Garmin
2 lurkers |
220 watchers
18 Sep
1:13pm, 18 Sep 2024
46,164 posts
|
SPR
larkim wrote: It'll never be perfect. And, to be honest, do any of us that run a fair amount really need the watch to give us an indication of how fast we could run those standard distances? Well it's a selling point of the watches and it has the potential to be the best predictor possible for an individual given it has each individual's training data. |
18 Sep
1:31pm, 18 Sep 2024
25,632 posts
|
larkim
Agree it is a selling point. But still, we all take these things with a pinch of salt don't we? It won't necessarily know weight (in terms of optimal for the individual), shoe choice, course nature, psychological factors etc. HR + training volume + training types will give a decent perspective on it, and it's getting better all the time. But still nearest 20s for 5k, nearest 30s for 10k, nearest 60s for HM and nearest 120s for marathon would really be a more honest assessment wouldn't it? |
18 Sep
1:46pm, 18 Sep 2024
70,986 posts
|
Velociraptor
I can see how it might think I could shave a few seconds off my 10k race time, @SPR, but a half marathon at just a little slower on average than my 10k pace is difficult to understand. I wonder whether the algorithm "thinks" that if my training pace and heart rate for 10k and 13 miles are similar the same will apply to my race pace for the two distances. The 5k prediction flatters my achievable leg speed. The marathon prediction may or may not be within reach with appropriate training. |
|
18 Sep
1:58pm, 18 Sep 2024
46,165 posts
|
SPR
We take it with a pinch of salt because we see issues with accuracy, if it worked well, it would be taken more seriously (and some that find it accurate do). No one is asking for magic (clearly if all my runs are at 9 min miling it's could struggle to nail a significantly faster 5km prediction) but with the data it has, it can do better (eg if I run 10km XC at a certain speed, that shouldn't be my 5km prediction) and they may have made it better. Saying it's a range assumes some level of accuracy which some of us (I for one) know it didn't have before. It certainly seems to me that a predictor with all your data should be closer than 20 secs to your 5000 time (especially after you've run one) or have some clear logic why. |
18 Sep
2:16pm, 18 Sep 2024
17,530 posts
|
chunkywizard
SPR, I have a 965 so it would be using the new algorithm. I am doing track every week (so it should have a good idea of my VO2Max, L/T) as well as I am running parkrun hard most weeks so it should know where I am at race pace. I reckon I could maybe go 10-20s better than my SB with the right course, weather etc but not 1:20.
|
|
18 Sep
2:36pm, 18 Sep 2024
46,166 posts
|
SPR
Will be interesting what it does when you run both 5k and 13 miles at higher intensities @Raptors claws are coming to town. Interesting to note that McMillan predicts slower for 5km and HM off a 10km of 49:33 @chunkyblizzard interesting. As someone doing regular parkrun, you'd expect it to work well. Maybe check HR max (they have sports HR and also if it's a new watch auto detect may have changed what you previously set) is accurate but other than that, I guess it's another example of things not quite working. |
|
18 Sep
2:39pm, 18 Sep 2024
17,533 posts
|
chunkywizard
SPR, my max HR is set correctly. Maybe it's dropped by 1 or 2, in the years since I last hit it, but it's relatively accurate I believe [goes off to check this years race/ intervals to check...]
|
|
18 Sep
2:45pm, 18 Sep 2024
17,534 posts
|
chunkywizard
I got to 169 on a track 3K this year and my max is in at 172, so it's not too far off.
|
|
18 Sep
2:49pm, 18 Sep 2024
17,535 posts
|
chunkywizard
Interestingly my 3K time was 12.04 and the race predictor is say 19:24 so I don't know where I'm going to find 8 secs/km extra
|
|
18 Sep
2:52pm, 18 Sep 2024
25,633 posts
|
larkim
It all just feels a little pointless. Run a 5k, read a table of comparable 3k or 10k times. That's good enough surely. It's overcomplicating things which are only really relevant if you're not actually ever doing any racing.
|
Related Threads
-
Garmin 735xt problem
Jan 2021
-
Garmin and Windows 10
Jun 2016
-
Help pls - what’s latest equiv to garmin vivoactive HR?
Jan 2024
-
Inporting Garmin historic data
Dec 2023
-
Garmin Connect IQ apps and widgets etc
Jan 2023
-
Polar Precision data
Jan 2022
-
PSA - Potential GPS watch stocking filler bargain?
Sep 2021
-
Running Apps Discrepancies
May 2021
-
Are replacement ANT sticks really *that* much?!
Sep 2013
-
Kindle
Dec 2024