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

Race predictor broken?

Yesterday I ran a 10 k race at 45:40 and today, after a good night’s sleep, garmin 10 k race predictor says 49:35. How does thar make sense?

  • Mine was suddenly fixed. It has estimated 5K at 30-31 minutes for months, despite me regularly running significantly faster in my workouts. Then suddenly it dropped to 23-24 minutes after I had an easy 10.3K run at 60 minutes or so. The difference was that I had no warmup, cooldown or stretching included in my run. I will usually log something like 90 minutes for a 50 minute run, and that seems to confuse the watch. 

  • It’s incredible that with all the technology nowadays the left hand doesn’t talk to the right hand. 

    Same thing happened for years. Last year I ran a HM and beat the estimate by a few minutes. The watch knew I had just ran the race yet continued to show the old, slower time. 

    Garmin refuse to publicly explain why this happens. 

  • The only things your Foreruner 255 measures with reasonable accuracy are distance and average pace (if your GPS signal is good) and heart rate (when you run slow or wear a heart rate belt).

    All the other numbers that your Garmin spits out (VO2max, anaerobic threshold, HRV, etc.) are calculations, not measurements. Those guesstimates can be off by 10% or much more, especially if you're much faster than average. As a result, Garmin's predicted race times are usually off by many minutes.

    Nobody in my run club races the times Garmin predicts. Everyone is quite a few minutes faster or slower.

    The only thing Garmin's race predictions may be good for is to spot overall trends: are my times going up or down during a multi-week training cycle?

    And even those trends can be highly misleading, so don't believe your Garmin when it calls your pre-marathon tapering "detraining" or "unproductive."

  • I see this quite a lot, an immediate change following a workout which then adjusts overnight.

    I suspect it does a rough, local calculation at the time and during the night its a server side calculation and push so its different in the morning.

    Warmup/Cooldown and cardiac drift seem to have large effects on my predictions too.