Baxter Revisited

Wagering on the outcome of National Football League games usually involves something called a point spread, or “line,” of X points. You have a choice of two ways to place your bet:

(C)

The stronger team will win by more than X. That's called “Covering the spread.”

(U)

The weaker team, known as the “Underdog,” will lose by fewer than X and might even win.

When placing their bets, gamblers pay their bookmakers a ten percent commission. It will be refunded if they win. Pocketing the losers' commissions provides most of the long-run profit for the bookies. They'd like to see at least as many dollars in losing bets as in winning bets. Therefore, when they “set the lines,” they hedge by choosing X in such a way that options (C) and (U) will attract roughly equal support. If they do, it matters little which option wins; either way, the bookies get about five percent of the total wagered.

Some fans have unusual insight into the relative strengths of the two teams, or at least they think so. These fans place their wagers confidently. But it has long been believed that if you're merely flipping a coin, (U) is the better guess. Take the underdog plus the points.

Why should that be true? On this website in 2006 I quoted Jay Kornegay, the executive director of the Las Vegas Hilton sports book, who said “the novice bettor, the casual fan, almost always bets on the better team.” To attract more money to the underdogs and balance the betting pool, the bookies offer a higher spread. With the bar set unrealistically high, the favorites are frequently unable to cover it.

Years ago I wrote about The Ted Baxter System, based on an episode of the Mary Tyler Moore Show in which the fictional newscaster wagered only on teams that were underdogs by 11 points or more. I ran an experiment which suggested Ted's strategy might be profitable. However, most gamblers don't want to bet on merely one or two lopsided games a week. They need more action.

This past summer, I learned of a way to overcome that last objection when I heard from David King down in Australia. He had come upon my article on the World Wide Web.

Very interesting indeed, as I had watched that show and tested the theory the same as you had, but on all sorts of different sports.

In Australia we have the Australian Football League (AFL) — not soccer, but they call it Aussie Rules football, vaguely similar to your football in the USA.

Just wanted to let you know that I bet the system in a different way. I bet all the underdogs but stake according to the points spread. Therefore, the bigger the points spread (underdog), the bigger the bet. So there's plenty of action (turnover), and the greater the points spread, the greater the profit.

I decided to check out David's idea. First I had to collect some data. I charted the 256 games of the 2017 NFL season, noting the point spreads (usually on Wednesday morning) and simulating three different systems on a spreadsheet.

(Why did Ulf and Ted wager such weird dollar amounts? I post-adjusted the spreadsheet, arranging things so that over the entire season each of the three gamblers risked the same: an average of about $78 per week.)

During the first third of the season (the six weeks from September 7 through October 16), underdogs came through as expected. Although they had only a 41-50 record on the scoreboard, against the spread they tallied 58 wins and 33 losses (.637). Ulf and David were able to celebrate the outcomes of almost two-thirds of the games.

The underdogs' best week against the spread was Week 3, when they went 12-4. A fifth of David's winnings that week, $13, came courtesy of the Houston Texans. As 13-point underdogs to the New England Patriots, they lost by only three.

But then, somehow, another sport changed everything. On October 19, baseball's defending World Series champions, the Chicago Cubs, were eliminated from the National League Championship Series. They lost at Wrigley Field by the humiliating score of 11 to 1. Who would have thought it?

After that, the prognosticators suddenly got a whole lot better at setting lines. (I know, it doesn't make any sense, but it happened. Maybe they gave up on baseball and began concentrating on football.) Now it became much less likely for an NFL underdog to win against the spread. All three of our hypothetical bettors, like Cubbie fans everywhere, lost their shirts.

In the final eleven weeks, the underdogs had only a 62-91-12 record against the spread (.405). The worst was week 12, when they went 3-12-1.

This ate up everyone's winnings. Over the entire season, the underdogs finished 120-124-12 against the spread. Our three gamblers each bet $1,326. Ulf lost $89 and David lost $90, or about -7% of the amount they'd wagered.

On the other hand, Ted won 10% more bets than he lost. With a record of 11 wins and 10 losses, he won $694 and lost $631. But he had to pay his bookie an extra 10% of his losings, a total of $694, so on a monetary basis he came out exactly even. The Baxter System is still the best!

The favorites won 68 percent of their games outright in 2017, as one might expect. More importantly for the gamblers: After a miserable first six weeks, the favorites rebounded to cover the spread in 51 percent for the season as a whole.

This outcome was unexpected, but not unprecedented. In an update to my previous article, I noted that in 2005 “‘the favorites covered the point spread an incredible 63 percent of the time,’ according to a column by Jim Armstrong of the Denver Post.”

The moral: Usually it's a good bet to take the underdog and the points — but not always. As the saying goes, “You pays your money and you takes your chances.”

As I had the data on hand, I've prepared some charts and graphs if you're interested.

Remembering one of my graphs from 2002 (shown at the right), I hoped that the histogram below might show a dramatic preponderance of the number of winning bets as the spread increased. However, the 2017 season didn't really turn out like that.

Which teams were the biggest disappointments? Obviously the New York Giants and the Cleveland Browns, who finished the season with the league's worst overall records: 3-13 and 0-16 respectively. But which teams did the best? Here they are, in green:

* As shown here, the Pittsburgh Steelers were favored in 14 games and won 12 of them on the scoreboard. However, they often disappointed their supporters by failing to cover, going only 6-8 against the spread.

** As shown here, the Minnesota Vikings were favored in 12 games and covered the spread in 10 of them. In fact, they won outright in all 10.

The best underdogs were the Tampa Bay Buccaneers, who won against the spread when they were 10-point and 7-point underdogs. But they also lost against the spread when they were 10-point and 7-point underdogs. So those four games canceled each other out, and David's net results involved only smaller wagers.