QualityQueryParser qqParser = new SimpleQQParser("title","body");
QualityBenchmark qrun = new QualityBenchmark(qqs, qqParser, searcher, docNameField);
SubmissionReport submitLog = DEBUG ? new SubmissionReport(logger, "TestRun") : null;
qrun.setMaxResults(maxResults);
QualityStats stats[] = qrun.execute(judge, submitLog, logger);
// --------- verify by the way judgments were altered for this test:
// for some queries, depending on m = qnum % 8
// m==0: avg_precision and recall are hurt, by marking fake docs as relevant
// m==1: precision_at_n and avg_precision are hurt, by unmarking relevant docs