Can you believe it, some computer scientists at the University of Alberta have worked out how to always win at Checkers (known in the UK as Draughts). To prove it they have created a Java applet which can play you and will always win.
Checkers has a search space of 5×10^20 (ie 5 followed by 20 0s) i.e. 500,000,000,000,000,000,000 possible moves. Compared to chess of course that is miniscule but it is a million times larger than Connect four.
This was a project that started in 1989 and an earlier version of the program won the World Checkers competition in 1994. Though it is a fairly trivial task- the research itself will have many uses. Other games that may get ’solved’ in the future will include Othello (also known as Reversi) and Poker.
Popularity: 5% [?]
Phi is a number approximately 1.61810339887… Not so well known as Pi or e, yet Phi is a very popular number. You can determine its rough value from the equation 1/phi = phi-1 which is almost magical. Ie 1/1.6181033 = 0.6181033.
Phi has other properties as well. It is the ideal ratio for width to height- to get the most pleasing view. It is known as the golden ratio and is found frequently in nature. This site will probably tell you more than you ever want to know about it! Some very nice illustrations as well.
Popularity: 3% [?]
One of the problems teaching statistics is that it can be a very boring subject. Yet there are probably more people who would benefit from knowing a bit more about probability- eg Poker players.
There can be other spin offs as well. A few years back I had to write a program to scan a rival company’s website to see how many houses they had for sale on it. They said 80,000 but we didn’t believe them. They had a feature where you could search by postcode district and it told you how many there were in that district. Now there are 2,800 such districts in the UK so I had the program do the queries randomly and slowly so it wouldn’t attract attention. As each district was done, it updated a count of all the properties and an average number of properties per district. Multiplying this by 2,800 gave an estimate of how many properties there were in total. After 40 or 50 the estimated figure had settled around 40,000 and stayed there. This is sampling theory in practice. There was no need to read all 2,800 districts - just 40 or 50.
Exploring data is a ten year old Australian website that tries to illustrate statistics in a more interesting and involving way, using real data and real problems. It contains
activities, worksheets, overhead transparency masters,
datasets and assessment to support data exploration plus an extensive collection of articles.If you don’t know the difference between a stem plot or a dot plot this is the place to look. There is a lot here.
Popularity: 9% [?]