Dave Morton - Nov 4, 2014:

What do you all think? It might take me a short while, but such a judging/admin script wouldn’t be difficult to put together.

Sounds really good.  A strick judging framework like you decribed makes for a more robust implementation of what I think Vince is valiantly trying to faciliate with the CBBRC (“to provide an informal environment within which botmasters can try their machines and themselves in the court of public opinion”).

All great suggestions guys. Dave I’d like to see an implementation of that. Email me when you get a chance. I was also thinking that you could eliminate a lot of the problems by automating the Interogs portion of the judging. Pandorabots already has an open API, all you need would be each bots ID.  SAIL has an API, and for most bots providing an API would be a simple matter, just create a basic page that would accept a POST from any location. Basically you are just looking at a QR pair. Still trying to get caught up, there will be more details on each bots judging.  The freestyle would still have to be a human. I agree that having multiple judges visit opens up too much of a question, what I was thinking is that you really do not have to have multiple visits, a single judge visits and posts the transcripts. Then multiple judges can view the transcripts and form their opinions.

Vince

Judging from a set of transcripts may not be a good idea, but it’s worth a try, I think. I already have a remote debugging tool for Program O that could be fairly easily converted to a multi-bot interface for use in judging. The only real problem would be with bots that don’t have a web interface, but that’s one area where judging from a single set of transcripts could solve several problems. That said, however, I don’t recall any of the chatbots in the most recent contest that didn’t have one. I’ll have to research that, just to be sure.

Email’s on the way, Vince.

This looks like fun, I have entered poppy-cat (http://www.talkigy.com/chat/poppy-cat).

She’s not a very competitive cat, but she does like bragging rights

I’m impressed with the way you have set this up.

Good deal David,

We are shooting for February for the next round. Does your service have an API by any chance?

Vince

Hi Vincent
I’d think it does, but I haven’t documented it or checked that it works as a standalone api.

You might be able to get a sense of the api by visiting http://www.talkigy.com/rsp

which returns json response such as {"tpc": "poppy-cat", "rsp": "Yaaaaaaaaawn, wake me with a new question on Poppy, the cat that I am", "nbr": 2124}

and to stimulate a response, you need to Post json such as {stm: are you a cat?}

where tpc is the topic, nbr is the user.

What would you like to use the API for?

David,

After the last round we have been talking about the idea of automating at least the Question\Statement—>Response section to eliminate some of the problems with copying and pasting.  We also might give Dave Mortons automated judging tool a whirl!

Vince

Yeah, still working out the logistics of that, Vince. but the project I was working on previously that was keeping the judging tool on the back burner is now completed, so I’m going to get to work on it tomorrow.

Just to let you know that the Bragging Rights prize money arrived. Cheers Vince.

Still waiting for the Robochat Challenge prize money though…