Algrorithm Semifinal 1: Dropping like flies
05.03.06 2:40 PM


By Kawigi
TopCoder Member

I got to the first algorithm semifinal right as it was being started. Of course, the coding phase didn't start right at the scheduled 9:00 time - that was when the ntrefz's-voice-through-a-filter competitor introductions started. Everyone was ordered off the competition platform and the round 1 competitors stood under a large monitor to be called back up to the platform in alphabetical order (sorry, Yarin and Ying.

The funny thing for me is comparing this sort of techno-intro to the pomp and fanfare I remembered from the 2004 ICPC finals. With ICPC, there was a chamber orchestra playing our official fanfare and it had a much more Armageddon feel to it. This had a more Mortal Kombat feel. All we needed was some fast techno music while the names were read.

The fun thing with being here is opening the arena and telling people what's going on. Since I'm here typing this, I'm not really up and looking at the monitors too often, but I'm sitting where I can watch the leader board and tell people what's happening (of course, Rustyoldman was mainly running the play-by-play).

Right now I have antimatter and misof theorizing about the problems and what's going to happen in the challenge phase next to me, with John Dethridge, reid and kalinov adding comments occasionally. Funny statement of the morning courtesy of antimatter: "I hope SnapDragon isn't in the WildCard if I'm there..." Related quote in the arena from radeye: "snap's not gonna be happy to have to come up through the wildcard." So I guess no one wants SnapDragon in the wildcard. But he just resubmitted the 500 for 162.59 points, and he's in 6th place, so it sounds likely.

Second funny conversation - people in chat referred to Yarin's strategy of not getting the 250 and going straight to the 1000 as a " John Dethridge." radeye said, "My definition of "pulling a John Dethridge" is to write code that is body-slammed against the left margin." I'm not sure either would be my definition so much as exhausting every valid single-letter identifier in a solution.

With two minutes left, it sounded like ntrefz found the reverb settings and announced that there were two minutes left with an electronically lowered echoing voice. Yeah, so under the 2-minutes-left pressure, everyone has to stop and laugh. The same overly-reverberated voice announces the end of the coding phase, and the start and end of the challenge phase.

The challenge phase wasn't any faster than one would think, but was still interesting, with at least 2 successful and 4 unsuccessful challenges. Petr looked like he was padding his lead with one successful and one unsuccessful challenge. What was really interesting was watching the system tests results come in at the end - Each person's row was updated starting from the bottom of the standings after the challenge phase, and we heard yells and jeers as it started to almost look like EVERY solution was going to fail. Finally, about half-way up the board, we started to see people pass one or the other problem, but not more than one until cyfra, who was previously in second place. When Petr and SnapDragon's 500s failed, it became apparent that Petr's challenging was not padding his lead, but allowing him to make it to the finals without having to go through the wildcard round by .25 points. That just proves that every second counts (for either one of them).

One thing I forgot to mention yesterday was the publicity here. People say that there's more than in previous years, but I don't have anything to compare it with. People are walking around with cameras and video cameras, interviewing contestants and spectators (I was sitting next to Ruberik explaining the challenge phase to one of them earlier). At some level, you don't really understand the pressure and nerves of the challenge phase unless you've done one where you knew that advancing, money, or your rating was on the line. The media isn't obtrusive, but they are walking around sometimes in places where it seems like they shouldn't be :-) I just ignore them and assume everything's ok.

Hmm, since writing that, I just got interviewed impromptu about Petr's narrow 2nd to advance. I shouldn't talk so much. Anyways, see you next round!