Text in the article refers to a paper co-authored by William Landes of the University of Chicago Law School:
Several studies have looked at the question, including one by USC law professor Lee Epstein, William M. Landes of the University of Chicago and Judge Richard A. Posner of the U.S. 7th Circuit Court of Appeals, based on statistical analysis of Supreme Court oral arguments from 2004 through 2007.
Their conclusion: "The number of questions and the total words in question ... provide a reasonable predictor of most Justices' votes," the clearest exception being Justice Clarence Thomas who, alone among his colleagues, almost never speaks during the court's arguments.
The least predictable member of the court in their study? That would be Justice Anthony M. Kennedy, who is often the swing vote in the court's closely divided cases and was the justice most closely watched in the healthcare arguments.
The title of the paper is INFERRING THE WINNING PARTY IN THE SUPREME COURT FROM THE PATTERN OF QUESTIONING AT ORAL ARGUMENT and the first paragraph states
Chief Justice John Roberts, and others, have noticed that the lawyer in an oral argument in the Supreme Court who is asked more questions than his opponent is likely to lose.2 Our main purpose in this paper is to provide empirical tests of that hypothesis and the related hypothesis that the number of words asked, as distinct from the number of questions, also predicts the outcome. We also explore the theoretical basis for these hypotheses, and we begin with a brief discussion of the theoretical issue.
Footnote 2 of the paper quotes CJ Roberts: “the secret to successful advocacy is simply to get the Court to ask your opponent more questions." John G. Roberts, Jr., "Oral Advocacy and Re-emergence of a Supreme Court Bar," 30 Journal of Supreme Court History 68 (2005).
The paper speaks of a "realistic theory":
The realistic theory is that judges usually make up their mind before oral argument. Indeed, in the case of judges, such as the Supreme Court Justices, who have discretion to decide which cases to hear, their minds may be made up when they decide whether to vote to hear the case. They use oral argument to try to persuade the other judges, and this implies asking more questions of the lawyer for the party they plan to vote against in order to punch holes in the lawyer’s case and perhaps prevent him from articulating his best arguments. This tactic is especially important because judges usually do not discuss the cases before argument, although, in the case of Supreme Court Justices, their votes on whether to grant certiorari will often indicate their leanings. Moreover, at their post-argument conference, at which they discuss the case and vote on the outcome, they speak (and usually, in speaking, indicate their vote) in a prescribed order. In some courts, the judges speak in reverse order of seniority; in the Supreme Court, they speak in order of seniority. A judge high on the seniority ladder in the first type of court, or low on the ladder in the second type, has to fear that a majority will have in effect decided the case before he gets to speak, and this motivates him to speak his mind at the oral argument in the guise of questioning the lawyers, in order to communi- cate his views to the Justices who will vote ahead of him at the conference.