Race predictor is garbage

Most likely your VO2 Max is over estimated, making the watch estimate you have the cardiovascular capacity to run 5k in 26mn and change. The race predictor takes your Vo2 Max and your recent training history to come up with predictions for races. Predictions for races tend to be more accurate for a specific race than for a generic 5k. The generic ones are for "ideal conditions".

In my case, my 5k predictions are spot on. They have been a bit optimistic in the past. I am going for a 10k now, we'll see how that one goes.

It is too bad that Garmin race predictions don't come with a range of error. Stryd predictions come with one and I always end up in that range for my 5k, even if those predictions are a bit optimistic. If the range of error for Garmin is the same as the Stryd one, despite the Garmin predictions having been too optimistic in the past, I would also have been in the range.

Maybe the vo2 max, maybe the max hr is too high. Both were automatically set by the watch and previous Garmin watches I wear 24/7 for 4 years.

Unless you have done super max intensity intervals that would drive your HR to your max (google max HR self test)… there is no way that your watch could know what your max HR is, plus even if you do a Max HR test unless executed perfectly with tons of adrenaline, caffeine, heat, cheering crowd... its probably going to need to be adjusted higher a 3-5bpm

Watches don't know if something is max effort , if you can still see straight or if you want to fall onto the ground. lol They are just watching for a new high bpm, but is it good data?  was it just a spike of optical HR?  Best to review data and manually set based on tests.  your recent/past 5k races if you had a good finish sprint would be good ones to look at!  

Need also an accurate updated weight to be set in Garmin for vo2max to be accurate.  You could have a genetically high vo2max but currently lack the fitness to make the most of it.  

At least for the weight I know it's accurate to a week ago :) 

Regarding the max HR and vo2 max: who knows. The "best" way for me to assess it is to compare my race prediction to my race time I guess.

Though it is really a good question, not just fir the watch but to me as well: how much should I kill myself during a race like that? The 2 races I got trophies were really hard, I mean I have all I had and couldn't move for a few days afterwards, but I'm not sure if that should be the case. Or maybe it's ok if it's within the "Garmin recovery hours range", meaning as long as it's less then 96 hours. Once it reaches 96, which to my experience seems to be some cap by Garmin, probably means: a. you don't know how long, b. maybe it was a but too much.

My experience is similar - the Garmin algorithm is too optimistic...possibly because it is too simplistic and weighted all or heavily on VO2max. In contrast, Runalyze has an algorithm that considers VO2max, weekly training mileage and long run frequency, which I find to be more accurate as a race predictor.

Races are to test yourself! (IMO)  So yeah it should be 100%+ and should put you in a world of hurt the next couple days. lol  The more trained you are the less that recovery will take, especially for shorter races.  I've had plenty of Garmin recovery time IN DAYS! LOL The first time I raced a really fast/hard 10mile race after it said "3 days recovery"...I was shocked, didn't know it would show it in days! LOL  (marathons, ultras, 100+ bike rides etc)

If your body/mind can do it... then you do it.  Ya don't limit your effort in a race! lol (besides obviously pacing it appropriate and adjusting for hopefully even/negative splits).  Pacing/goals can often be tricky cuz either you pace conservative and miss your true potential... or you go to hard at start and fizzle !  I always like the go out a bit too fast... and hold that pace til the end! LOL

I’m the opposite and beat the prediction. Last month I ran 3+ minutes faster than the prediction for 8k. My suggested threshold paces are also slower than I run them. 

The less data it has, the worse the prediction, so a year of not running will throw it way off.

I had been training regularly for my last 5k, several runs a week of various intensities and lengths.  The predictor said 25:18, and my chip time was 25:21 with absolutely nothing left in the tank (a great step up from last year at 29:11!).  So for me it was right on the money.