Race predictor

Your personal best results are just some data points that the watch takes into account as long as they are recent.

The race predictions are based on your VO2 Max and your recent training history.

It follows that if the predictions are off,

- either your VO2 Max is off,

- or your training history doesn't capture your actual capabilities.

To the first point, the most common issue with an inaccurate VO2 Max is an inaccurate Max HR value. The VO2 Max algorithm has a 5% error on average with a formula-based Max HR, lower with the actual Max HR. An error of 8bpm in Max HR can triple the errror rate on VO2 Max. Another common source of error in Vo2 Max is some wrong HR or pace data, or a bias in the training history (see following point)

To the second point, make sure your training history has a balanced focus to expose the watch to your capabiliites across a range of durations and effort. From time to time, I recommend you substitute your training plan or daily suggestion pace target with a maximal steady pace for the duration of the intervals.

Check also your threshold estimate on the watch. If you adjust your Max HR, run the threshold test again. In the end, you should be able to run the 10k at your threshold pace. Users who tend to have optimistic race predictions tend to find their threshold pace too optimistic and vice-versa. Research has found that threshold performance is a better race predictor than VO2 Max, and it is possible that this data point is also used by the Garmin algorithm for the race predictions.

Appreciate the thorough reply.  Neither apply here as both watches have the same Max HR & VO2 Max.  I have been trying to train more in Zone 2 since getting the Fenix and with a HR monitor as well so it could be related to these factors.  You would think that if you ran a 5k well below what the watch estimates, it would be smart enough to drop the race predictor into that range immediately and make the normal nominal adjustments due to daily training from there.  Again, appreciate your reply!

ajhdr0p said:

Neither apply here as both watches have the same Max HR & VO2 Max.

Well, not so fast. Both watches have different training history data, correct? Also, both watches run different firmware and Firstbeat algorithms.

ajhdr0p said:

the race predictor still thinks my 5k pace is slower than what I ran the half in

Under the performance widget, what is the lactate threshold pace that the watch is currently tracking? 

And what is your 10k prediction?

6:11min/mi could give you a 10k within 38.20mn and a 5k within 18:05 assuming a 10k is run at threshold and a 5k is run 6% faster than a 10k. So the predictions are kind of aligned with your threshold data for the 5k and a bit conservative for the 10k, although within 1mn, which is the +/- error range I see for my Stryd predictions for 10k.

I recommend you perform a new lactate threshold test on the Fenix. I would also perform a couple of maximal effort anaerobic and supra-threshold workouts over the next couple of weeks. I would expect this would update your predictions.

Thanks for the responses and suggestions Etupes.  I'll continue the training and see if it comes down.  I didn't even know about the lactate threshold workout available.  I do wish there were more "instant" results based on racing quicker times, but I'll continue watch it (hopefully trend down).  Thanks!

ajhdr0p said:

I do wish there were more "instant" results based on racing quicker times

Yes, I do too, but it is not the way the watch works.

ajhdr0p said:

I didn't even know about the lactate threshold workout available

Do trust the lactate threshold built in test, but be weary of the threshold auto-detection in particular after interval workouts, unless the intervals are in the "threshold range" (ie above 20-60mn maximal effort). Auto-detection after anaerobic, VO2 Max or sprint tend to give me erroneous threshold paces.

If you look at the Stryd model, it gives you race estimates with about 30s, 1mn, 2mn, 4mn error ranges for the 5k, 10k, HM and M. So it is fair to expect Garmin has about the same ranges.

Once your estimates are within the right range compared with your PB, you know that the watch has a tuned dataset and all the training metrics (VO2 Max, Training Effect, Load, Targetst for daily suggested workouts, etc.) are as accurate as they can be.