While I understand that we are taught how to hedge our bets with statements in law school, the statement “This idea [,] however, might have some imperfections.”, should read “This idea has imperfections.”

Why the certainty?

Very simply even taking the drastic measure of not looking at the case law, one can see that the lawyers not only misapply statistics, but also use classical logical fallacies to support their “argument”. While both are common features in lawyers, both are somewhat inexcusable if one believes that one should not lie to the court (But see Zacharias, Fred C.,Fitting Lying to the Court into the Central Moral Tradition of Lawyering(June, 23 2008). Case Western Law Review, Vol. 58, June/July 2008
Available at SSRN: http://ssrn.com/abstract=1150450).

As to the statistics, it seems that the little book that has been around 40 years, How to Lie With Statistics, still should be relevant to [these] lawyers (if not all professionals). Social statistics are products of our social arrangements, despite their power to summarize.
First, there is an obvious conscious bias where there is a selection of favorable data and a suppression of unfavorable data – where is the statistical methods in the lawyers argument that mitigates bias? Simply citing Google is not enough.
Second, (even though the statistics seem to have the Google brand name on them) while the data came from Google, the conclusions are entirely the lawyers own. The real question is under what circumstances Google stands behind the data.
Third, the sample that was selected is biased. For instance selecting Google as the means inherently limits the data set to certain types and classes of actors or expands the field too much (under and over sampling). This is much like the Market problem in Antitrust. Let us not even go into representativeness.
Fourth, the numbers that are presented inherently lose meaning because intentionally there is no means of comparison to other numbers. Getting any meaning from this depends on your knowing something about the rankings of searches outside of the selected searches. Few of us actually know this data, but it is relevant. The absence of other figures take away most of the meaning. These are just a few concerns.
Next, the lawyers did not have to know statistics to avoid these misses of the truth. There are obvious logical fallacies (which are avoided in good statistics). Take the fact that the argument draws a conclusion that is stated more strongly than the evidence actually supports (Inductive Hyperbole). Or, the argument draws a conclusion from observations that were obtained in a manner that would be likely to tilt or corrupt the results (Tainted Data). Or, one event (community values) is cited as the cause of another (ranking), but, while there may actually be a connection between the two events, the hypothesis mis-locates it, either making the effect into the cause, or treating as cause and effect two events that are independent results of a common cause, like apple pie and XYZ (Non Causa Pro Causa). Or, the argument misrepresents a position that it seeks to refute, what community standards are and instead puts Google as the proxy for such standards; y refuting the position as misrepresented, the argument creates the impression that it has refuted the position that is actually held by opponents (Straw Man). Or, the argument misrepresents the consequences of choices that are available when making a decision, or fails to present all the choices available, such as the internet is used for the avoidance of community standards (False Dilemma). Or, the argument draws a conclusion from observed cases that are only superficially or apparently similar to the unobserved cases about which the conclusion is being drawn (False Analogy). Or, the argument draws a conclusion based on a comparison between two (or more) groups, even though some important difference between the groups, other than the difference specified by the experiment, may be responsible for the results obtained (Uncontrolled Factors). Or, the argument depends upon an ambiguity in the meaning of a word, community values (Equivocation). Or, the argument moves from a claim about the distributive sense of a class (i.e. each of the parts taken separately) to a claim about the collective sense of a class (i.e. the class taken as a whole)(Distributive Fallacy – Composition).

William Love, J.D.