I’m working on a little side project to make a decent looking power analysis web UI. Because seriously, pretty much all of the available stuff is just awful.
One thing I’d like to integrate is Normalized Power, Intensity Factor, and Training Stress Scores. These ARE valuable metrics, and Peaksware declares they are registered trademarks. That being said, I can find no Trademark records anywhere for these methodologies.
There is plenty of documentation available that describes how to determine these values. I know this is a tri forum, but I’m wondering if anyone knows whether its possible to use these values without incurring the legal wrath of Peaksware.
If I can’t, it essentially means that Peaksware has pretty much cornered the market on most advanced training software since these are fairly valuable training metrics.
Yes, it’s called Intellectual Property. Welcome to half of Silicon Valley
You can trademark their names. You can (theoretically) patent their algorithms or methodology. I would think that they have done the former only.
The formula does not need to be disclosed and may be maintained as a trade secret. The same way coca cola hides the true formula for its product.
I’ve basically written something really close to WKO+ with a lot more customization and tools. It’s been me and a couple friends using it with no real plans to release. One of the reasons is all those trademarked terms that are used.
BTW NP and all the metrics that depend on it may vary depending on your implementation … especially if there are any big gaps in your data, since the algorithm works off 30s averages. I just zero them out … but you can get fancy with how you deal with it.
You can trademark their names. You can (theoretically) patent their algorithms or methodology. I would think that they have done the former only.
The algorithms for NP, IF, and TSS were discussed publicly possibly even before there was a Cyclingpeaks – we were discussing them on the Wattage list in early 2003. I did think it was kind of funny that Trainingpeaks (nee Cyclingpeaks) trademarked the name “normalized power.” Andy’s original name for it was “corrected power.” “Normalized power” was a suggestion that came from someone on the Wattage list.
The formula does not need to be disclosed and may be maintained as a trade secret. The same way coca cola hides the true formula for its product.
If you do not disclose the formula, then you cannot receive a patent. Coca-Cola’s formula is not patented. In theory, anyone can produce Coke…if they were able to stumble upon the right combination of ingredients.
By disclosing the details of a product, you are then granted a period of exclusivity (the patent). However, once the patent expires, anyone can produce that product. Hence the reason you see “generic” drugs eventually hit the market. The patent expires and then any pharma can then produce a “Claritan”, for example.
The formula does not need to be disclosed and may be maintained as a trade secret. The same way coca cola hides the true formula for its product.
If you do not disclose the formula, then you cannot receive a patent. Coca-Cola’s formula is not patented. In theory, anyone can produce Coke…if they were able to stumble upon the right combination of ingredients.
By disclosing the details of a product, you are then granted a period of exclusivity (the patent). However, once the patent expires, anyone can produce that product. Hence the reason you see “generic” drugs eventually hit the market. The patent expires and then any pharma can then produce a “Claritan”, for example.
Right, but patents are given in order to encourage innovation by awarding exclusivity. I try my best to know as little about patents as possible but, in general, that means you have to show that one else has previously done the thing you’re trying to patent and that you “invented” it – in a practical sense, it means you have to apply for a patent either before something becomes widely known (or relatively soon thereafter) and then you disclose the details in the patent so when the patent expires others can build on your work. In this particular case, Andy had openly discussed the algorithms for NP, IF, and TSS before (I think) Cyclingpeaks and WKO existed, and got a lot of feedback from the Wattage community on testing and refining and validating and using the ideas. Subsequently (I think) he hooked up with Hunter, who was the first to see the commercial value in these concepts. Much later other commercial products started using the same concepts and terms since they had been discussed publicly – and Andy went ballistic. It was too late (I think) to patent the algorithms but Cyclingpeaks trademarked the names (and abbreviations), so I guess we should either be writing NP™, IF™, and TSS™ or possibly NP(R), IF(R), and TSS(R) if the terms are registered trademarks.
The formula does not need to be disclosed and may be maintained as a trade secret. The same way coca cola hides the true formula for its product.
If you do not disclose the formula, then you cannot receive a patent. Coca-Cola’s formula is not patented. In theory, anyone can produce Coke…if they were able to stumble upon the right combination of ingredients.
By disclosing the details of a product, you are then granted a period of exclusivity (the patent). However, once the patent expires, anyone can produce that product. Hence the reason you see “generic” drugs eventually hit the market. The patent expires and then any pharma can then produce a “Claritan”, for example.
Ya I actually did not say it was patented I said it could me maintained as a trade secret, but it would be just as exclusive so long as you can maintain the secrecy of the formula.
I searched the United States Patent and Trademark Office website and could find no registrations for these metrics.
As for Peaksware’s claims, they (to my knowledge) do not claim they have registrations. They claim the marks are “trademarks of Peaksware.” (see http://home.trainingpeaks.com/) They also do not use the (R) designation but instead use the ™ designation in certain spots. Further indication that the marks are not registered.
This does not mean that they have no rights in the marks. They would just be limited to their common-law rights to the extent they haven’t been impaired by any of their actions or lack of action.
If I understand you correctly here, then I could use the same methodology for determining something like Normalized Power or Training Stress Score, but I may need their permission to use the mark “TSS”?
Golden Cheetah uses or used Phil Skiba’s similar metrics due to the this very issue. I don’t know the particulars but know that Phil was very accommodating, I am pretty certain no money changed hands, but I could be wrong. I think Phil calls them xPower and xPace.
If I recall correctly the main difference was that Phil used an exponentially weighted moving average instead of a 30 second moving average. If you pick the alpha of the exponential weighting correctly, you can make the rolling averahe and exponential weighted moving average quite close.
Golden Cheetah uses or used Phil Skiba’s similar metrics due to the this very issue. I don’t know the particulars but know that Phil was very accommodating, I am pretty certain no money changed hands, but I could be wrong. I think Phil calls them xPower and xPace.
xPower, BikeScore, and Relative Intensity.
I don’t think there was anything nefarious going on. I think the issue was that when GC came along, Cyclingpeaks had not yet figured out how it was going to handle requests to use NP, IF, and TSS and while they were figuring that out, GC needed to keep going. Phil was very accomodating, as you say, so GC went that way. The next version of GC (v. 3) will use either Phyfarm’s xPower and Bikescore or else Peaksware’s NP, IF, and TSS.
If I recall correctly the main difference was that Phil used an exponentially weighted moving average instead of a 30 second moving average. If you pick the alpha of the exponential weighting correctly, you can make the rolling averahe and exponential weighted moving average quite close.
Often, but not always. You can create oddball data streams where they differ by more than a little – but you’re right, usually they’re quite close.
Welcome to my world. The terms “Normalized Power” etc. may be trademarked. I do not know if they have been but a trademark application certainly could have been filed and granted. The algorythms are most likely NOT patentable by themselves. There is numerous caselaw on this issue. Gootschalk v. Benson (409 U.S. 63 1972) basically held that a method of converting binary coded deciamls to binary numbers was not patentable. I would imagine that formula using power data would similarly not be patentable.
The formula could be kept as a trade secret, but if someone idenpendently comes up with the formula then there is no recourse against that person as they did not obtain the information from the holder of the trade secret.
The trademarks are another matter. A trademark, if properly maintained, can go on indefinitely. You can certainly coin a different term(s) for the values you want to measure.
If I understand you correctly here, then I could use the same methodology for determining something like Normalized Power or Training Stress Score, but I may need their permission to use the mark “TSS”?
Why not come up with your own catch phrase for the same thing?
Something along the lines of “Training Impact Metric”, or come up with a completely unique term.
John
Just so you know I’ve already trademarked “craptacules”. This is the amount of bad workouts in minutes multiplied by your “apathy factor” (also trademarked).
Thanks DougRob! That is an exceptionally clear explanation. While I certainly won’t stake my future on forum legal advice, this is enough to allow me to proceed with this little project.
I should note to the GC crew, I really don’t mean any disrespect… Your efforts for the benefit of the community are laudable and certainly better than anything out there right now. I’m just another designer/developer with an itch to scratch.
The algorythms are most likely NOT patentable by themselves. There is numerous caselaw on this issue. Gootschalk v. Benson (409 U.S. 63 1972) basically held that a method of converting binary coded deciamls to binary numbers was not patentable. I would imagine that formula using power data would similarly not be patentable.
Maybe true, but a processing chip with those instructions may very well be patentable…
If I understand you correctly here, then I could use the same methodology for determining something like Normalized Power or Training Stress Score, but I may need their permission to use the mark “TSS”?
If I understand what’s happened in the past, Peaksware will let you use the marks so long as you give them credit.