Data Mining
By Kris Lazaro (11/7/2006)
Definition
The term data mining refers to the searching of sports data over the years and isolating specific situations to bet on in the future.
Handicapping
Data Mining is indeed a part of handicapping. In looking for favorable situations, the astute handicapper must make sure that the same conditions exist for a particular angle. For example, in the National Football League, there is currently an obscure angle that is currently 23-0 as of the second week of October 2006—mainly, it states to fade teams that had previously won at home by 7 or more points AND had twice the amount of offensive yardage as compared to their opponent. Such obscurity can only really be analyzed by a computer program that has been fed statistics. In short, one cannot solely rely upon one angle in determining which side to take in any particular event. One must combine this with other factors to successfully arrive at a play.
Statistics
Statistics can be skewed; one can make a case for either side of a sporting event if one chooses to. For example basketball team A might be 23-4 against the spread (ATS) after scoring more than 100 points in their last game AND be 14-0 ATS off back-to-back home losses, BUT if their opponent is 25-6 ATS after holding their opposition to under 90 points AND 16-0 ATS in the second game of a homestand, THEN neither side should have an advantage. In this case, the two angles for team A will cancel out the two angles for team B. In general, if one side has more angles going for it than the other side, the prior side ‘should’ have an advantage. But, the author cautions this line of thinking, since one angle maybe more superior than another. For example, angle A might be more powerful and relevant than angles B, C, and D combined. Part of the thrill and enjoyment of sports betting is to isolate the best plays from all the games on the board, and hopefully find the right, winning side after the event has been played.
Bettor beware
Since the advent of the Internet, sports statistics has been more readily available to the general public. With this knowledge, oddsmakers have somewhat skewed their odds likewise. Although very hard to detect, some lines have been shaded than most. For example, the zig zag theory used to work perfectly in the NBA. But until recently, this angle has been shaded in favor of the sportsbooks.
Summary
In conclusion, data mining is the act of isolating specific situations in the past to bet on future games. |