Saturday, April 26, 2008

kayry vs irunforbeer vs sparky heartrate data

I just learned something important - I'm different :-) (many have been telling me this for years - now I think I believe them). I was comparing heartrate data vs mileage for myself vs a couple other runners at Boston. Irunforbeer (John) ran Boston at 2:46 and Sparky ran Boston at 2:40 while I ran it at 2:42. Below is a comparison of our heartrate data normalized to try to take out the hill variables and also normalized to the average heartrate of the miles represented. The normalizing didn't work too well for the very hilly miles 16 thru 21 so I took that data out as well as a couple other outliers. (click on graph to make it bigger).

I find this to reveal a very important flaw in my marathon pace strategy for the last 4 marathons. Unlike irunforbeer and sparky - I speed up as the race goes on as I try to maintain the same heartrate while they in contrast would slow down if maintaining constant HR. I have end up failing at some point and dropping off the HR target for the last 4 marathons. At NYC I had to pull back my effort around 18, at Houston around 16, at Seabrook at 14 (a training run so I didn't push it that hard) and at Boston around 16. This now helps me understand why - I'm speeding up as the race goes on and eventually I get to a speed I can't handle. Next race I think I will try to ratchet down the heartrate target every 4 miles or so based on the slope of the data and see if I can keep the pace more constant to the end and avoid the dropping off target.

I thought I would look at this historical for my past marathons - all seem to give me the same indication - HR slows down with distance - I sure wish I'd connected this together before Boston.

I knew this before but I had always discounted it thinking a large carb load pre-race was causing it but both Boston and Seabrook I didn't do the large carb load yet same results. Live and learn - but now that I see this - it makes my race day strategy much different - no longer constant heartrate beginning to end - start at a target and notch the heartrate target down every 4 miles or so appears the more even paced strategy for me.

This also convinces me I'm not as fast as I thought - I thought I might be good for a 6mpm pace - if I had followed this declining HR target at Boston starting 165 and down 1 bpm per 4 miles I would have gone 2:27 faster (exactly 2:40:00 as it happens) - fueling/hydration may still have been a factor at Boston but now I'm thinking my race strategy had me pour to much effort into the race in the middle miles which cost me in the latter miles. At Houston using same target - I would have been 1:25 faster (2:39:21) I don't think I am quite 6 mpm ready for a marathon......yet.

Summary / Bottom Line Racing by heartrate is not as straightforward as I thought - can't just pick one heartrate and stick to it from beginning to end - can't even learn from others it would appear as there appears wide variation in how the ticker works between different runners. With constant effort being the objective (ie for a flat course - same pace every mile) - my heartrate goes down over the miles - for others - it goes up over the miles. I don't know why there is a difference and in the end I figure it doesn't really matter all that much - I just need to know how mine works if I'm going to use it as a tool in racing/training - which I am.


Gavin Boyles said...

sparky here - interesting stuff, kayry! For what it's worth, at Boston I only used the HRM as a ceiling. That is, I would look at it occasionally, and if it was over 170 I'd think about slowing down. Later in the race, I was feeling good, so I changed that mental threshold to more like 172. In the very early miles (1-5 or so) I tried to keep it under 170 by a decent margin. Interesting to compare our respective first-mile splits at Boston - I was 6:20!

squirrel1.1 said...

well if there's a way to find error in going from 2:56(?) to 2:42 in a few months, you've certainly nailed it. :)

Great analysis!

kayry said...

Hey sparky

First mile for me I did not holding back on that downhill - and it felt ok - 5:23 - it seemed ok at the time and on an equal effort it is about what I was expecting but maybe it was a touch fast.

Looking at your splits it looks to me like you ran a very good equal effort beginning to end race adjust splits for hills. The graph above was not your actual splits but trying to adjust to IF you had run at a constant heartrate (assuming 3 sec/mile per bpm off average) - you would have slowed far to much at the end but you were smart in the race to let the feel over-ride the heartrate info you were getting and run at a higher heartrate. For example - Mile 24 has your highest heartrate at 90% max but your pace (6:06) for that near flat mile was right about average for the race - had you tried to keep the heartrate you averaged for the race that mile you would have slowed unnecessarily to more like 6:15-6:20 pace. Incrementally increasing your heartrate target thru the race appears the right answer for you ~1 bpm per 2 miles by my graph.

I appear to need to reduce my HR target thru the race or I end up running too fast. I'm kinda excited to figure this out finally as I had felt like I was not powering mentally hard enough to maintain to the finish for the last several races and now I can see I was just being too hard on myself. I was inadvertantly speeding up about a second per mile until I was just running too fast. At Boston I went in resolved to push to maintain the target better than I did at Houston which was exactly the wrong answer and cost me a little. Ended up putting 1-1.5 minutes of greater effort into the first 16 miles as I more strictly maintained my heartrate target - ignoring my bodies signals it was too fast - especially ~10-16 miles in - which resulted in bonking a little across the hills.

Anyway - great race out there - I think you ran it about perfect and 17 minute PR is AMAZING!!! It beats my 16 minute PR at Houston in January.

I know that 4 seconds from 2:40 is going to knaw at you though.....for me it's 46 seconds from 2:40 and it sure is knawing at me :-).

kayry said...


Thanks squirrel - Always trying to improve :-).

I'm thinking I've inadvertantly received lots of benefits from training at constant HR now that I see the pace has been steadily increasing.....For example my long runs tend to finish faster than they start because I'm trying to maintain constant heartrate - or many times I try for fast finish last few miles and try to bump up the heartrate to my beginning marathon heartrate - which actually is probably nearly my Tempo heartrate at the end - so in otherwords - I've been sneeking in more quality into the end of my workouts than I thought which probably has benefited me over the last year+.

irunforbeer said...

Hi Kayry, Interesting discussion as always. I wanted to look at my 5K splits and found a free program called SportTracks
which allows you to import Garmin files and choose different splits (including custom splits) and charting options. On chart is called accumulated pace which reminded me of this topic. Anyway, you might want to try it out since you are an advanced data cruncher.


kayry said...

Hey Mr. irunforbeer,

I've got a claim to fame for you - you have the fastest veteran (>50)finishing time from California for the whole decade at Boston. As it happens I've got the fastest Master's time (>40) from Texas for the decade. Since race results aren't on line prior to 2001 there's no way I know to check prior to that so I suppose it's possible we have those titles for all time but somehow I doubt it for me - but could very well be true for you :o).

RE Sport Tracks - I've been using that for over a year - it is my "official" logbook as I also add my non-Garmin runs into it (treadmill or runs w/o Garmin or periods of a run sometimes if I forget to turn it on or pause it by accident or something). I like its features of being able to fix heartrate data - can delete or change the raw data - also can split apart a run which is sometimes handy. I wish you could cut and paste data into excel in it or could create printouts but can't do that - have to use MotionBased for that - and of course Motionbased tacks on the weather too which I like for that.

Great Race at Boston - your LSD approach to running has me really thinking of this strategy over the summer - after Anchorage - maybe - we'll see.


irunforbeer said...

Hey 'Mr. Fastest Boston Master from Texas of this Century",

Cool that you looked up those factoids, it would be great if the records went back farther. Maybe Boston Marathon has a 'historian' for requests like this?

I'm liking SportTracks a lot. It has some nice features as you say. Cheers from Team John!

kayry said...

Hey John - I sent baa an email with the question - I'll let you know the response. I mentioned how important it was to have the facts correct to properly engrave beer mugs :-).

I noticed yesterday I was using the old version of sporttrack so I just downloaded the new version yesterday - looks like there are more functions and better graphics - I like the improvements!