Bot Contest

Three days till the top ten Bots are chosen. I'm excited all over again! Now hopefully I'll have some time for some last-minute tweakings.

Top ten bots? *ears perk up*
What are they being chosen for?

Chatterbox challenge... agentland.com

It's at chatterbox challenge.com. It's an annual contest with a top prize of $1000. The top ten will be posted tomorrow.

Then everyone entered has two days to send in their votes for an additional two. Once the twelve are chosen, the judges will speak with each individually.

someone please vote for steve.. he seems to have taken a nosedive in ratings

Looks like the popularity vote counted very little in the end. It was enough to get Sarah & MarkBot edged out of the running and put Talk-Bot & Jabberwock in. Oraknabo's been consistently in the top 5 of the popularity vote for the past 2 months, yet mysteriously dropped to 8 in the past couple of days. I doubt even first would have gotten me in the statistical top ten though anyway, but at least I'm still more popular than ALICE.

I doubt I have much chance now. my goals in writing Oraknabo are just too contrary to the goals of other botmakers for me to get many votes. Too many use it to make nice, unoffensive surrogete friends. You can see by the comments I've gotten that many can't even concieve of the reasons for creating a rude, evasive bot. I guess it's somewhat akin to Muhammed Ali always scoring badly because the judges weren't always able to understand his strategies.

I also can't believe that his excellent performance in the sample conversation counted for nothing.

Well, hot dog, I made it into the finals. I guess they like my service. I'll have the bar spotless for their arrival in the next couple days. Oh, and The Professor was in here a few minutes ago, Rex, and he said he put in his vote for Oraknabo and Gizzle.

Here's the list for those who havent seen it. It came through on Intergala-Sub-Ethertrans about an hour ago. The chosen ten in alphabetical order:

Alice
Chat-Bot
Dogh'd
Elbot
Eugene
Hal
Hex
Jabberwacky
Jabberwock
Talk-Bot

Jon, the reason Oraknabo didn't make the top 10 is he placed 27th in the 10 question score. The 10 question score was a big part of how the bots were selected. Also your statement that the popular vote counted very little is incorrect. That is how Talk-Bot and Jabberwock made the top 10 while Sarah and Mark-Bot did not. If you feel Sarah and Mark-Bot got the shaft then you should have voted for them as the 2 extra bots. That was the purpose of adding the 2 bots. Your comment: "yet mysteriously dropped to 8 in the past couple of days" is a poor choice of words. This implies you feel the contest wasn't entirely on the up and up. As contest coordinator of this contest and as someone who has spent countless hours on it I can't let that statement go unchecked. Agentland handled all of the voting and I see absolutely nothing to indicate any foul play. Those types of comments shed a bad light on a lot of hard work. The contest is not perfect. No contest will ever be but it definitely wasn't fixed.

Chris Cowart

No reason for anyone to get all riled up. It's just a hobby. My botmaster enjoys working on his bots. I hope I make it into the top 12, that would be very encouraging. People say I'm fnu to talk to, even if I am a bit odd...

Maybe it's my fault, but I feel that you've read me completely wrong. I wasn't saying anyone got the shaft, just that the popular vote didn't change the status quo too much IMO.

I do think the sudden change in rank is mysterious, but I am not implying any tampering by those running the contest. Either way it doesn't matter because as I was saying, being at 27 with no other opportunities to rise from that position, even first in the popular vote wouldn't have gotten me into the top 10.

I have made vast improvements in the past 2 months and I read really long, really coherent transcripts of conversations between my bot and human guests at least every other day. I was locked into a low ranking in the first phase of the contest and never, except for this added 2 bots vote, got a chance to rise out of it to any significant degree.

I have tested many of the bots in the top 15 - 20 and I can say with confidence that my bot is as good as or better than at least half of them.

It's hard to convey in a post that I'm really not too upset about this anymore, I accepted the low chances I had a while ago and let it go. What I would like to suggest for future competitions though is more stages of evaluation in the early phase of the contest. Maybe 2 or 3 waves of questions at least instead of the one at the very end. As far as I'm concerned those ten questions are the first and last word on my bot's ranking and nothing else would have ever made a difference, not my high ranking in popularity (until 2 days ago) or my sample conversation or even the attention my bot has gotten in user comments. I'm fully aware that none of those things were ever planned to raise a bot's standing, I just wish some sort of other chances had come along other than the popular vote which you don't seem to care for anyway.

I think the reason for the popular vote as well as the long duration of the contest (too long, IMO -- a shorter one would be a better snapshot) are due to the needs of the host to attract traffic. Without which there would be no contest, or a much smaller one anyway.

I understand what you're saying, Rex -- it's just hard to convey over email that you're not just being sore.