See introduction here
The Game: Centipede (1 player - timed)
romname: centtime
Screenshot:

Reward: $100 (usd)
Deadline for submission to MARP: March 11, 2005 2359 hours GMT
Further Bounty: none
Centipede. This has to be my favorite videogame of all time. Nothing else ever matched the speed, the rhythm, the sense of controlled utter destruction of high-level Centipede play. Those that don't understand, well, just won't understand.
Although it may not have been done in MAME, the game is marathonable by people sufficiently talented. (I never got there.) There are, unfortunately, no tournament settings widely recognized, as per TG, nor are limiting the number of lives possible via settings on the machine. we do, however, have a ROMset which allows a time limit to be set.
I think this originated when in 1981, Atari sponsored a national competition to find the best player on the game. The firm hired by Atari to run the contest actually didn't do a very good job, but the contest was won by Eric Ginner. The machines for the contest were set to three-minute play. I assume this is where the centtime ROMset comes from. The tournament was head-to-head single elimination. I have read that his score in the final was 52,000, and that his best for the tournament was 55,000.
Let's see what we can do. Set the DIPswitch to 3 minutes as per MARP rules for the ROMset.
The best guide to the game was "The Video Master's Guide to Centipede" by Ron Dubren. Dubren had this to say about the 3 minute game:
Dubren's last remark might suggest a further bounty, but watching Don Hayes' (previously) first place recording, I see he was on the next-to-last wave when time ended. I'm guessing someone concentrating on that instead of scoring points could do it.Ron Dubren wrote:By introducing a time limit to CENTIPEDE, the entire complexion of the game changes. Unlike the long game, where the object is to stay alive as long as possible and accumulate the most number of points, in the short game there is a theoretical limit to the number of possible points you can score. There are just so many Spiders, Fleas, and Scorpions that can turn up as potential targets, just so many Mushrooms you can wipe off the screen, just so many attack waves you can clear. We know for example that a Spider appears about 4 seconds after the last one has been shot. Dividing four into 180 seconds, (e.g. 3 minutes) gives us 45, the theoretical maximum number of Spider kills that are possible in the short game.If every one of these Spiders were nailed the instant it appeared (for 900 points each), you could theoretically score 40,500 points from Spiders alone. Similar calculations of "perfect" scores for the other targets are not as easy to determine, but in practice, top scores in the 3-minute game are usually in the 50,000 to 60,000 range.
As far as I know, nobody has ever been able to get through the entire attack-wave cycle in 3 minutes. This may not be possible. Remember, the first attack-wave cycle includes those extra slow waves up to 40,000 points.
Speaking of Don, he is one of the reasons I hadn't used a Centipede/Millipede version yet for GEGOTW. He'd be tough for anyone to beat. Perhaps with only a week he will have trouble beating his own previous score. I was able to do my 53,000 in one week, but it was a week of pretty serious effort.
The emulation of the sound chip for these early Atari games was changed recently to facilitate recording and playing back with sound. It should be all straight in .93, but I haven't tested it yet. The "fix" was removed in .92, so I will avoid it. As of .91 there were still some issues with random number generation, but this doesn't significantly alter gameplay, or affect competition on the game so we will use WolfMAME .91 for this week's competition.