The Principle of Maximum Confusion Revisited

According to his obituary, which appeared in the New York Times on May 7, 2009, Burton Paul Fabricand was a physicist, economist, financier, author and World War II veteran, born in New York City on Nov. 22, 1923.

He was also one heck of a handicapping theorist.

In his book “Horse Sense,” Fabricand showed why with a brilliant idea he termed “The Principle of Maximum Confusion.” 

Armed with the knowledge that the top public betting choice typically loses about nine cents on the dollar — a situation, it should be noted, that no longer exists (the loss is much greater) — Fabricand set out to find a means of turning that red ink green.

Toward this end, he postulated that favorites are most likely to show a profit when there are other horses in the field with similar past performances. The logic being that the crowd must have some reason for making the favorite the top betting choice, even though its merits don’t exactly scream out from the pages of the Brisnet Ultimate PPs.

Fabricand then presented seven rules to determine when the favorite’s past performances are comparable to those of one or more of his foes. Unfortunately, a couple of these rules are highly subjective and/or dated, so I came up with a simpler set of criteria and I also narrowed the scope of Fabricand’s Principle a bit:

1) Only non-maiden races are considered for play.
2) The morning line favorite and the second choice must have the best last-race Brisnet speed figures and the best Brisnet Prime Power Ratings in the field.
3) Both horses must’ve raced in the past month.
4) Play is on the morning line favorite if his/her last race Brisnet speed figure ranks second in the race.

Two perfect examples of this revised Fabricand approach came on March 20, 2015.

In the fourth race at Hawthorne Park, a $5,000 starter affair, Indian Artifact was the 9/5 morning line favorite, while Masquerade Fashion checked in at 5/2. They’d both raced in the past month and they ranked 1-2 in Prime Power and Speed Last Race.

As the morning-line choice with the second-best last-race Brisnet speed figure in the field, Indian Artifact qualified as a bet and won by a length and a quarter, paying $4.20. Masquerade Fashion finished third as the 9-5 second choice.

The fourth race at Turfway Park, a straight $5,000 claiming event, looked eerily similar to the race above, only this time the prime contenders were Citizen John (2/1 on the morning line) and Followthemoneytrail (5/2). Both had raced less than 30 days ago and both ranked in the top two in Speed Last Race and Prime Power.

And, just like before, the morning-line favorite (Citizen John) came away victorious, narrowly defeating Son of Sparta and returning $5.00 as the 3-2 favorite. Followthemoneytrail staggered home fifth and, like Masquerade Fashion, was over-bet, going to post as the 2-1 second choice.

Of course, I realize that two examples do not prove anything, so I tested this revised Principle of Maximum Confusion on my database of over 14,300 races.

The results were encouraging.

Plays: 152
Winners: 68
Win Rate: 44.7%
$2 Net Return: $2.15
ROI: +7.30%

Obviously, one should test this method further before cashing in the 401(k) and playing the races for a living, but it sure looks like Fabricand may have been on to something.