Brooksee has been aggressively promoting their Laurel Timing technology. The reality is that races may want to consider the several downsides to their approach.
“At the finish line, Laurel by Brooksee delivered a 99.95% read rate, outperforming a leading legacy timing system used at the event, which recorded a lower 99.82% read rate.” from Brooksee Press Release.
We did a deep dive into several timing technologies in this blog a few weeks ago, including LoRa, which is the technology behind Brooksee’s Laurel timing. There are two problems. First, the accuracy is only between 20-200 meters, which means a time for a 6 minute per mile pace might be off by 4-40 seconds. The second problem is that each runner has to wear (and dispose of) a tag with a lithium battery on it.
We applaud trying new technology, but this technology has been tried in other settings (like military positioning of soldiers on a battlefield and proven not to be accurate enough). However, from what we have seen, races should think long and hard before implementing Brooksee Laurel Timing for two reasons.
- Will runners lose confidence in the times posted by a race, as well as confidence in the quality of the race if they have Brooksee Laurel Timing?
- Will runners lose confidence in a race who promotes a clearly non-sustainable bib tagging technology with lithium batteries?
Inaccuracy
The misleading statement in the Brooksee press release might lead people to believe that Laurel is as accurate (or better) than RFID. No, it is not. RIFD is accurate within tenths or hundredths of a second, where Laurel is accurate between 4-40+ seconds.
To solve for this, Brooksee uses Chronotrack for their (accurate) start and finish times, but uses Laurel for (inaccurate) splits (yet throws Chronotrack under the bus in their press release (when there are plenty of other reasons, including operator error, for a 0.18% missed read rate)).
Inaccuracy Example 1 – Fast Runner at Mesa Marathon
This can be seen at the Mesa Marathon (the subject of the press release) with an example runner – Spencer Brown, who ran an accurate (Chronotrack) 1:05:51. Here is a quick comparison of his splits as provided by Brooksee Laurel on the left and his watch on the right:
The splits are way off. In the last mile, Brooksee has him at 4:20 while his watch has him at 4:45. If you watch Spencer’s video, he explains how he went out conservatively and had a lot of “gas in the tank” at the end – but not a 4:20! Ten of the thirteen splits were off by over 12 seconds. Spencer is obviously a solid runner who knows his pace – imagine if he had been getting real time feedback on his splits during the race and the confusion it would cause. Imagine his friends following him seeing these splits and wondering what the heck was going on?
Here are screenshots of his splits on the Mesa Marathon website provided by Brooksee Laurel and his watch splits:
Rory Linkletter was the winner of the Mesa Half and he had similar inaccuracies according to his Strava Activity with half of the splits off by over 12 seconds with the worst being 30 seconds. This made the little “AI Generated” writeup irrelevant, even though Brooksee highlighted it in their press release. You can see Rory’s Youtube recap here.
Inaccuracy Example 2 – Houston Half
Two participants of the Houston Half shared their watch data as examples of the inaccuracy of the Laurel Splits:
This race showed similar traits with half of the splits being over 12 seconds off, with the largest split difference being over a minute.
Inaccuracy Example 3 – RunningUSA Morning Run
Brooksee paid RunningUSA to sponsor the morning run at this year’s Conference and time it. As part of that sponsorship, later that day Brooksee’s founder and CEO, Phil Dumontet, did a presentation that turned quite contentious. There were a number of the industry’s best timers in the room and they called into question his claims of accuracy. Phil double downed on his claim of 99% accuracy. The timers then used a number of examples from race results they had seen. Most alarming were the reports from the inaccuracies from the morning run:
- Multiple large (20-60 seconds) differences between watch splits and start/finish times since Brooksee used Laurel for the start/finish times.
- What seemed like “averaging” of splits for some participants. Here are a couple of examples we were sent:
- One person dropped out at the mile point, yet was listed with 2 mile, 3 mile and finish times – and even was an award winner. This type of error where the participant was actually off course or had health issues is a serious safety concern. Not to mention confusing to spectators.
- Two runners ran with each other and their overall times were almost a minute apart.
Comments on Splits and Impact on Reporting
We had this one thoughtful email from a timer:
“For me, outside of the obvious dishonesty in this reporting, there is a discussion of what exactly is a split and what is its purpose. Having a handful of splits that read at a reasonable rate and are publicly shown make sense, and I’d have less issue with this if the majority of these “splits” were just being used to bolster the tracking and not displayed as actual reads when they’re guesses. Bolstering tracking for folks not using GPS tracking has a lot of utility, but shouldn’t be presented as real splits at this point. Furthermore, splits are supposed to be reported to the BAA when data is submitted – are these “interpolated” times being reported? Does the BAA know about this? Because I think there’s a strong case that could be made for any race timed by Brooksee to be heavily reviewed by an organization like Boston, or NYRR, or anyone who requires a qualifying time.“
Lithium Batteries
Unlike RFID bibtags, Brooksee Laurel requires a lithium battery (CR1216) in each bib.
Lithium batteries contain hazardous chemicals, including lithium, manganese dioxide, and organic electrolyte. While the hazardous materials total only 68 milligrams per bibtag, races like Mesa had 10,000 participants and Houston had 7,000. Those are large enough quantities that many organizations would want to look for ways to avoid that hazardous waste.
Of course the other consideration for races is whether some participants might choose more sustainable races to participate in.
Brooksee Pricing
Brooksee is using very aggressive pricing. We have heard quotes around $2.50-$3.00 per participant. They are losing money on these bids since the cost of the Laurel tag is approximately $2.50-3.50, plus they still have to use a real RFID bibtag, plus all of the normal personnel and equipment costs. This is obviously unsustainable and is meant to try to get them either into long term contracts where they will raise prices or reduce their costs (maybe eliminate Laurel). So if a race is entertaining the idea of using Brooksee for timing, study the contract carefully.
Summary
As we started out, we love new technology and have an open platform that welcomes all types of timing technology. We would love to see the LoRa technology improve what races can deliver.
But the fact is, the LoRa technology is not appropriate for races. And all of the data we have seen in the base research of LoRa (accuracy of 20-200 meters (vs. GPS used in watches, Strava, RaceJoy, etc. is 3-10 meters)) as well as the results that Brooksee themselves publish show it is inaccurate. With the additional concerns of hazardous material disposal, we recommend caution before using Brooksee’s Laurel technology.
Appendix
We have received other examples of inaccurate data:
