Linked by Thom Holwerda on Sun 14th Oct 2007 15:12 UTC
This week's 'big' news on OSNews was about software patents. You know, those things that say you cannot stack four pixels on top of one another unless you pay money to the guy who invented four-pixel-stacks (or the guy who bought the guy who invented four-pixel-stacks). A company called IP Innovation, LLC, has sued Novell and Red Hat for infringement of the company's IP portfolio. Software patents are of course generally completely ridiculous, so I will not focus on that here. I want to focus on something else.
Permalink for comment 278303
To read all comments associated with this story, please click here.
To read all comments associated with this story, please click here.
Member since:
2006-04-20
To use your coin toss example, it could very well be a coincidence. A 1/8 chance of something happening is by no means improbable. A lot of people have very funny ideas about statistics that don't add up in the final analysis. Because the probability of one coin toss does not predict or affect the outcome of another coin toss, some seemingly very improbable things are actually not as improbable as they seem. For example, I could toss three coins and have them all land heads up three times in a row. That would not be terribly amazing. Now, if I tossed those coins two hundred times and they consistently landed heads up more than chance predicts, then you might say there was something odd about those coins or the person tossing them, but even then, unless you had evidence that there was something fishy going on, the laws of probability allow for random chance to produce the same coin toss results ad infinitum, because one coin toss is unrelated to another.
You cannot aggregate the probabilities of isolated events unless there is a causal relationship between them. A whole bunch of improbable things can occur by chance, and it is only when there is a common thread binding them together that you can make a linkage between them, and say it is more than just chance.
Probability alone is not sufficient to rule something out (or in). Winning the lotto jackpot is very improbable for any given ticket buyer (usually something like 1/50,000,000), yet enough people buy enough tickets that it is a near certainty that one or more of them will win lotto.
Now, in the case at hand, there is insufficient evidence to draw linkages between the isolated points and to start saying there is something sinister and underhanded going on. Thom has analysed the issue correctly IMO.