Bejeweled Blitz seems to be the latest craze on Facebook, if the number of my friends playing it is anything to go by. Back in my university days, Bejeweled was a big distraction when I should probably have been revising. I used to play it on a Sony Clie - a Palm OS 4 based device. Bejeweled was well suited to the stylus and high resolution (320 x 320!!) colour touchscreen, while the 33 MHz processor could just about handle it smoothly. I used to play for hours on the time trial mode, racking up enormous scores with each game lasting for what seemed like an age. A year ago I tried to recreate those days on my Nokia mobile phone. Not much had changed - except that it had been snapped up by a big publisher (EA). Using the key pad control system made it frustrating to play, particularly in time trial mode. I played it a few times, and then thought perhaps my Bejeweled days were over. That's until I was introduced to Bejeweled Blitz on Facebook.

The big difference for the Blitz game is the introduction of the fixed 1 minute play time. Time is not extended for achieving matches or combos. Other differences are the power gems awarded for matching 4 gems in a row, and the hypercubes awarded for matching 5 gems in a row.  Given the short play time, the key to the game is racking up points with a big speed bonuses and multipliers in hand.

After playing a few games, I began to tire of Bejewled Blitz. How big (or small) a score I achieved seemed to be more by random chance than by my actions. For example, sometimes I would rack up a huge score just watching gems fall and match instantly and continue to chain with others for quite some time, without any interaction from me. Of course part of the skill could be in setting up these kind of chain reactions - or at least swinging the layout of the board in favour of them - but for me it just feels like there is too much random chance for it to be compelling. The person on the top of my leaderboard (i.e. out of my Facebook friends that are playing Bejeweled Blitz in that week), the people who end up being top of the list are the ones who are investing the most time in it. This leads me to my theory that the chance of achieving a disproportionately large score is just as probable as a small score... so that if you play enough games, you are likely to get a score good enough to make it to the top of your personalised leaderboard.

Given that I don't have time to sit down and play hundreds of games of Bejeweled Blitz, there was only one way to test this theory: by writing a Bejeweled Blitz Bot. In part 2 I will go into detail on the making of the bot. Until then, why not have a go at Bejewled Blitz and see what you think of it?