[Jack Simth]

You'd lose that bet - there are a few too many types of things (and a few too many specifics under each type) of what it might be for me to run through them effectively in the allotted time.

Some samples of possibilities:

1) Each grouping could stand for a single letter, in which case (for out purposes) it is an unbreakable one time pad.

2) The groupings could be a distraction, and it's just 36 letters scrambled in a particular fasion.

3) The groupings could be a distraction, and it's a simple letter-code which then can be broken into words by adding spaces at appropriet locations

4) it could be coding for a small ASCII picture when arranged in a particular fasion (36 letters can be evenly arrayed out into 1x36, 2x18, 3x12, 4x9, 6x6, 9x4, 12x3, 18x2, or 36x1). If the spaces are meaningful, then there are 47 or 48 characters to deal with (47 is prime, and can be discounted for picture purposes).

5) it is possible some set of symbols needs to be filtered out to get the message

6) it could be random garbage

7) it could be padded with random garbage

8) something I'm not thinking of at the moment

9) it could be a specific, used form of encryption

10) it could be an idiot code, where each segment means something long

11) a variation on the above

12) just about any combination of the above

Without a much larger sample size and/or further data, no algorythm will produce better results than guessing or exhaustive search (algorythms specifically designed for a type that the code just happens to be count as guessing, as they eliminate the other types by random chance of being that particular algorythm).