race prediction inacurate, slower than already recorded trainings

The predictions seem to be pretty close to the real performances. The predictions are mostly based on the VO₂Max measurements which are quite approximative, so do not expect that it will give you times to the second accurate. 

thank you very much for your quickbreply. 

but i do not agree. 

Garmin says, the predicted time should be reached after a profund training for thst distance.

i did not yet specifically train for that.

if you understand about running and pace, you will know that 10min more or less at a half marathon is not at all about "a couple of secs". if it was so unprecise there would not be any use of it!

especially as i mentioned already i am not at my best form at all and already undercut the predicted time. garmin should know that more is possible after specific training. and that actualy is what garmin advertised this gimmick for. 

does anyone know sth about the algorithm, how it is calculated? 

Garmin says, the predicted time should be reached after a profund training for thst distance.

i did not yet specifically train for that.

Why are you then surprised the estimate is not accurate?

does anyone know sth about the algorithm, how it is calculated? 

Yes, Garmin. Respectively Firstbeat (who is now Garmin,... anyway). Do not expect they would reveal the algorithm.

i have my Garmin watch since April, and it shows me a VO2max of 55.

when saying "i did not yet specifically train for that", it means that with a specific training, i would most likely get faster than i am today. not?  but how can Garmin tell me my race time would be 1h44 if i already showed him that i can do it in 1h 35min. ? you dont get the point?

Somehow the race predict should always be minimum same fast as the recorded training runs, otherwise it is nonsense.not?

Somehow the race predict should always be minimum same fast as the recorded training runs, otherwise it is nonsense.not?

Why nonsense? Do you believe that your form is always only increasing?

Anyway, I think it is a bit naive to believe that the device can predict your true time based just on some VO₂Max estimates, that already alone are not accurate (FirstBeat claims the accuracy of 95%). Now, take 100 athletes with identical VO₂Max values, and I bet that their times will fluctuate certainly more than the 10% that we can see in your case (1h44 vs 1h35). The pace depends on many more factors, including the physiology, physignomy, technique, external conditions, and many other factors. The perfomance prediction still can be a usefull tool, when you watch how it chages with the time, but personally I would not expect any miracles from it.

thank you very much for your thoughts. indeed i think i expected too much from this calculation. 

just one last word:

yes i DO expect my form to get better, if i specifically train for something.

And garmin knows, how my form changes, and that its getting  better. at least it is telling me that. 

yes i DO expect my form to get better, if i specifically train for something.

Yes, you may expect it, but the reality may be different. Even at the top athletes, the performance varies during the season, and even between individual races. And then, of course, it changes with the age as well. Not speaking about the fluctuations of the form due to bad sleep, alcohol consumption, illness, stress, injury, or other factors that can influence not only the VO₂Max readings, but also the actual performance.

The discrepancy in the prediction simply comes predominantly from the VO₂Max estimates taken during some of your runs. So if the vlaue was not yet stable before your 1h35m run, or was accidentally slightly higher then recently, then of course, the prediction for the same distance will be lower.

It's even worse for me to be honest. I'm only just starting to look at what the different features mean. I've had the watch (245 no music) for about 8 mths or so. I've done nearly 400 runs/cycles. For my running the race predictor suggests 28:10 for 5k, 1:00:51 for 10 km and 2:31 for a half marathon. I'm doing upto 22 for 5k, 48 for 10k and 1:40 for 20k (I've not done a half marathon). So compared to my race predictions, yours look pretty accurate 3406754! My VO2 max is about 40, I saw someone suggesting theres is 55 here. does that make mine pretty low? What the implications of that?