Lawyers for the Department of Justice and for the Penguin Random House delivered their closing argument on Friday in a case that will determine whether the publisher, the country’s largest, can buy one of its rivals, Simon & Schuster.

The case, which will be decided in the fall by Judge Florence Y. Pan, focused on the effects of consolidation on publishing, an industry that has already been dramatically reshaped by mergers in recent years.

The government sued to stop the deal on antitrust grounds, saying it would diminish competition among the biggest houses and push down advances for some authors. Bertelsmann, the parent company of Penguin Random House, argued the deal would bring its supply chain and distribution muscle to a longer list of authors, to their benefit, and that the industry is large and varied, made up of many important players beyond the biggest firms. Judge Pan expressed skepticism over several of Penguin Random House’s main arguments.

Beyond the legal debate, the three-week trial offered an unusual glimpse into the world of publishing, offering observers a parade of high-profile publishing executives, agents and authors speaking frankly and on the record about how books are made.

Here is some of what we learned.

What makes a best seller?

No one — not even the top executives in publishing — has figured out the mix of factors that can guarantee a best seller.

Markus Dohle, the head of Penguin Random House, said everything in the book business is “random.” The judge described the industry as one of “risk-taking” and said every book was “a gamble.” The publisher of a major imprint called the business “fickle.”

The difficulty of predicting success was a consistent theme during the trial. It sets the publishing world apart from many industries, which can use the past sales of similar products — like a type of washing machine or cereal — to predict future sales. But every book is unique, so while educated guesses can be made, especially concerning established authors, nobody really knows what the future holds for any given project.

“Everything is random in publishing,” Mr. Dohle said on the stand, recounting the company’s origin story. “That’s why we have that name. So the founder thought: Everything is random. Success is random. Best sellers are random. So that’s why we are the Random House.”

How do publishers decide how much to pay an author?

How much a publishing house offers an author for their book is usually a closely guarded secret — a number based roughly on sales expectations, competition for a project and very subjective elements like how badly editors want it.

During the trial, several of these payments, which are called advances, were discussed with uncommon candor.

Jennifer Bergstrom, the publisher of the Simon & Schuster imprint Gallery, said they paid “millions” more for a book by the comedian Amy Schumer than the sales estimates would have dictated.

A few authors, like the actor Jamie Foxx and Jiayang Fan, a writer for The New Yorker, were discussed at the trial by name — both of them were offered seven figures. Other writers were discussed in code to preserve confidentiality.

“Author A is a reality TV star,” Ms. Bergstrom said. “She’s a housewife on Broadway — excuse me, Broadway? She wishes. Bravo.”

She continued: “Author B is the — was the star, he is now in prison, of a major Netflix show. We paid 175 for that.”

Penguin Random House executives testified that the approval of Madeline McIntosh, the chief executive of Penguin Random House U.S., is needed for advances of $1 million or higher, and Mr. Dohle’s for payments above $2 million. For advances of $75 million or more, Mr. Dohle needs approval from executives at Bertelsmann — but he’s never had occasion to ask, he said.

What books drive the industry’s profits?

By most measures, publishing is thriving. In any given year, hundreds of publishers in the United States release around 60,000 books. From 2012 to 2019, print book sales grew by more than 20 percent, from nine billion to 11 billion, Mr. Dohle testified. And book sales were strong during the pandemic, rising by another 20 percent from 2019 to 2021, he said.

But the trial highlighted a surprising fact: A minuscule percentage of books generate the vast majority of profits.

During their testimony, Penguin Random House executives said that just 35 percent of books the company publishes are profitable. Among the titles that make money, a very small sliver — just 4 percent — account for 60 percent of those profits.

“That’s how risky our business is,” Mr. Dohle said. “It’s the books that you don’t pay a lot for and become runaway best sellers.”

The trial offered examples of books that publishers paid a relatively small amount for, and that turned out to be a great hit. Sally Kim, the publisher of the Penguin Random House imprint Putnam, said they acquired “Where the Crawdads Sing” for “mid-six figures.” That book has gone on to sell around 15 million copies worldwide.

The issue is industrywide, and has been exacerbated by the rise of online retail, which tends to reinforce the visibility of best sellers. In 2021, fewer than one percent of the 3.2 million titles that BookScan tracked sold more than 5,000 copies.

How does consolidation impact publishing, and authors?

The publishing industry has been transformed by consolidation in recent years. Big media conglomerates have gobbled up smaller companies, giving them huge catalogs and a competitive advantage over small and midsize publishers, who are typically shut out of auctions for blockbuster books by brand-name authors.

The repercussions of those changes were hotly debated during the trial.

Justice Department’s argument emphasized the growing influence of the so-called Big Five publishing houses: Macmillan, HarperCollins, Penguin Random House, Hachette and Simon & Schuster. In a pretrial brief, the Justice Department stated that the Big Five make up 90 percent of the market for “anticipated top-selling books.”

“They are the only firms with the capital, reputations, editorial capacity, marketing, publicity, sales, and distribution resources to regularly acquire anticipated top-selling books,” lawyers for the Justice Department wrote.

Bertelsmann’s attorneys presented a very different portrait, emphasizing that some midsize and independent publishers have been thriving and increasing their market share. During his testimony, Mr. Dohle noted that in 2021, roughly half of the books sold in the United States were from publishers outside of the Big Five.

Penguin Random House also made the case that reducing the number of major publishers would not result in authors losing money, because the remaining houses will be forced to compete even more aggressively for potential best-sellers and will offer bigger advances to counter the combined heft of Penguin Random House and Simon & Schuster. They also claimed in a pretrial brief that savings resulting from the merger will result in more money to spend on books, enabling the company to “offer higher advances” and “bid on more books from more authors.”

Judge Pan questioned those assertions. On Wednesday, she ruled that a significant piece of Penguin Random House’s argument — its evidence that combining the two companies would lead to cost savings and efficiencies that would ultimately benefit authors — was inadmissible because it had not been independently verified.

In an extended back and forth, Judge Pan also seemed skeptical of assertions by Penguin Random House’s economic expert witness, Dr. Edward Snyder, a professor at Yale, that reducing the number of major publishers would not result in lower advances. She said the publisher would not need to make an affirmative decision to lower advances, but that it would happen anyway because there would be less competition.

“I don’t see that there’s going to be a lessening of competition,” Mr. Snyder said.

Judge Pan replied: “That’s the ultimate question.”