This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

PacePro could be so much more

So I did my first race following a PacePro strategy. Despite other threads in this forum it worked quite well, once you came to terms with how creating and uploading a strategy works.

I am now planning my next race and since PacePro has been a major selling point for me, I can not help but notice, that for the money involved I expected it to be more than something you can re-create with a spreadsheet.

Here is in a nutshell what PacePro does in preparation of a race:

  • Analyse a given course and calculate split times depending on your desired overall pace or race duration.
  • Split times will generally speaking reflect slower paces for uphill splits and faster paces downhill.
  • Calculations can be adjusted in two dimensions:
    1) estimated effort it takes to run uphill (will increase the min/max range of paces, making uphills slower and downhills faster),
    2) overall positive/negative split (positive split will calculate splits early in the race faster than similar splits late in the race, negative splits vice versa).
  • Splits can be set in three different ways,
    1) create a split every 1km,
    2) create a split every 1mi,
    3) create a split before every significant elevation change (personally I find #3 the most useful, since it is the "truest" reflection of the elevation profile, yet other people might not be comfortable with odd splits).
  • The strategy will than be stored as a split table and can be synced to your watch.

That surely sounds like a helpful addition, but it's 2019 and we're talking about a device probably more expensive than your average smartphone. Reading through the lines of a static pre-calculated table is a pretty low bar for a flagship device, proudly introduced in front of the UTMB community.

So here is what I hoped my fenix 6 could do, what I think it is able to do and what I would love to see Garmin add to PacePro with upcoming updates:

  1. (Re-)Calculate splits dynamically.
    Every strategy goes out of the window on first contact with reality (e.g. the weather). So if the initial goal had been set to finish a race in 4 hrs, I may be well ahead or behind that goal after a few splits. The watch already displays the time deviation and an estimated finishing time, but it would be much cooler, if it would adapt the strategy, lightly increasing/decreasing upcoming split times to compensate for the "error".
  2. Adapt to performance
    Similar to 1., recalculate the strategy on the fly, but from a different angle. PacePro currently requires you to determine your strategy (see above, those 2 dimensions) ahead of a race but you might feel different for better or worse once you're at it.
    So instead of sticking to the set strategy, a constantly occuring deviation in target split time and achieved split time could influence the remaining strategy. For instance, constantly going slower uphill and faster downhill could be an indicator for setting the uphill effort too low. Or, if you're increasingly running behind your targeted split time you probably set a too optimistic negative split strategy which should be corrected.
    You could even go as far as to allow for the finishing goal to be corrected, maybe allow a percentage or absolute number of minutes that the strategy may deviate from the original goal.
  3. Learn from activities
    This is probably the greatest misconception I had about Garmin, the level of prediction their devices offer based on the data they gather.
    Instead of letting me guess how the function curve of split and uphill strategy would look look like to achieve my goal, this could be learned from my past activities and improvements.
    I bought the fenix imagining that it could at least do the work that I already did tediously with Excel and a lot less data.

Maybe I aimed a bit too high, but even in its current form, PacePro seems to be too "simple" to serve the purpose it advertises.
Other runners had related issues with the strategies.

  • Strategies seem to be limited in the total duration or distance, which seems counterintuitive when especially mountain races may take longer than similar distances on flat roads.
  • The slider that controls the effect of uphill effort doesn't go far enough, so minimum uphill paces are too fast (or vice versa, downhills are too slow).
    Walking uphill to conserve energy is not uncommon even among strong athletes, plus highly technical trails will slow that down even further.
  • Runners tend to split their races not only in distances or changes in elevation but also between aid stations.
    Maybe allow for waypoints in a gpx file to serve another split indicator

Again, maybe I just fell for my imagination when I read about PacePro, but there is real potential for something that makes impressive use of the existing data and computing power we carry around our wrists.

Just for good measure, here are some other threads I dug up on PacePro.
PacePro strategies not suitable for slower runners
forums.garmin.com/.../pacepro-strategies-not-suitable-for-slower-runners

Pacepro post-activity analysis
forums.garmin.com/.../pacepro-post-activity-analysis

Pacepro for Ultras Longer than 24:06:59
forums.garmin.com/.../pacepro-for-ultras-longer-than-24-06-59

  • I absolutely agree. However, I can't quite see how a device as small as the F6 could have the computing power to do all that. Granted there's elements of analysis that could be done in Garmin Connect eg developing a strategy based on previous history, but changing the strategy on the fly would require a lot of processing power that might not be available, or consume too much battery. 

  • I was looking for something like that but managed to ignore it somehow. Thank you!

  • Unfortunately there is little info to find about the computing power actually available. I just figured, if it is sufficient enough to create it before the race, and if the older models for years has been suffciently powered to do routing and navigation... at least re-using the same method should work. Granted the strain on battery life but the fenix 6 has plenty of that.

  • So far in my experience, pace pro works great while going on a loop course or a 1-way course although it could be much smarter like you listed. The worst part is when using it on a there - and - back course. You have to start and turn around at the exact spot listed on the course map or the ahead/behind times are totally wrong and don't get corrected; you can beat your goal but still the watch can show you're falling behind

  • That's the unfortunate truth, and as I understood from a review, it is intentional / by design. PacePro seemingly compares your progress in distance against the splits it created, and not the splits against your progress on the track your following. I had the same issue when having to take a small detour. Since then every new split notification was too early because I was still going uphill when the split changed to the following downhill pace.

  • I'm not sure exactly how you mean, but for me the feature that Pacepro compare times against your position on the course and not against your distance on your gps-track, is the best thing about it.

    I've never done a city race where I have trusted the distance/pace on the watch and use that information to calculate an expected finish time.

    I've done one race with pacepro and it worked great for me.

    From Dcr review:

    "What’s notable here is is that it locks to your GPS location on the predefined course (think of it like a train track, or roller coaster ride), rather than your watch distance the GPS has measured. This has its pros and cons."

  • Agreed, I had to read that a few more times because the example of his NY track confused me a bit.

    It maps the position you are currently at to the position you are supposed to be according to your track, and then give you the details for that split. And that's the right way to do it.

    Now I have to look elsewhere why new splits started a bit earlier than expected but maybe that was just due to a positioning error in the woods.

  • I would be content if we could at least define the splits/times manually in Connect. Currently with automatic splits it seems to be aimed exclusively for road races (with maybe a few hills). Trying to use it with a trail race results in splits that are typically too high for uphills and/or too low for downhills. If at least it was manual I could set the splits based on my fitness and knowledge of the course (grade, surface, aid stations, etc

  • Yeah, other runners over on reddit expressed the same. Either have some sort of pace input option for gradients or gradient ranges (e.g. slowest pace uphill = X, calculate the rest from there)

    Don't know if the math still works in this case. If you set distance and absolute numbers for split paces as constants, than there is little room for calculating variables that achieve the overall time / pace that you also defined.

    On the other hand, the fastest and slowest pace in the strategy are already a result from using the slider controls, so if you would allow to type in one of those values it should be able to come up with the according opposite value.