March 13, 2017
By Marc Tracy
Starting in the 1970s, eight prosecutors in the Justice Department’s Public Integrity Section set up a standing meeting. Once a year, they would gather in the deputy attorney general’s conference room and run an elaborate pool for the N.C.A.A. men’s basketball tournament.
“We mostly followed the law,” said Reid Weingarten, one of the pool’s charter members.
The pool lasted decades, surviving long after the colleagues had left for other jobs. Like most N.C.A.A. tournament pools, it brought friends together for good old-fashioned fun. Any financial stakes paled in comparison to the competitive ones. Participants got carried away with preselection research. And they were merciless in lambasting one another’s pet predilections.
“Eric” — that would be Eric H. Holder Jr., the future attorney general — “would always pick U.C.L.A., because he was a Kareem guy,” Weingarten said, referring to Kareem Abdul-Jabbar’s alma mater. “And he would always pick the Ivy League team, because he went to Columbia.”
But what made the bracket truly unusual was its structure. The participants did not pick games; rather, they drafted teams. (Random selection dictated the drafting order.) Owning the team that won it all merited the largest reward, of course, but it also was worth something to have a team that made the Final Four or the lowest-seeded team to make the round of 16.
This was not, in other words, your ordinary office pool.
“I find blind pool picking tedious, ridiculous and not fun,” Weingarten said.
That is not to say there is anything wrong with simple pools; millions of Americans take part in one every March, often merely for the prize of office or family bragging rights. In fact, one recently existed in a place even higher than the Justice Department. For the previous six N.C.A.A. tournaments, Kyle Lierman, then of the White House Office of Public Engagement, ran BOTUS — Bracket of the United States — for hundreds of members of President Barack Obama’s staff. (He will run the pool again this year; it has been rechristened BOTUS 44.)
There were no financial stakes, Lierman said, and the president’s bracket, publicly filled out on ESPN, was automatically entered every year. “He did well a couple years,” Lierman said. “Other years, not so well.”
The N.C.A.A. tournament, which kicks off Tuesday, is the Super Bowl of sports wagering. Actually, it is the Super Bowl twice over. The American Gaming Association, a trade group for the casino industry, estimated that Americans will wager more than $10 billion on the tournament this year, most of it illegally, compared with $4.7 billion on the Super Bowl. And that figure does not include the multitudes of office pools, with their $5 or $10 entry fees. Officially, the N.C.A.A. opposes “all forms of legal and illegal sports wagering.”
“Despite the N.C.A.A.’s protests against gambling,” said Geoff Freeman, the chief executive of the gaming association, which supports the broad legalization of sports wagering, “they’ve created the biggest gambling event in American history. They wisely did it with great marketing, tremendous use of the bracket and infiltration into the American cultural lifestyle.”
Most pools are straightforward: Take a blank bracket, predict winners for every game and submit it before the first one tips off. But for a certain breed of fan, like those aforementioned anticorruption lawyers, the standard one-size-fits-all pool is not enough. Several have devised sophisticated, even byzantine alterations on the theme.
Noah Chestnut, a product manager at the sports website Bleacher Report, crafted a draft-style pool several years ago that is similar to the one used by the lawyers. In his pool, drafting is auction-style. Entrants are allocated a $100 budget that must be spent on four to six teams; whoever offers the most for a given team buys it. (When Kentucky entered the 2015 tournament undefeated, the successful draftee paid $96 for the privilege of selecting the Wildcats, who lost in the semifinals.)
“I’ve always preferred to draft to everything else in life,” Chestnut said. “I don’t actually enjoy the games as much as the time before the games.”
A common tweak in irregular pools is to adjust how many points a participant receives for picking an underdog in order to make every game closer to an effective tossup. Picking a 14th seed that wins its first game, some reason, should be worth more than picking a No. 3 seed that holds serve.
Chestnut, for example, weighted the scoring so that additional points were available when a team upset a favorite: 3 points, for instance, when a No. 10 beat a No. 7. And The Wall Street Journal reported several years ago on a biostatistician who used sophisticated mathematical formulas to determine point values for every matchup based on how previous, similar matchups had turned out.
Similarly, in the early 1980s, Stephen Brown raided the microfiche at the University of Virginia where he was a graduate student. Armed with pencil and paper, he consulted past bracket results from The New York Times and, based on previous outcomes, calculated upset bonus points, which increased as the bracket narrowed.
“I was supposed to be writing my dissertation on an 18th-century poet named Edward Young,” said Brown, now an English professor at Rhode Island College.
The bonus points, though, were secondary to Brown’s core innovation: round-by-round picking. Instead of filling out an entire bracket before the first game, participants picked games in the second round after knowing the results in the first, and so on. The structure keeps more participants in the running, as anyone with a bad first round, say, can make up for it with a good second round — particularly if there are more upsets, and thus more available upset bonus points.
“I never liked the idea of filling out the brackets,” he said, adding, “It makes so much more sense, and would take more judgment and skill, if you knew who was actually playing every game.”
Brown’s pool survives to this day, administered by his nephew. So does a decades-old offshoot centered in the Washington area. (Full disclosure: The offshoot is this reporter’s annual pool of choice.)
Finally, for those who prefer getting ready for the party to the party itself, Bryan Cimorelli’s pool, the Conference Tournament Challenge, is for you. It covers not the one N.C.A.A. championship tournament, but the 32 Division I conference tournaments, which this season comprised 294 games, per his calculation.
Each tournament is treated like the N.C.A.A. tournament, with point values increasing each round. Additional points are awarded for upsets, calculated by a method too exasperating to explain here. Values are additionally multiplied depending on how prominent the conference is: Conferences like the Southwestern Athletic are the baseline, while points in the Ivy League are multiplied by 3.5; the Atlantic 10 by 4.5; and the six power conferences by 5.
“I like things that are really complicated and take a lot of thought,” said Cimorelli, perhaps stating the obvious.
Cimorelli, whose day job in budgeting also involves spreadsheets, emails a preview every morning and a recap every night to ensure his pool’s 40 participants remain engaged. A recent preview was more than 15,000 words. It informed competitors that only one had picked third-seeded Richmond to win the Atlantic 10, discussed coaching seniority in the Southeastern Conference and made the observation that while more than half the pool had picked Rice to upset Texas-El Paso in one game, none had the Owls reaching the Conference USA final.
On Sunday, the C.T.C. ended and, soon after, the selection committee handed down the N.C.A.A. tournament bracket. For Cimorelli, it always is perhaps the year’s biggest letdown.
“I don’t even want to look at a bracket,” he said. “I fill one out because it’s my sports fan’s obligation, but after 300 games in 13 days, I get to the actual tournament and I’m like, This is boring.”