A Load of Bull

Dungeons and dragons in 8bit!

2005-10-19T19:17:00.000+10:00

I'd almost forgotten about this 8-bit D&D flash. There is another one out there that uses the same audio, but has different video where the characters are from various 1st-person shootum ups (I think). If anyone knows where that one is, add the link as a comment here!

Dark Dungeons

2005-07-04T21:09:00.000+10:00

Dark Dungeons, by Jack T. Chick is a huge laugh. He seems to be one of those bible bashers that believe that roleplaying is inherently evil. Nothing wrong with being a bible basher, but roleplaying = evil is
not just narrow minded - it's plain wrong. I don't have any figures or proof, but I reckon roleplaying would have saved a lot of lonely teenager's lives because it's a good way to get social and chase that angst away.
Anyway, if you want a laugh, or want to see just how evil roleplaying really is, check this out.

Lightning Bolt! Lightning Bolt!

2005-07-03T17:39:00.000+10:00

The classic Lightning Bolt video is a cack! It's also known as Ogre Battle. It's a bunch of geeks dressed in ye olde costumes pretending to beat up an ogre. One of the wizards of the group must be really powerful, because he can cast about 3 lightning bolt spells every second, and looks to be able to cast hundreds of them per day too. I bet you're impressed! The funny thing is, his lightning bolts don't look like lightning bolts at all.... wierd. There are even remixed versions of the video out there, done to various musical scores and with added special effects. Just go to Google and you'll be able to find the remixes from there.

Another, called Cauldron is also worth a look at if you liked Lightning Bolt. Cauldron is 15 minutes long though, and not as funny.

I also came across another Live Action Roleplaying (LARP) video recently, called Lightning Bolt, the Motion Picture. Again, it's not as funny as Lightning Bolt, but it still is good to laugh at. I like laughing at sad displays like this. I've also done it in real life too, and it's as sad as it looks.

If anybody knows of links to other LARP videos that are funny, even though they weren't intended to be, let me know! I'd love to check them out.

There are some other really good roleplaying game (RPG) parody videos and web sites out there that are worth a look at. Most of these are funny because they are meant to be, but some are funny because they are just sad. I'll post some links to these in my next entry!

Kiwiflew

2005-06-27T20:48:00.000+10:00

I got my first link today from pretty Japanese girl. She's also my number one commenter! In appreciation, I've added her blog, Kiwiflew, on my blogroll! Check it out.

Bull.doc

2005-06-25T19:50:00.000+10:00

Which File Extension are You?

How boring. I'm a .doc file extension. I would have expected something more exotic. Of course the actual questions have nothing to do with the final analysis. I demand a retrial!

Christmas Wish

2005-06-23T17:27:00.000+10:00

Ram-A-Lam-A-Ding-Dong

2005-06-23T17:26:00.000+10:00

Invaded

2005-06-21T20:51:00.000+10:00

I might branch out and use some other office stationary next. Like those phone call message pads or maybe a napkin from the work cafeteria.

Surreal Banana Peel!

2005-06-21T20:49:00.000+10:00

I told this joke to a mate of mine whose wife is an artist. He laughed for several days straight. True story.

It's life Jim, but not as we know it.

2005-06-21T20:48:00.000+10:00

Did you know that Pacman originally came from Japan, where it was called Puckman. When U.S.A. decided to release it there, they were afraid that "Puckman" is too easy to modify into a rude name. Hence "Pacman" was born.

Based on a true story

2005-06-21T20:45:00.000+10:00

Here's the second. As you can see, my real name is Robert Winkel, but everyone calls me Bull in real life. I put the copyright notice on these because of the irony - nobody would try to steal these comics I'm sure!

Phone Home

2005-06-21T20:43:00.000+10:00

Here is the first of a series of cartoons that I am calling "Wokka Wokka", for now. They are a tribute to Doug Savage's Savage Chickens. As you can see, the inspiration came directly from Savage Chickens. My aim wasn't to create anything new or imaginative here, but just to have some fun and share it.

Terry Tourette

2005-06-19T16:19:00.000+10:00

Here are some comics that I drew a few years back. As you can see, a lot of thought and effort went into them. I thought that these artworks are too good to keep to myself and have therefore decided to share them with the rest of the world.
Enjoy...

Using linear programming for game playing

2005-05-11T19:56:00.000+10:00

For my course in Optimization Techniques I have to propose my own assignment. All of my past proposed assignments have involved building a computer opponent for some game. So I wanted to do the same thing for this course.

The Optimization Techniques course is really just an intro to linear programming. I've trawled the internet, but all I could find was pages on "game theory", which discuss, with very simple examples, methods of using linear programming to solve a specific type of game - one where both players move simultaneously, each player knows the "payoff" for each move, and the payoff gain for one player is the payoff loss for another. In this case, linear programming is used to assign the best probability distribution for each move.

What I wanted to use it for was for games like tic-tac-toe, Connect 4 or Othello, where players take turns to move. I didn't even know this was possible. After trawling the net I still didn't know, so I asked on news:sci.op-research, where I got some help!

If you are interested, check out this pageweb: http://ite.pubs.informs.org/Vol3No3/ChlondAkyol/index.php. It describes how integer programming (a subset of linear programming) can be used to play Nim. I will read this paper through thoroughly a few times, but from my skim-reading it looks like something I can use. Yay!

THE END

Next challenge?

2005-05-02T20:47:00.000+10:00

I always like to keep my mind active, whether it's reading books or downloaded a thesis, or setting myself a new challenge. Often, I set myself a new challenge, do it, feel good about it, and then don't do it any more. That doesn't really bother me. I like to learn, and continually test myself with new things.
The next thing I intend to try is fantasy art drawing. I don't know exactly when I'll start it, or if I'll be any good, but once I start, I know I'll give it a good go. When that happens, I'll scan various stuff in, and you can see my work. If I get really good at it, I hope to be able to fairly quickly sketch any roleplaying character that I play.
Time will tell.

A Picture is Worth 1000 Words

2005-05-01T10:59:00.000+10:00

In the past I got into playing with Photoshop. I found a pretty cool website that runs competitions daily to test your Photoshopping skills (http://www.worth1000.com). So I tried it out for a few weeks, won a few awards, and that is the end of that story.

Here are some of my better attempts. You can click on the image for a better view. The first one is called "Atlas Skull" (you will have to click on the image to see it properly).

The second one is called "Rose Bulb". I recently received an email from an artist that wished to turn this image into an oil painting. I asked for a photo of the finished product, so I'll post that here if I actually get it.

This last one is the most popular one that I have done. It is called "Egg Vend". It is a crowd favourite, but not my favourite, simply because it has a few technical mistakes in it. It has won several awards, gotten "Image of the Day" on the Worth1000 website, and I have even seen it in a few random "funny photos" websites.

Can a Computer Win a Footy Tipping Contest?

2005-04-27T19:30:00.000+10:00

That's what I intend to find out!
Every year, there's a few footy (Australian Football) tipping contests at work. I intend to evolved a neural network, from a single neuron, that will reach around 75% accuracy at predicting the winners. Once tested, I'll enter it into the competition. I don't consider this cheating, because I will let everybody know about it, and if it wins, I'll donate any money back to the social club.
It's been done before, with only training the neural network to pick a winner based on the two teams previous N (3 I think) games, and whether or not they have the home game advantage. This neural network acheived a 72% accuracy, on average. And this is without knowing which teams are playing (i.e. Hawthorn vs. Geelong), whether it's raining or not, and whether key players have been injured. I intend to include subjective information like this!
How? Well, there are plenty of regular footy tipsters out there to gain data from (e.g. "experts" from papers, sports betting agencies, etc.). These tipsters are all subjective to some degree. The sports agencies have many experts with their ear to the ground! Therefore, I intend to use some of this data as inputs to my neural network. This way, the neural network can learn which tipsters to trust under certain circumstances, and how much to trust them. Therefore I don't need to worry about a certain key player being out for 3 games because of a drug scandle or the like, I let this information be captured by the expert tipsters tips, and then I input these tips into my neural network.
It will be interesting to see how much the neural network relies on objective information and how much it relies on subjective information. If the neural network ends up doing any good, I'll write a paper on my findings.
The next step would be to evolve another neural network to do something similar. This time, it wont be designed to pick the winner. It will be designed to get the maximum amount of profit based on the odds given by online sports betting agencies. The neural network will decide whether or not a game is worth betting on, and how much should be bet on it. I could do something basic, like start the season with $100, and put any winnings back into the bets. Hopefully, if it works nicely enough, with the componded winnings, I'll be a MILLIONAIRE!!! BWAHAHAHAHAH!!!! If not, then I'll just be a $100 short, but have had a bit of fun in the meantime. :)

THE END

Evolving blind tic-tac-toe players

2005-04-26T20:21:00.000+10:00

I'm studying a Masters of Electrical Engineering at the moment, and have just finished a Masters of Information Technology. I like to study, especially when it's not me that's paying for it (my work is!)
I like doing any subject that has some sort of Artificial Intelligence in it, as I get to experiment with various ideas I've had while laying wide-away at 2am in bed. So far, my passion has been using A.I. for game playing.
I wrote a Connect-4 playing computer opponent, using neural networks and evolutionary algorithms. It beats me every time!
I also wrote a "blind" tic-tac-toe neural network, also grown from evolutionary algorithms. The tricky thing with "blind" tic-tac-toe is that you have to remember the previous moves, because you are not allowed to look at the board. This may not sound like a hard thing to program, but the idea was to experiment with evolutionary algorithms to see whether they could successfully evolve a neural network that had memory (in the form of recurrent links). It worked fine, and it plays an OK game. Not as good as a human, but the experiment was still a success. Interestingly enough, the "blind" tic-tac-toe neural network that I evolved played a better game than "seeing" tic-tac-toe neural networks that were evolved over the same number of generations.
If you're interested in why, my theory is that initially, the topology of the neural networks are minimal, and for the “seeing” neural networks, they have a 9 inputs at various states at any moment in time, where-as the “blind” neural networks effectively only have one of the inputs active at any one time (representing the opponents last move). This could be a form of information overload for the “seeing” neural networks during the early generations. It would seem that the “handicap” of limited memory for the early generation “blind” neural networks outweighs the handicap of information overload for the early generation “seeing” neural networks.
If you're not interested, then why have you read this far? :P
Anyway, I'm willing to talk about it more, just leave some comments. Or I could post the papers here.