Dr. Dobb's Journal June, 2005
OAF Again
Dear DDJ,
I have been looking at John Trono's OAF algorithm ("Applying the Overtake & Feedback Algorithm," DDJ, February 2004) with great interest. This is a fantastic piece of work, which I am very keen to understand and hopefully apply, and was wondering if John could please explain the following sentence with regards to working out the "Distance" factor a bit further: "...where distance is the largest, integer multiple of SQRTN less than or equal to the separation between the two team's rankings, which in this case is floor(57/11) or 5." For example, if there are 20 teams in my league and the two teams involved are ranked 3 and 12, what would the distance be? I live in Manchester, England, and was wondering whether I would be able to use this method to create some soccer ratings? The winning score margins for each game are going to be a lot lower than for American football, the most frequent being just 1 or 2. Should I multiply these scores by another constant to fit them into the algorithm? Also what do you think should happen to each teams rating if a Draw (Tie) result occurs?
Ian Broughton
tracy.jones63@ntlworld.com
John responds: Ian, thank you for your interest in (and kind words about) the OAF algorithm. Though I left out how ties games are handled in the brief summary that is on my web page (http://academics .smcvt.edu/jtrono/OAF.html) (since they no longer are a possibility in NCAA football), such an outcome is treated almost like a loss. When two teams' ratings are averaged in OAF, the victor gets the average +0.4 and the loser the average -0.4 as their updated ratings. When a tie occurs, the team that had the higher rating before the tie game has 0.4 added to the average, and the other team receives the average -0.4 as its new rating.
If there are 20 teams, then the integer used for √N is the floor of the rounded result of √(20), which in this case is 4. If a team ranked as #3 loses to a team that is ranked #12, then the modified update rule would compute that team #12 is (12-3)/4=2 "intervals" outside of the region where the pure OAF update would be applied and so the denominator of 4 would be used instead of 2, when computing the "average" of the two opponents' ratings. (Teams 4-6 would use a denominator of 2, i.e., the true mean, 7-10 would use a denominator of 3, 11-14 would use 4, etc.) I hope this clears that up for you. Please go back and reread the last three paragraphs on that web page to see if you follow what I describe here. And remember, there is the caveat about using the larger update of the two that are computed, as mentioned in the penultimate paragraph on my web page that you quote from below.
I would be curious to hear about your results when applying OAF to the soccer games in England. I don't think you would have to do anything special to adjust the scores because of the smaller differentials; you might just end up with ratings that are closer, that's all. (Since touchdowns in football are like goals in soccer, you might try multiplying all your scores by 7, but again, I am not sure if that would be necessary.)
Good luck; don't hesitate to ask for clarification if my description above is unclear, and keep me posted about how your study goes when applying this to soccer.
Silent Update
Dear DDJ,
I appreciate the skill that went into Zuoliu Ding's article "A Silent Component Update for Internet Explorer (DDJ, April 2005). As a user, I ask, how do I prevent a software company or malicious agent from silently updating the software on my hard drive? Will a firewall like ZoneAlarm do?
Bill Weitze
bweitze@california.com
Zuoliu responds: Bill, thanks for your interest in my article. Regarding your question, I would like to say there is nearly no way to prevent such an update with a third-party product. This is because such a silent update is triggered by the specific software that has been installed in your local machine based on your trust already. The download and update is guided by the software resided. If a hacker hijacked such a company download site and knew the update mechanism, your computer would be vulnerable and defenseless. For some general invasion, you may take [precautions] like blocking some ActiveX downloads from the prompts by XP/SP2. But for such a silent update, the individual company should take full responsibility in designing secure updates, such as encoding download information, dynamizing update sites, and so on. Hope this helps.
Is Wind Power Hot Air?
Dear DDJ,
As a part of Jonathan Erickson's "looking back" perspective in the February 2005 issue of DDJ, he effused about the amount of energy being created via wind-generated electricity. However, nuclear energy is an idea whose time has come and gone—and is come again with a vengeance. Simply put, it is the cleanest, safest, best way to create energy for people on this Earth. Wind power doesn't even come close to fitting the needs of this world. Note that a "wind farm" would require about 300 square miles of space to produce the equivalent output of one nuclear plant. The wind turbines slice-and-dice birds with a terrible regularity.
Don't take my word for it. China researched it and is now committed to building a minimum of 25 new reactors in the next decade. All completely fail-safe. The design is called "Pebble Bed Reactors." You can fire these puppies up and walk away from them forever. They won't blow up and they won't melt down. I refer you to the February 2005 issue of Wired magazine, which covered this issue in some detail. The biggest threat to the reintroduction of nuclear power to America is, of course, the Luddites—people who believe that "nuclear is bad." Again, Wired covers these types pretty well. And so I ask Jonathan to reconsider the scientific evidence that bodes extremely well for nuclear power, and drop the hippie-cult thing with "wind power." In the end, "Wind Power" is just a bunch of hot air.
Peter Andre
peter.seattle@gmail.com
Jonathan responds: Thanks for your note, Peter. Did I say that windpower ought to replace nuclear power? I don't recall doing so. Should there be alternatives to coal-based power plants? Yes. Coal-based plants screw up the environment when you dig the coal and when you burn it. I don't have a problem with nuclear power. But I also think that places rich in wind (Washington DC doesn't count) can harvest that energy and provide an economic boon to the local economy. This includes, say, western Kansas, the Dakotas, and other such windblown areas. It probably isn't feasible to build nuclear plants in western Kansas, but wind turbines do make sense.
DDJ