rows below the top border and there is no cage above, you won’t find any if you dig that direction.

The cacti walls are revealed by digging, so sometimes a tunnel can give further information for planning your path. This only applies if you have enough food/hope left.

To summarize the strategies:

  • try to reach upmost row or rightmost column of the cages but pass the center of the level
  • don’t dig to places where there can be no cage (2 spaces between player and border)
  • don’t hesitate to search rooms for food
  • progress to keep your hope filled

Headed to the center and didn't took the direct path to replenish food and hope. Else I would have died short before exit (Difficulty: Hard).

With that you should be able to beat all difficulties after some tryouts. At this point the random generator sets in. The design I used is rather insecure in providing the difficulty level. Instead of spreading a fixed amount of cages for each difficulty, I implemented to check a percentage to set a cage for every field of the grid (e.g. 33:67 that there is a cage on a field in normal mode). Instead of filling one third of the level with cages, this leads to very unsteady number of cages. By that, an “insane” level could have more cages than an “easy” one. Same applies to walls. By that the difficulty selector is a rather vague adjuster. But nevertheless you will notice the difference if you give it some tries (one playthrough is quick; around 15 seconds). So go and play. Good luck.


Just a game design related teaser for the post mortem coming soon.

Tags: , , , ,

One Response to “Strategy Guide and Tech Talk”

  1. foobaz says:

    I could be wrong, but I don’t think the random cage generation would make such a big difference. I think the binomial distribution is sufficiently concentrated that you would get close to your expected number of cages every time. I think that the variance on the number of cages on the top and right border is greater and, in my opinion, impacts the outcome much more.

    Basically, on insane, I would just try to go straight to the exit by first finding a cage on the top or right side. This is to avoid running out of food (I haven’t looked at the code but it felt like visiiting cages yield an expected loss in _food_ (not just hope) at insane).

    If you wanted a more consistent difficulty by altering the distribution of cages, I would suggest (also) fixing the number of cages on the sides. In fact, if this information is given to the player beforehand. it could make an interesting part of the game. (If they “saw” enough empty squares, they would know the remaining are cages).

Leave a Reply

You must be logged in to post a comment.

[cache: storing page]