Problems Persist with Gmax Timing System

|

Problems with the Gmax timing system continued during the Bing Crosby meet at Del Mar | Benoit

By

Despite a press release in late August from Equibase in which the company announced it would be conducting an “extensive analysis” of its Gmax timing system, the technology has continued to produce a number of inaccurate times at the racetracks it services.

The TDN has found recent examples from Del Mar, Tampa Bay Downs and Penn National that show there were problems with the Gmax timing of the races. Gmax, which is operated in the U.S. by Equibase, relies on Global Positioning System (GPS) to record times, as well as providing additional information such as the running order of a race.

Equibase issued the press release after the TDN and other media outlets reported on a series of problems with the timing system since it was first introduced to American racing in 2018 and subsequently suffered a number of timing issues during this year's Del Mar summer meet. Starting with the Aug. 2 card, Del Mar took the unusual step of relying not on Gmax, but on hand times, for all its grass races through the remainder of the meet.

The more recent problems were particularly pronounced during the early days of the meet at Tampa Bay Downs, where the Gmax system is being used for the first time. The chart for the first race on opening day, Nov. 25, notes that the race was hand-timed. During the running of the first three races of the day, no fractional or final times were posted on the toteboard as the races were being run.

After an uneventful day Nov. 27, the second day of the Tampa meet, there were several problems Nov. 28. The charts of five of the 10 races on the card noted that the races were hand-timed. In some of the races, no times were posted on the toteboard or through the simulcast feed while the races were being run. In others, the times listed were clearly incorrect. In the fifth race, the opening quarter time was posted as 33.81. In the sixth race, after the horses crossed the wire a time of 1:29.94 was listed for the quarter-time. In the seventh, the quarter-time was posted as 1:04.39 and the final time for the six furlong race went up as 1:10.71. In the official race chart, the running time is listed as 1:11.70, nearly a full second off the Gmax time originally posted on the toteboard.

Tampa Bay Downs President Pete Berube told the TDN that some of the problems with the posting of inaccurate fractional and final times were the result of errors being made by an outside vendor contracted to handle Tampa's television graphics package. Berube added that he had not lost faith in the Gmax system.

“I am very confident with the times,” he said. “I know there has been a lot of talk about timings and things like that from the speed figure guys and I certainly appreciate that. But I am very pleased that Equibase is doing the video control with the timing to make sure we are putting out correct times because I know how important they are to the bettors.”

Over the Aug. 1-2 weekend, the original times of seven races at Del Mar were updated in the final chart. The altered times came in both turf and dirt races and the differences were as small as 0.07 seconds and as big as 1.19 seconds. Addressing the problems after the meet concluded, Del Mar officials announced that they had re-surveyed the turf course to “enhance timing and tracking accuracy with the state-of-the-art GPS system.” There was no mention of correcting problems with the timing for dirt races.

Speaking to the California Horse Racing Board in September, Del Mar President Josh Rubinstein told the regulators that the track would not have to rely on hand timing for the Bing Crosby meet because problems with the Gmax system had been corrected.

But the official charts for five races run at the Bing Crosby meet list that they were hand-timed. Two of the races were run on the dirt and three were turf races. Asked by the TDN to address the five hand-timed races, Rubinstein pointed to extenuating circumstances that led to the problems. Some of the problems, Rubinstein said, involved miscalculating how far the run-up distance prior to the start was and how far out the rails were on the turf course.

There were a handful of additional problems during the meet. Over the three-day span beginning Nov. 20, there were at least four races where Gmax times that went up on the toteboard while the race were being run were changed in the final chart.

Equibase President and COO Jason Wilson said there were issues with “probably 10 races” at the Del Mar fall meet.

Despite the problems with Gmax, Rubinstein said he remains a fan of the system.

“Del Mar continues to believe in the platform and GPS technology for timing and tracking” he wrote in an e-mail  “We're working with Gmax and Equibase to provide the most accurate information possible.”

During November, there were numerous occasions at Penn National where the Gmax time was later corrected, with an adjusted time going into the official charts. On the night of Nov. 13 alone, the were seven races in which the final Gmax time was adjusted before going into the official chart.

Chris McErlean, the vice president of racing for Penn National Gaming, referred Gmax questions to Wilson.

Among the first to discover that there were problems with Gmax, the team that puts together the Beyer figures stopped using the Gmax times when making their numbers. Instead, they used a computer program that, they said, allows them to get accurate times by watching the replays. Wilson said that Equibase is now relying on a similar video timing program whenever it comes to their attention that a Gmax time may be inaccurate. By doing so, it appears that Equibase has been able to catch most of the mistakes and correct them before they become a permanent part of the charts and a horse's past performance lines. But Wilson admitted that it would be preferable for there to be fewer mistakes in the first place.

“We use video timing as a way to check for races where there needs to be some investigation as to whether a time is good or not,” he said. “We will go in and review those races and make changes as we need to. People have probably seen some of that. I think we need to work on how we communicate those changes to people. A lot of this is growing pains and, unfortunately, it has been a bit more painful than we thought it was going to be. We are getting there. Obviously, video timing every single race is not, in the long term, sustainable. It's just not a good use of resources.”

Gmax was developed by the British company Total Performance Data. In an Oct. 11, 2018, press release, Equibase first announced its partnership with Total Performance Data and that Gmax had been installed at Woodbine, Golden Gate Fields, Laurel and Pimlico. Gmax was being touted as an efficient and inexpensive timing system that could not only time races but provide tracks with such things as automated charts and dynamic video graphics. The problems began at the outset. Theracingbiz.com website reported that during a four-month period at Laurel beginning in Feb. 2019, 10 track records were set at Laurel and that it was later found that in all 10 cases the Gmax time was faster than the time recorded by traditional timing methods.

Wilson said that Equibase realized early on that an effort was needed to continually improve the system.

“When we went into this project, we didn't necessarily look at it that we were buying something off the shelf, had to install it and that was that,” he said. “We definitely looked at it from a standpoint of how can we make the overall environment better and bring in more research and development. It's not just times. It is information in general. How do we improve on that and make it better for everybody?”

That 21 months have passed since the first signs of problems at Laurel and Gmax is still creating a number of inaccurately timed races has raised questions as to whether or not Gmax will ever work properly and whether or not GPS is a good means of measuring time. In the meantime, other sports where timing is part of the equation have made great strides in their timing methods. The times for Olympic events are now so accurate that races can be measured at one-millionth of a second. Gmax times have been known to be off by as much as a full second.

Wilson said that tests have shown that Gmax is getting better all the time and he listed a number of steps Equibase has taken to improve the technology. Wilson acknowledges that the goal should be for the Gmax times to be so reliable and so accurate that they no longer come under question. He said he was confident that day would come.

“We are in the business of continuous improvement,” he said. “We don't want to have to check the times with video timing going forward. We want to be to the point where we don't have to go and check those times because we are confident they are accurate to, say, a tenth of a second 99% of the time.”

Editor's note: Barry Weisbord, the founder and former publisher of the TDN, is the Chairman of Trakus, a competing timing and tracking system. 

Not a subscriber? Click here to sign up for the daily PDF or alerts.