Garmin Race Predictor or VDOT Calculator

The VDOT Calculator works really well for me. My Garmin thinks I am way faster than I really am. I love that it is so supportive of me but trust VDOT.

Recently, on this forum, I've probably written the most pertaining to this topic, in various threads. It is obvious that Jack Daniels VDOT calculator is far superior to Garmin's race predictor, since the latter can be prone to errors for multiple reasons. That said, it can provide a useful up-to-date look at one's relative fitness (i.e., how it is changing) IF (a big IF) one understands some of its limitations. For example, if all recent workouts were done under the same, consistent, conditions, then it would become less subject to error in its averaging of workouts being performed under varied (sometimes slower) conditions.

Garmin will give good estimation +-5%, if you setup exact known max HR, and this a problem...

Garmin has 2 versions of race predictor: one (older) is a simple VDOT table lookup based on it's already quite flawed VO2Max estimation.

The newer one is a bit smarter, but still not very predictive.

Both have no idea about your endurance or anaerobic/aerobic balance, so you can treat them as your potential if you put in the required endurance or speed work. Only problem is that once you do that, your VO2Max estimation is likely to increase :)

Agree that VDOT tables/calculators are more accurate than Garmin. My Fenix 5 (and the 3 before it) think I am way faster than I really am - predicts 30 seconds faster than I have ever run a 5k in close to ideal conditions. While I have an idea of what my max HR is, I probably have it just ever so slightly off...

VDOT was spot on. I ran 19:44

I got a Fenix 5.

Did the following workout:

1 mile warmup at 8:20 pace

Then (14x1 float--1 minute at 1600m pace, 1 minute at probably MP)

1 mile cooldown at 8:20 pace

It says my VDOT is 50. I don't buy it. It was also 82'F out so that skewed it.

https://www.coacheseducation.com/endur/jack-daniels-nov-00.php

A VDOT of 50 is this: 5:50 mile/19:57 5k/41:21 10k/1:31 half/3:10 full

My last marathon was a 3:24, so about a 46 VDOT (6:17/21:25/1:38:27 [close to my 1:39 half PR, although that included a 2-minute bathroom break]/3:24). I do think I'm in better shape than I was in 2019, but I'm not quite at sub-20 5k level yet or 1:31 half, but probably just a little bit above that.

VDOT is probably the most accurate. Garmin 5K off the v02 is just like an ego boost. Runalyze says I'm in shape for a 2:18 marathon which again is nice of them to say that but lol no...

Garmin's race predictor drops off your predicted pace quite a bit on longer distances. Has anyone found their predictions accurate? Most other prediction methods have a more even, gradual slowing of pace as it predicts longer races.

jecht wrote:

I got a Fenix 5.

Did the following workout:

1 mile warmup at 8:20 pace

Then (14x1 float--1 minute at 1600m pace, 1 minute at probably MP)

1 mile cooldown at 8:20 pace

It says my VDOT is 50. I don't buy it. It was also 82'F out so that skewed it.

https://www.coacheseducation.com/endur/jack-daniels-nov-00.php

A VDOT of 50 is this: 5:50 mile/19:57 5k/41:21 10k/1:31 half/3:10 full

My last marathon was a 3:24, so about a 46 VDOT (6:17/21:25/1:38:27 [close to my 1:39 half PR, although that included a 2-minute bathroom break]/3:24). I do think I'm in better shape than I was in 2019, but I'm not quite at sub-20 5k level yet or 1:31 half, but probably just a little bit above that.

The Garmin is providing you an estimation of your "VO2max", which is technically, not the same as Jack Daniels' VDOT. The latter is an "Aerobic Power' metric, which theoretically is combining one's VO2max with the effect of their 'running economy'. It stands to reason that VO2max will be higher than VDOT.

28 minutes alternating mile pace/marathon pace is a stiff workout. I don't know if I could do it. I'm pretty sure I couldn't. Just doing that workout at even pace would require a near-race effort. Or did you mean 7 x 1 minute at mile pace with 7 x 1 minute at marathon pace as the float? That would still be a challenging workout. In any case, the workout you did is a good indication that you're fitter now than before.

I find that the "vO2max" number my watch shows me runs about +4-5 beyond my VDOT based on my race results. My all-time high VDOT is 60.2, while my watch topped out at 64 that week.

This is all based on estimates, but this suggests to me that your workout was roughly equivalent to your previous PRs, and since you were capable of running that effort in a workout in 82 F weather, you're probably capable of running PRs in good race conditions.

The McMillan calculator is pretty good:

https://www.mcmillanrunning.com/#run-with

The Garmin "race predictor" is the closest its ever been for me right now, but it is still estimating a 5k time ~1:00 faster than I could run (It's estimating 16:47, so that's way off).

My garmin race predictor thinks I am much slower than I am. 16:58 5k to 2:58 marathon. My guess is I am in about 15:30 shape at the moment. I could probably gut out a 2:50 marathon but it would hurt. The day I ran my 5k PR (15:23) Garmin guessed 16:45. Not sure why it is SO far off for me. Presumably because I am aerobically under developed and my zone 2 running is slower than everything else.

Runalyze on the other hand is pretty much spot on for me @ 5k. I had to tinker with it a good amount and apply a slight correction factor but I seem to have it to a point where it’s within a few seconds of my 5k fitness at this point in good weather conditions. Admittedly, the heat has been mucking it up recently.

My one complaint with all of these calcs is that they look at things ways too longitudinally. They need to do a better job of recognizing really hard workouts or races and appropriately weighting those into the prediction. It just makes garmin look stupid when you literally run an all out 5k and their prediction is more than a minute off that time following the effort.

Actually, I believe VDOT by the nature is VO2Max at average running economy.

JD had a big array of data of VO2Max values of athletes he tested together with their race results, so he just created a table from it.

In other words, it's lower than VO2Max for runners with relatively worse economy and is higher for runners with relatively better economy.

highhoppingworm wrote:

My garmin race predictor thinks I am much slower than I am.

My one complaint with all of these calcs is that they look at things ways too longitudinally. They need to do a better job of recognizing really hard workouts or races and appropriately weighting those into the prediction. It just makes garmin look stupid when you literally run an all out 5k and their prediction is more than a minute off that time following the effort.

I have the same issue where Garmin predicts too slow and this is a great point. I can understand not adjusting if you "underform" and run slower, but if a 5k or 10k is blown away by multiple minutes then it needs a more intelligent adjustment like the Runalyze correction factor. Even with an HRM and an accurate max HR too.

Runalyze is dead on, VDOT is pretty good without any HR info, but Garmin is just seems to miss even with all the info it has.

Darby wrote:

JD had a big array of data of VO2Max values of athletes he tested together with their race results, so he just created a table from it.

Daniels' VDOT Calculator gives equivalent-times predictions that are not too far off from McMillan's Calculator predictions. They seem to work, generally, or they would have probably received significant push-back by now.

I'm curious whether Daniels (and Jimmy Gilbert) decided to, in essence, incorporate a 'knock-down' factor into their equation so that their VDOT's typically would come in under VO2max's garnered from treadmill tests?

Dr. Kenneth Cooper--as well as others (e.g., Balke)--had an array of test subjects on which their VO2max values, determined from treadmill tests, were correlated with their 12min Time Trial results. In the following table, we observe how Jack Daniels' VDOT compares with Cooper's VO2max (extracted from a 12min Time Trial, using the following equation*: VO2max = (dm -504.9)/44.73):

12-min TT Distance..................Cooper VO2max...............Daniels VDOT

3110m...................................................58....................................50

3360m...................................................64....................................55

3605m...................................................69....................................60

3845m...................................................75....................................65

4082m...................................................80....................................70

4315m...................................................85....................................75

*Cooper's VO2max equation is found in the Wikipedia article titled "Cooper test".)

I suppose Daniels, if he had wanted to (without actually changing his race-equivalency predictions), could have skewed his VDOT values, making them roughly 8-9 points higher across the board, so that they would more closely align with VO2max values determined from prior treadmill tests as observed by Cooper, eh?