FGS strikes back...

   As noted in my Crystal Castles - the booklet writing, I had my (former) world-record score of 898,815 beaten on Crystal by my friend Frank Seay (FGS).  On 08-09-1987, he made an incredible 901,290, making him the first player to cross the difficult 900,000 barrier.  Of course, I would have liked to have been first, but, I hoped to beat his world record score in the months that followed.  Unfortunately, it was not in the cards, and I never have topped his (or even my own) score.

   On his record game, he had a 133,000 time bonus, and of course a 60,000 life bonus.  He got both witches and all last gem bonuses, but died once, on 8/2.  At the time, we discussed what we felt was an ultimate potential score that could be attained.  I thought a reasonable maximum was around 904,000, while Frank considered a 902-903 the max.

   Well, it turns out that we were both wrong, though not in the way you would think.  To not jump too far ahead, let me say that Frank topped his 901,290 a number of times (around 10 or so, by his estimate) in the remaining months of 1987.  His best score was a 902,265, and he had one other 902 score as his second best (902,105).

   Now, on to the interesting part.  Frank got to thinking about how we had point-pressed the game in level 8 and above, and began to consider the possibility of doing the same with the lower levels.  Of course, he and I had both considered starting at other points in order to have a net gain over the way we normally played.  But, we had previously only considered 7/1 as a viable possibility, as we didn't want to lose a lot of time bonus.  While starting there might give a slight net gain, it would add four more screens to mess up on, and this hardly seemed worth the small potential gain of one or two thousand points.

   What Frank and I had basically not considered was the possibility of forgetting about time bonus altogether, and starting at (or near) the beginning.  This was perhaps because getting a good time bonus was so ingrained in us.  Now, because of the fact that the first level has a relatively low amount of points, and considering time bonus trade-offs, the only other place to start a game and realize a decent net gain in points over playing 8/1 up is level 2.  So, after confirming that the potential was there, Frank set off on this new quest.

   Note that while the lower levels (say, five and below) are relatively easy, there are still a large number of boards to play, and even missing just a gem or two on each one has a debilitating cumulative effect.  So, it is still not easy to play a game like this, especially since it takes a lot longer (20 minutes or so, versus less than five our "regular" (8/1 up) way).  But, two huge advantages are not having to worry about playing quickly for a good time bonus, and being able to afford to take a few chances in the lower levels, if necessary, since you only need a couple of lives going on to 7/1 in order to have the maximum life bonus at the end.  Still, consider that using this method requires you to not only have to play the levels below 8 in a good fashion, but even if you have a super game going to that point, you still have to play 8/1 up in a great way in order to max the score out (the ultimate goal, of course...).

   Well, Frank practiced the lower screens, developed some great patterns, and played a number of months before finally beating his own world record score.  His first time to beat it, he got 904,168 points (unfortunately, he doesn't track things as steadfastly as me, so the details on this, and other scores, are unavailable).  He later had a 906,128 game.  Then, on 03-17-1988, he set his current world record: 910,722 points.  The top three scores on his machine are now the latter two mentioned, and a 904,866.  Not too shabby...

   Interestingly, in my original Crystal Castles booklet, I estimated a theoretical maximum score for the game (before he made his score), and it turns out that I calculated it to within 4or 5 thousand points.  Not bad, and with a recent update I've done to make my estimates more accurate (mainly for the level 6 tree waves), I've calculated the maximum final score as 909,999 points - less than a 1000 away from Frank's score !  Of course, he's taken great delight in pointing out that he's exceeded my "theoretical maximum" !!!

  Speaking as someone who is intimately familiar with the game, Frank did an incredible job to get a score so high.  He played every board in a near-perfect fashion, and his patterns for many are very difficult to execute, so, to have played a total of 33 boards and had very few come out below near maximum is quite incredible.  Still, I have seen Frank play through the game, and there are places that the patterns can be bettered, and with a minimum of extra risk.  If I have time in future years, I will pursue chasing his score.  The only question is, which one do I try for first - 8/1, or 2/1, up ?  Stay tuned, and, try checking back in 10 years or so for the answer...

Click above to go back to the Crystal Castles page...