Page 5 of 11

PostPosted: Fri Jan 09, 2009 7:48 pm
by Switch
trell wrote:Probably i am missing something..How set up the start date? I can always see all my history, but i desire see only for ex from 01/01/09 :roll:

By default, the current version of the plugin looks 4 months back from the current date. Cutting of at 01/01/2009 would actually reduce your CTL, as it looks back 42 days (defined by the CTL setting.

PostPosted: Sat Jan 10, 2009 2:20 am
by HENNES
Switch wrote:
HENNES wrote:spot on!

Well, almost!


No, not me or my training - I meant your analysis was brilliant: spot on!

End of May your TSB was even higher.


12 days after the marathon, may 16th I ran 10k in PB and was very suprised about the time (but that generally happens 2-4 weeks after a marathon)

In other words, more tapering.


Brilliant analysis! Exactaly what I wanted to hear (but didnt tell anything) as I figured that out after the race. I was totally exhausted right from the beginning and therefore have adopted some changes to my new plan for Rotterdam april 5th:

- weekly km peak 3 weeks before the marathon
- tapering starting 3 weeks before the marathon

Is that "in line" with your findings?

This should be reflected by changing your ATL setting from 11 to something like 14, 15, 16.


I think I really should invest some time to study this, as you found out some real things without knowing what I did, I must be able to finetune much better with knowledge of this nice plugin and my training!

HENNES wrote:But its probably more interesting with the correct settings Ive hopefully applied now.

I do not understand why you have different settings for TRIMP and for your ST Standard HR zones?


I tried to "explain" , I have no clue about that and just fiddled around - which doesnt mean that this must be right - obviously not.

I suggest you adapt your ST HR settings to match your TRIMP settings. Your current ST Standard HR zones are too "flat" in my opinion.


Will have a look how to change that.

Many thanks!

PostPosted: Sat Jan 10, 2009 7:18 am
by mechgt
Switch wrote:I do not understand why you have different settings for TRIMP and for your ST Standard HR zones? I suggest you adapt your ST HR settings to match your TRIMP settings. Your current ST Standard HR zones are too "flat" in my opinion.


Where do the TRIMP plugin HR factors come from? I was trying to figure out where I should get mine from, and originally tried to model mine after the TRIMP plugin; but found some other calculations (originally from mazoaguirre, thanks!) and supported by this:

Modeling Human Performance in Running, Morton, et al wrote:For quantification purposes we have chosen to use [factor] = e^(bx) in accordante with the commonly observed exponential rise in blood lactate with exercise intensity x reflected by the HR ratio. Values for b were chosen to match the shape of the increment curve in blood lactate concentration (in mM) with increasing work rate and heart rate in men and women as reported by Green et al. (6). These data demonstrate that the male and female responses are sufficiently different to warrant description by separate b values for men (1.92) and women (1.67).


There's a tad more to the factor equation, but this is what it's based on. The HRratio ('x' above) is based on the HR zone, and your max & resting HR. Because the equation for these factors is based on HRmax & HRrest (which will change over time) in an upcoming release, there'll be an option to use those HR values from the date of activity, but it won't work with manually defined factors, you'll have to use the equation.

I guess in all this, I just wanted everyone to know where the factor equation came from and that I didn't just make them up (doesn't mean they're right and others are wrong, just means they actually came from somewhere.)

NOTE: Everybody gets the male factors for now (sorry girls, there's no male/female athlete setting in ST, we need to request this!)

PostPosted: Sat Jan 10, 2009 8:45 am
by HENNES
@mechgt, as I spoiled a bit my HR-zone with fiddling around and there is no "reset" in STs and you looked into that topic, what would recommend to set up there. Same zones like in TRIMP`?

Do those make any sense???



Image



60% HFmax or 80% make a lot of difference in training, however with factor 1.1 and 1.2 they are very similar treated. Looks strange/wrong to me...

EDIT: just realize that TRAINING LOAD uses my zones which are more spread across and therefore apply better factors - to my believe. Maybe HRzones should be even further split apart in steps of 5BPM ????


Image

PostPosted: Sat Jan 10, 2009 8:58 am
by mechgt
HENNES wrote:@mechgt, as I spoiled a bit my HR-zone with fiddling around and there is no "reset" in STs and you looked into that topic, what would recommend to set up there. Same zones like in TRIMP`?


What I do is use a SingleZone in TL - maybe a TRIMP heart rate category.

I'd to go into STs HR categories (Select View > Categories > Heart Rate Zones) and create a TRIMP heart rate category. Give it a lot of zones... mine has 10 zones starting at 0 - 60% (zone 1), and incrementing by 5% -> 0-60%, 60-65%, 65-70%, etc. Feel free to make as many zones as you want: the more zones, the better resolution.

Then go into TL and use this new TRIMP heart rate zone as your SingleZone. I'm a fan of TL's default factors, but you can setup your profile as you see fit.

This gives your TRIMP calculations very fine resolution, but you don't have to look at all these HR zones in other views.

PostPosted: Sat Jan 10, 2009 9:46 am
by HENNES
mechgt wrote:I'd to go into STs HR categories (Select View > Categories > Heart Rate Zones) and create a TRIMP heart rate category. Give it a lot of zones... mine has 10 zones starting at 0 - 60% (zone 1), and incrementing by 5% -> 0-60%, 60-65%, 65-70%, etc. Feel free to make as many zones as you want: the more zones, the better resolution.


Ah, I see we are in line with my EDITED comment from above. With my categories I didnt have % but numbers to fill in. I wonder whether it wouldnt be better to use log changes instead of linear changes. The change 187-5bpm is obviously very different than 125-5bps - or is this already convered by the very different factor weightings. Im not clear on this....(aswell :lol: )

Then go into TL and use this new TRIMP heart rate zone as your SingleZone. I'm a fan of TL's default factors, but you can setup your profile as you see fit.

This gives your TRIMP calculations very fine resolution, but you don't have to look at all these HR zones in other views.


Sounds good - and done: looks like this and I assume all charts and graphs are different now, right?



Image



thanks

PostPosted: Sat Jan 10, 2009 9:52 am
by mechgt
HENNES wrote:...I wonder whether it wouldnt be better to use log changes instead of linear changes. The change 187-5bpm is obviously very different than 125-5bps...

This only changes where you want the most resolution in your data. This is less important than the factors - that's where the magic happens. What you have below looks great in my opinion.
HENNES wrote:Sounds good - and done: looks like this and I assume all charts and graphs are different now, right?

Yes, when you go to the TL view, your charts will reflect the new settings.

PostPosted: Sat Jan 10, 2009 10:26 am
by TX911
HENNES wrote:Image


I'm not sure, but I believe the curve should be more exponential than linear. Thus, as you approach higher and higher heart rates, the run beomes exponentially more difficult. i.e., it is much harder to go from 85 to 95% heart rate than 55 to 65%. Does this make sense?

PostPosted: Sat Jan 10, 2009 10:47 am
by HENNES
TX911 wrote:I'm not sure, but I believe the curve should be more exponential than linear. Thus, as you approach higher and higher heart rates, the run beomes exponentially more difficult. i.e., it is much harder to go from 85 to 95% heart rate than 55 to 65%. Does this make sense?


Makes sense and is what I said above - but the answer from the boss is true too:
the faster increasing factors do solve this problem. See the factors!

See the steepening curve in my last post: Posted: Sat Jan 10, 2009 9:46 am

PostPosted: Sat Jan 10, 2009 12:54 pm
by mechgt
TX911 wrote:I'm not sure, but I believe the curve should be more exponential than linear. Thus, as you approach higher and higher heart rates, the run beomes exponentially more difficult. i.e., it is much harder to go from 85 to 95% heart rate than 55 to 65%. Does this make sense?

The factors are in fact exponential, it's just not as accentuated as what you may have been expecting. You're still getting 2.5 - 3x the 'points' for a minute at 90% HR vs 60% HR. Note that the percentages here are % from HRrest to HRmax not 0 to HRmax.

PostPosted: Sat Jan 10, 2009 4:15 pm
by HENNES
mechgt wrote:Note that the percentages here are % from HRrest to HRmax not 0 to HRmax.


Why is this and can you tell why this is more right or important? (only if the answer is 2-5 sentences and not an essay) :evil:

PostPosted: Sat Jan 10, 2009 5:33 pm
by TX911
mechgt wrote:The factors are in fact exponential, it's just not as accentuated as what you may have been expecting. You're still getting 2.5 - 3x the 'points' for a minute at 90% HR vs 60% HR. Note that the percentages here are % from HRrest to HRmax not 0 to HRmax.


You're right, the default factors are exponential. I just switched back to the default factor values using a 10-zone HR reserve, and I'm happy with it.

I was just looking at the curve of HENNES, and it looked fairly linear.

PostPosted: Sat Jan 10, 2009 9:17 pm
by mechgt
HENNES wrote:
mechgt wrote:Note that the percentages here are % from HRrest to HRmax not 0 to HRmax.


Why is this and can you tell why this is more right or important? (only if the answer is 2-5 sentences and not an essay) :evil:


It was just a clarification of what I was writing. No one is interested in HR from 0 to resting (unless you're dead-ish!) so it makes sense to only measure and analyze your full operating range.

PostPosted: Sat Jan 10, 2009 11:19 pm
by Gunnar
ildibad wrote:By design, rolling trimp on 7 days should be very similar to ATL and rolling trimp on 28 days not very far from CTL.

It would seem at first that this may be, but ATL and CTL are weighted averages. The rolling TrImp totals are more of a way to keep in check that you don't overtrain, in a specific block of time.

ildibad wrote: Your post was interesting for because I used to display "only" CTL, TSB and forecasted influence.

Your explanations tried to convince me that i needed to display also ATL, because it represents the workload I can handle.

But even without displaying ATL, one should be able to manage the fatigue vs fitness by managing the variation of CTL and keeping under control the TSB.
The Ramp rate is just interesting information, but the rolling TrImp total is really just another metric to keep an eye on.

ildibad wrote: Sorry to contradicting you, but I try to keep my graphs as simple as possible (keeping my brain below the steaming point ;-) )
It's only a suggestion, as there are always those of us who would like other information than what is currenty provided. If I knew anything about programming I would attempt to write a plugin myself. And really for me, I'm not really interested in charting the rolling TrImp (even though I do in my spreadsheet), I'm most interested of the number and it's approch to the Breakeven point.

:D

Again, great work on this already wonderfully usefull plugin.

And for those of you who love the information that this plugin gives you, but don't really know what it all means, I have also found that the Wattage google group ( http://groups.google.com/group/wattage?hl=en ) has a ton of info about TSS,CTL,ATL,TSB and a lot more. You need to request membership, and it's geared a bit more for training with power, but it has proven very helpfull.

PostPosted: Sun Jan 11, 2009 1:40 am
by racerfern
There seems to be a discrepancy somewhere. I can supply data if needed. The following charts are all from the same data, one is the training load, one is the training planner and the other one is the Performance Manager Chart from WKO+.

While it is difficult to pinpoint differences, you can see that both the Training Planner and the PMC chart have a significant drop in CTL in early November followed by a seesawing CTL after that.

However the Training Load plugin has a significantly increasing CTL after the drop in November.

Any ideas?

Image

Image


Image

PostPosted: Sun Jan 11, 2009 6:28 am
by gerhard
racerfern wrote:However the Training Load plugin has a significantly increasing CTL after the drop in November.

Any ideas?


Is the results similar with Single Zone split in many parts?
(I saw some unexpected results when changing training that happened to just be in a certain zone.)

TL uses TSS from TRIMP where TP/WKO+ uses TSS based on NP. If you have a different type of training, the result will be different. With TP I get larger differences, but then the Power when running is estimated.

PostPosted: Sun Jan 11, 2009 6:39 am
by gdean
Is it possible to use Training Load to visualise effects of a planned training period. E.g. I'd like to add some 'dummy' activities for next week (or longer), add some nominal TRIMP scores against them, and then see their effects on CTL/ATL... ??

PostPosted: Sun Jan 11, 2009 7:44 am
by mechgt
racerfern wrote:...both the Training Planner and the PMC chart have a significant drop in CTL in early November followed by a seesawing CTL after that.

However the Training Load plugin has a significantly increasing CTL after the drop in November.

Any ideas?


It looks to me that there is a different starting point. All three charts are tending toward a CTL value of ~65ish, except for whatever reason, TL is starting around 0, and the other two are starting from a higher value.

PostPosted: Sun Jan 11, 2009 7:53 am
by gdean
gdean wrote:Is it possible to use Training Load to visualise effects of a planned training period. E.g. I'd like to add some 'dummy' activities for next week (or longer), add some nominal TRIMP scores against them, and then see their effects on CTL/ATL... ??


I've hacked this for now by simply changing the date on my computer, then Sport Tracks displays everything upto 'current' date and hence the Training Load graph does.

PostPosted: Sun Jan 11, 2009 7:54 am
by mechgt
gdean wrote:
gdean wrote:Is it possible to use Training Load to visualise effects of a planned training period. E.g. I'd like to add some 'dummy' activities for next week (or longer), add some nominal TRIMP scores against them, and then see their effects on CTL/ATL... ??


I've hacked this for now by simply changing the date on my computer, then Sport Tracks displays everything upto 'current' date and hence the Training Load graph does.


I'll change TL so that it does not cut off at today, so that if people put future activities in, it'll catch them.

PostPosted: Sun Jan 11, 2009 8:02 am
by racerfern
It looks to me that there is a different starting point. All three charts are tending toward a CTL value of ~65ish, except for whatever reason, TL is starting around 0, and the other two are starting from a higher value.


That's it! I discovered this because Gerhard made me look at the zones. At that point I realized the starting CTL and ATL points were at zero. I adjusted those variables to 80ish and it's fine now.

Then I went to post and realized another posting had been made by mechgt, which nailed it. Thanks guys.

Those variables don't always get remembered.

PostPosted: Sun Jan 11, 2009 8:06 am
by mechgt
racerfern wrote:...
Those variables don't always get remembered.


I believe you use multiple logbooks... right? All of the TL settings are stored in the logbook, so just remember that you'll have to set them in each logbook.

PostPosted: Sun Jan 11, 2009 9:17 am
by gdean
mechgt wrote:
gdean wrote:
gdean wrote:Is it possible to use Training Load to visualise effects of a planned training period. E.g. I'd like to add some 'dummy' activities for next week (or longer), add some nominal TRIMP scores against them, and then see their effects on CTL/ATL... ??


I've hacked this for now by simply changing the date on my computer, then Sport Tracks displays everything upto 'current' date and hence the Training Load graph does.


I'll change TL so that it does not cut off at today, so that if people put future activities in, it'll catch them.


Thank you - very useful. The TL plugin has been extremely useful to me, thanks again.

PostPosted: Sun Jan 11, 2009 10:54 am
by gerhard
Exception:
* View certain category
* Selected certain activity older than 4 months, zoomed out
* Selected Settings - no changes
* Pressed back. Default zoom (OK), but table is empty (should not be like this). Calender marked.
* Click table header - get exception below. If graph is selected instead, table is back

Workaround: Just press Continue

************** Exception Text **************
System.NullReferenceException: Object reference not set to an instance of an object.
at TrainingLoad.UI.View.ViewTrainingLoadPageControl.treeActivity_ColumnClicked(Object sender, ColumnEventArgs e)
at ZoneFiveSoftware.Common.Visuals.TreeList.ColumnsImpl.OnMouseDown(MouseEventArgs e, Rectangle drawRect)
at ZoneFiveSoftware.Common.Visuals.TreeList.OnMouseDown(MouseEventArgs e)
at System.Windows.Forms.Control.WmMouseDown(Message& m, MouseButtons button, Int32 clicks)
at System.Windows.Forms.Control.WndProc(Message& m)
at System.Windows.Forms.ScrollableControl.WndProc(Message& m)
at System.Windows.Forms.ContainerControl.WndProc(Message& m)
at System.Windows.Forms.UserControl.WndProc(Message& m)
at System.Windows.Forms.Control.ControlNativeWindow.OnMessage(Message& m)
at System.Windows.Forms.Control.ControlNativeWindow.WndProc(Message& m)
at System.Windows.Forms.NativeWindow.Callback(IntPtr hWnd, Int32 msg, IntPtr wparam, IntPtr lparam)

PostPosted: Sun Jan 11, 2009 11:03 am
by gerhard
I have divided the HR zones so each zone is just a couple of beats to reduce margin effects. When studying similar activities (distance, route, TSB, fitness) but different effort (time, average HR), the TSS seem to relate more to the distance than the perceived effort. It therefore seems like I should have steeper factor. I have played around a little with the factors, but getting the exponential curve manually is a little hard.

Can the steepness of the curve be customized or is the factors scientific?