Specific 2k Pacing (Per 500m)

General discussion on Training. How to get better on your erg, how to use your erg to get better at another sport, or anything else about improving your abilities.
Posts: 1
Joined: April 13th, 2006, 12:21 pm

Re: Specific 2k Pacing (Per 500m)

Post by bigsteve964 » April 13th, 2006, 12:26 pm

RR1 Kirk wrote:
Adrenaline wrote:I'm hoping to get low 7:30s or even break 7:30 in a 2k test on Wednesday, but I've never been good on planning out how to pace for something like this. I'm a small guy, so i don't know if that has to do with pacing at all.

I'm trying to figure out whether my first 500 should be something fairly fast, like 1:49s or if I should just hold a steady 1:50/51 for the first 1500 then pick it up in the end?
Remember that you only have to average 1:52.5 to hit 7:30. I would settle into my planned pace after 100 m or so, keep to it and then hit my sprint at about 200 to go.

Treat it as a set piece, not a race and stick to your plan. Learn from whatever happens.

Good luck
I agree wholeheartedly. For my 2ks i start with a hard (-10 2k pace or so) power ten and then i settle into my pace for the rest of the piece until the last 400m when i start to bring the split down and then a final sprint at about 250m to finish it up strong. It really helps me to do the piece this way but it may be just me.

Good Luck!

500m Poster
Posts: 61
Joined: March 16th, 2006, 2:41 pm

Post by rlholtz » April 14th, 2006, 1:17 pm

Wouldn't mixing it up a little, altering pace and spm, give certain musculature periods of respite resulting in a faster time? Or am I totally wacked?

Posts: 5
Joined: March 31st, 2006, 1:27 pm

Post by HH » April 15th, 2006, 6:07 pm

To the guy who says don't nail it off the start: Graham Benton hits 1:08's of his 2k start and he pulls a bloody fast time: if your hard enough go for it!

Posts: 8
Joined: March 20th, 2006, 1:15 pm

Post by ivanans » April 20th, 2006, 5:19 pm

Hmm... I'm always doing that fly & die technique and I haven't some special problems, although it's a lot easier if I keep it relatively steady (-3 first 500m, -2 at last 300m, not big jumps) and I always break it

Posts: 19
Joined: March 16th, 2006, 1:41 pm
Location: Fremont, CA, USA

Post by dougsurf » April 20th, 2006, 7:23 pm

Clarification on Fly 'n Die:

Sometimes the descriptions of fly 'n die make it sound like any kind of up front, fast start it that. But some very renowned rowers/ergers successfully start out strong, settle for the middle, and then hopefully have something left for a sprint.

I would characterize true fly 'n die this way. It is the most enthusiastic of novices who just can't convince himself that he's pulling hard enough, unless he is feeling discomfort the whole time. I've seen it often, in masters and juniors alike, in their first year. They just blast away as hard as they can, until it hurts from the get go, and then all of the things Mike C. describes above happen. They start way over, like 15 seconds faster than they should. And then they follow a curve of equal consistent pain, and straight line or downward curving decreasing speed often diminishing to a pace well over 15 seconds slower at the end. Hopefully they learn from the experience, with their coach's help.

That is distinguished from a more thoughtful but still fast start, that many use. One is Xeno Muller, who I'll let explain:

" This is how I race a 2000 meter Indoor Rowing Race on the Concept 2 Rower, CRASH-B or Long Beach Sprint 2006

On the ergo you have immediate feedback what your pace is. This is not the case in racing on the water. On the water it was by feel and race experience. On the ergo you aware of every split time and second you pull.

So this is how I personally row a 2K.

First I determine what I am capable of for 2k, by rowing a few race pace 1000 and 500m, and past ability and my current endurance ability for steady state.

In 2004 I pushed 5:53
In 2005 6:02 ( I believe)

For the sake of simplicity I will show the race plan for a 6:00 2K

First eighteen strokes bring the average per 500 meters down to 1:27ish without overdoing it and using adrenaline. Be very carful not to be blinded by adrenaline, if you do that mistake you explode somwhere between 1200m and 800 meters to go.

After the first 18 strokes find your race pace, which DOES not have to be 1:30, you have a head start because of the start. So I would push 130-131 occasionaly seeing 1:32. I am carfully monitoring the total average split per 500. So slowly the "start-lead" fizzels down to a total average of 1:30. This occured to me at 900 meters to go. So now it is down to buisness. I couldn't immediately adjust to maintaining 1:30 constant overall average so I lost in overall average 0.4 seconds bringing it to 1:30.4. The further you go into the piece the less the total average per 500 meters fluctuates! Once I find the pace at exactly and consistantly 1:30 I am at roughly 650 meters to go. Cool I see the end of the tunnel, yeah! I also do not feel totally tired either because I chose the right pace and only had to row the 1:30 for roughly 500 meters. So by the time I hit 350 meters to go I sprint because I know it is roughly thirty stroke. That is psycho babble of coursem, because it is going to be more like 38 strokes, but who cares about an extra 8 strokes when you only have 50 meters left. Believe me, when the sprint is set up right, for the remaining 350 meters, YOU CAN DROP YOUR OVERALL AVERGE by 0.5 seconds or even 0.8.
I hope this helps.
I realize that this is different from negative splitting. I am NOT a negative splitter.
We should have a get together at the IRON OARSMAN one day and eat at the outback stake house down the street.

Hope this is of interest.

- Doug

Mike Caviston
1k Poster
Posts: 167
Joined: April 20th, 2006, 10:37 pm
Location: Coronado, CA

Post by Mike Caviston » April 20th, 2006, 10:45 pm

Clarification on Specific 2K Pacing:

It is more effective to start with an opening 500m very close to your final average pace. I believe a slight negative split is most effective (i.e., only a couple tenths of a second per 500m slower than your final average), though even-splitting or a very slight positive split can be effective as well. The faster you start relative to your potential, the worse you will ultimately do.

For the past several weeks I have been researching pacing strategies used by athletes in various sports at the Olympic & World Championship level. I have gathered data from men’s & women’s 1500m running, men’s & women’s 400m & 800m swimming, men’s 4K & women’s 3K individual pursuit cycling, and men’s & women’s 5K & women’s 3K speed skating. I have results from every A-Final for Olympic & World Championship regattas from 1996-2005. I’ve also looked at results from CRASH-B & BIRC for multiple categories from various years, concentrating mostly on the men’s & women’s Open events. My basic model has been to look at each quarter of each race as a percentage of final average velocity (or Watts for indoor rowing). So, someone who does an extreme “Fly-and-Die” might do the first 500m @ 108% of final Watts, the second 500m @ 105%, the third 500m @ 96%, and the fourth 500m @ 91%. Someone with a more balanced strategy might go 99.5%, 100%, 100%, and 100.5%.

The results have been consistent and conclusive. Different sports have different characteristic pacing profiles, but in every single sport or event I’ve looked at, the “winners” (top half of the field) have started more conservatively than the “losers” (bottom half). Sometimes I’ve used individual races (running, cycling, skating, indoor rowing) and sometimes I’ve pooled results from different races (swimming & outdoor rowing) to get enough statistical power. For some events, the correlation coefficient between slow starts and higher finishes reaches -.9 (i.e., almost perfect correlation between starting slower and finishing faster).

For OTW rowing, it’s not possible to show the potential effectiveness of negative-splitting because only a tiny fraction of crews actually race that way (the average opening 500m being about 103% of final average velocity). But across all events, the medal winners start (relatively) more slowly than places 4-6. For indoor rowing, it is easy to show statistically that those who use a negative-split approach average faster times than those who don’t. And it’s easy to quantify the Fly-and-Die approach – the more you Fly, the more you Die.

I have been compiling many tables & graphs as I go along to illustrate all these relationships. I am willing to share, but I’m not going to take the time and effort when it appears my potential audience has already made up its mind. It’s true that some people have had remarkably fast times with an aggressive start. I congratulate them, but would also respectfully suggest they would be even faster with a more balanced pacing strategy. I have to chuckle at those who say “So-and-so has a monster start, and they manage to win!” That’s like saying that your great uncle Ernie smoked two packs a day and lived to be 92, so that proves smoking isn’t harmful. Look at the big picture.

Mike Caviston

User avatar
6k Poster
Posts: 772
Joined: March 16th, 2006, 11:25 am
Location: castelldefels - barcelona

Post by ancho » April 21st, 2006, 4:53 am

Mike, welcome back!
I'm glad you still are sharing your informations.
Thank you very much!!
yr 1966, 1,87 m, 8? kg
Be Water, My Friend!

6k Poster
Posts: 775
Joined: March 16th, 2006, 10:48 am

Post by TomR » April 22nd, 2006, 10:42 am


I hope you're going to show us at least some of the charts/tables that depict pacing in indoor races. Facts on the Forum are a bit of an oddity, but I, for one, would like to see the results of your research.


Mike Caviston
1k Poster
Posts: 167
Joined: April 20th, 2006, 10:37 pm
Location: Coronado, CA

Post by Mike Caviston » April 22nd, 2006, 7:35 pm

Just waiting to be asked.

To summarize my basic premise, I’m saying that for middle-distance events that require substantial energy from both aerobic and anaerobic sources (such as 2000m rowing or other events that last roughly 3 ½ - 8 minutes), the most effective pacing strategy is to complete the first portion of the race at a pace that is no faster than the final average pace. In other words, negative-splitting is more effective than otherwise. Furthermore, the positive-split or “fly-and-die” approach will have negative consequences in proportion to how far beyond average pace the initial portion is completed. The physiological rationale is as follows. Performance is affected by energy production (more is better; a positive effect) and fatiguing byproducts of metabolism (K+, H+, NH3, etc.; these have a negative effect). The key to optimal performance is to use a pacing strategy that maximizes energy production while keeping fatiguing agents within tolerable limits. Too much initial intensity disproportionately accelerates the fatigue process to the extent that energy production in the latter portion of the race becomes severely compromised.

I know the benefits of pacing not only from my own extensive racing experience, but from working with and charting the performances of hundreds of athletes over the past several years. Collegiate athletes typically do three 6K and three 2K tests per season. By working with some athletes for two, three, or four years, I have been able to see some individuals perform more than twenty tests over their careers while noting such things as the training they did preceding the tests and the pacing strategy they used during the tests. Well, guess what – the vast majority have performed better when negative-splitting or at least being more conservative at the start. While I am talking specifically about indoor ergometer tests here, the same phenomenon applies to races on the water.

A couple months ago I was reviewing some published literature on Pacing Strategies:

Code: Select all

Sandals, 2006:  “Influence Of Pacing Strategy On Oxygen Uptake
   During Treadmill Middle-Distance Running"
   International Journal of Sports Medicine, vol. 27, pp. 37-42
Albertus, 2005:  “Effect Of Distance Feedback On Pacing Strategy 
   and Perceived Exertion During Cycling”
   Medicine & Science In Sports & Exercise, vol. 37, pp. 461-468
Foster, 2005:  “Regulation Of Energy Expenditure During Prolonged
   Athletic Competition”
	Medicine & Science In Sports & Exercise, vol. 37, pp. 670-675
Garland, 2005:  “An Analysis Of the Pacing Strategy Adopted By Elite
   Competitors In 2000m Rowing”
	British Journal of Sports Medicine, vol. 39, pp. 39-42
Rauch, 2005:  “A Signalling Role For Muscle Glycogen In the Regulation Of Pace During Prolonged Exercise”
   British Journal of Sports Medicine, vol. 39, pp. 34-38
Ansley, 2004:  “Anticipatory Pacing Strategies During Supramaximal Exercise Lasting Longer Than 30s”
	Medicine & Science In Sports & Exercise, vol. 36, pp. 309-314
Ansley, 2004:  “Regulation Of Pacing Strategies During Successive 4-km Time Trials”
	Medicine & Science In Sports & Exercise, vol. 36, pp. 1819-1825
Foster, 2004:  “Effect Of Competitive Distance On Energy Expenditure During Simulated Competition”
	International Journal of Sports Medicine, vol. 25, pp. 198-204
Hill, 2004:  “The Relationship Between Power and Time To Fatigue In Cycle Ergometer Exercise”
	International Journal of Sports Medicine, vol. 25, pp. 357-361
Foster, 2003:  “Pattern Of Energy Expenditure During Simulated Competition”
	Medicine & Science In Sports & Exercise, vol. 35, pp. 826-831
Thompson, 2003:  “The Effect Of Even, Positive and Negative Pacing On
   Metabolic, Kinematic and Temporal Variables During Breaststroke Swimming”
   European Journal of Applied Physiology, vol. 88, pp. 438-443
Bishop, 2002:  “The Influence Of Pacing Strategy On VO2 and Supramaximal Kayak Performance”
	Medicine & Science In Sports & Exercise, vol. 34, pp. 1041-1047
Jones, 2002:  “Bioenergetic Constraints On Tactical Decision Making In Middle Distance Running”
	British Journal of Sports Medicine, vol. 36, pp. 102-104
Leveque, 2002:  “Effect Of Paddling Cadence On Time To Exhaustion and
   VO2 Kinetics At the Intensity Associated With VO2max In Elite White-Water Kayakers”
   Canadian Journal of Applied Physiology, vol. 27, pp. 602-611
Billat, 2001:  “Effect Of Free Versus Constant Pace On Performance and Oxygen Kinetics In Running”
	Medicine & Science In Sports & Exercise, vol. 33, pp. 2082-2088
Fukuba, 1999:  “A Metabolic Limit On the Ability To Make Up For Lost Time In Endurance Events”
	Journal of Applied Physiology, vol. 87, pp. 853-861
Liedl, 1999:  “Physiological Effects of Constant Versus Variable Power During Endurance Cycling”
	Medicine & Science In Sports & Exercise, vol. 31, pp. 1472-1477
Palmer, 1999:  “Metabolic and Performance Responses To Constant-Load 
   Vs. Variable-Intensity Exercise In Trained Cyclists”
   Journal of Applied Physiology, vol. 87, pp. 1186-1196
Palmer, 1997:  “Effects Of Steady-State Versus Stochastic Exercise On Subsequent Cycling Performance”
	Medicine & Science In Sports & Exercise, vol. 29, pp. 684-687
Craig, 1995:  “Influence Of Test Duration and Event Specificity On Maximal Accumulated Oxygen Deficit
 	Of High Performance Track Cyclists”
	International Journal of Sports Medicine, vol. 16, pp. 534-540
Foster, 1994:  “Pacing Strategy and Athletic Performance”
	Sports Medicine, vol. 17, pp. 77-85
Foster, 1993:  “Effect of Pacing Strategy On Cycle Time Trial Performance”
	Medicine & Science In Sports & Exercise, vol. 25, pp. 383-388
Ariyoshi, 1979:  “Influence Of Running Pace Upon Performance: 
   Effects Upon Treadmill Endurance Time and Oxygen Cost”
   European Journal of Applied Physiology, vol. 41, pp. 83-91
Robinson, 1958:  “Influence Of Fatigue On the Efficiency Of Men During Exhausting Runs”
	Journal of Applied Physiology, vol. 12, pp. 197-201
In particular, the article by Garland got me thinking. Garland looked at OTW results for HW men & women from the 2000 Olympics and 2001 & 2002 World Championships. He used the format of determining final average speed and calculating each 500m segment as a percentage of the final to determine pacing strategy. Garland concluded there was no significant difference in the pacing strategies used by “winners” vs. “losers”. However, his study utilized preliminary heats as well as finals. He excluded races where competitors appeared not to be giving maximal effort (such as crews that shut down in a qualifying heat), but this was somewhat subjective. Garland also looked at results from the British Indoor Rowing Championship for 2001 & 2002 and again concluded there was no difference between “winners” and “losers”. Not much information was provided about this analysis, except that Garland used time rather than Watts for comparison. Here are a couple of comments from Garland’s paper: “The data presented in this paper do not fully express the magnitude of the increased effort and increased physiological load at the start of the [OTW] race, because only times and speeds are recorded, rather than power output—a superior index of physiological load. The first 500 m section was the fastest despite the inclusion of the initial acceleration from a stationary start. It requires a much higher average power output to complete the first 500 m in 100 seconds, for example, than it does to complete the second 500 m in 100 seconds. …Power outputs to accelerate the boat from standing to race pace are as high as 700 W compared with 350–500 W later on in the race. These additional considerations make the adoption of a fast start strategy even more remarkable.” And: “It should be noted that this study has only shown the self selected pacing strategy used by rowers, but has not necessarily shown the physiologically optimal pacing strategy, nor provided physiological evidence as to why the first 500 m sector is rowed faster than the other sectors.”

I wondered if I could find data to replicate Garland’s approach. I decided to look not only for rowing results but also results from other sports such as running, swimming, cycling, and skating. I was only able to get complete pacing breakdowns for other sports from select years, but what I’ve been able to get has been very useful. I’ve been able to find pretty much all the results from rowing that I could want, with only a few gaps here and there. For OTW rowing, I wanted more data than Garland used for greater statistical power. I also decided to only use results from A-Finals to eliminate as much as possible any concerns about athlete motivation. For indoor rowing, I converted all pace times from seconds to Watts for analysis. Everything I have looked at so far has confirmed my initial hypothesis: faster athletes start their races more conservatively than slower athletes. This can be seen even within the limited confines of an Olympic final.

I gathered results from Olympic running, cycling, and skating at http://www.aafla.org/search/search.htm;
Olympic & WC swimming at http://www.fina.org/main/results/results.htm ;
Olympic & WC rowing at http://www.fisa.org/results/default.sps ;
and indoor rowing at http://www.concept2.co.uk/wirc/results.php .

Besides illustrating my major thesis, a few other interesting things have come to light which I will point out as I find time to show my results. Hopefully I’ll get some results posted in the next day or two.

Mike Caviston

Mike Caviston
1k Poster
Posts: 167
Joined: April 20th, 2006, 10:37 pm
Location: Coronado, CA

Post by Mike Caviston » April 23rd, 2006, 5:59 pm

I am interested in the common perception among athletes that taking an early lead in a race provides a tactical or psychological advantage. This is certainly true among OTW rowers. As Garland notes (see previous post) and as I have also found, in international 2K racing the first 500m is almost always completed at a significantly faster pace than the remaining segments, including the last 500m. I don’t see that as a good thing, but I’ll come back to that topic in the future. The first racing data I finished analyzing was for Olympic running events. I began by looking at whether taking an early lead for 800m, 1500m, 3K or 5K races was a successful strategy. I can at least see some potential tactical advantage to getting ahead of the field early: avoid getting boxed in and force others to run a longer course if they want to pass you. But what actually happens? From 1984 to 2004, in 36 Olympic finals (men’s’ & women’s 800m, men’s & women’s 1500m, men’s 5K, and women’s 5K [’96-‘04] and 3k [’84-‘92]), the early leader went on to win exactly four times (11%). “Early leader” means being in first place @ 400m (for 800m or 1500m races) or @ 1K (for 3K & 5K races). So in track running, the strategy of getting ahead early hasn’t been particularly successful. Anecdotally, I noted the change in strategy and results by Great Britain’s Kelly Holmes in the women’s 800m & 1500m from ’96 to ’04. In ’96 & ‘00, Ms.Holmes favored faster starts, but finished 4th in the 800m in ’96 and 3rd in 00’; in the 1500m she finished 9th in ’96 and 7th in ’00. (In ’96, she lead through 1200m but faded in the final 300m.) In ’04, Holmes was content to stay in back of the pack for the majority of her races before kicking hard to take gold in both events. Meanwhile, Jearl Miles Clark (USA) and Natalya Yevdokimova (Russia) were taking early leads but fading out of the medals.

I could only find complete split data for all 1500m competitors for the ’84 and ’88 Olympics, but this data included results for qualifying and semifinal heats as well as finals. I combined the qualifying heats and semifinal heats, separating the automatic qualifiers (by place) from the non-qualifiers. The finals were treated individually. The results can be seen below.


For each category, you can see the raw time as well as speed in m/s for each segment. You can see the velocity of each segment as a percentage of final velocity. For example, in the ’84 women’s final, the top six finishers completed the opening 400m @ 98.1% of their final average speed, while the bottom six finishers did the first 400m @ 99.7% of final speed. In other words, the “winners” started more conservatively than the “losers”. Note the correlation coefficients between initial pace and final time: slower starters tend to finish faster. Below each comparison for mean values is the P-value for that comparison. For those unfamiliar with statistics, the P-value expresses the probability of making a Type I error (which in this case would be concluding the samples are different when they are not. Different values could occur just by chance.) So the smaller the P-values, the more statistically significant the findings. In most published research, values <.05 are considered significant; in some cases, stricter values of <.01 may be required or less strict values of <.1 may be satisfactory. I’ve included all values and highlighted the ones <.1 in bold. (Note: all calculations and all charts & graphs were performed with Excel.)

Here are some graphs of the men’s & women’s 1500m finals from 1984:


The first two compare average speed for “winners” and “losers” during each segment, while the final two show relative pacing for each group. Note how the “winners” tend to start slow but keep accelerating while the “losers” start fast but can’t maintain the same acceleration.

Mike Caviston

Mike Caviston
1k Poster
Posts: 167
Joined: April 20th, 2006, 10:37 pm
Location: Coronado, CA

Post by Mike Caviston » April 23rd, 2006, 6:05 pm

Below are similar analyses for the available data for individual pursuit cycling. In the qualifying round, competitors are paired up for simultaneous time trials, with the fastest eight finishers moving on to the semifinal round.


1k Poster
Posts: 122
Joined: March 16th, 2006, 8:48 pm
Location: New Jersey

Good to Read Mike's Thoughts

Post by tennstrike » April 24th, 2006, 9:54 am

Let me add a welcome back, Mike. Thanks for sharing.
6'1" 192lb 60
500 1:38.7 | 1K 3:29.2 | 2K 7:16.9 | 5K 19:14.0 | 6K 23:12.3 | 10K 39:40.5 | Started rowing June05

Mike Caviston
1k Poster
Posts: 167
Joined: April 20th, 2006, 10:37 pm
Location: Coronado, CA

Post by Mike Caviston » April 24th, 2006, 3:03 pm

My apologies for distorting the aspect ratio of the page with these big tables. I’ve tried reducing the jpegs in my Photobucket album, but it won’t appear any smaller here.

Here is some data for men’s & women’s swimming, using Olympic & WC 400m & 800m finals. The first three analyses combine the ’00 & ’04 Olympics with the ’05 World Championships. The men’s 800m is a non-Olympic event so only the ’05 WC was used. Interestingly, even though there are only four subjects each in the “winners” and “losers” brackets, the results are still quite strong statistically. The women’s results, for whatever reason, are not as conclusive as the men’s, but still tend to indicate that more conservative pacing is ultimately more effective.


The speed skating data (from the 2002 Winter Olympics) included splits for the first 200m and each subsequent 400m. I’ve limited the table to only the first 200m and final times for simplicity. The major difference between “winners” and “losers” was that “losers” were faster in the opening 200m and slower in the penultimate 400m.


Jim Barry
Posts: 49
Joined: March 28th, 2006, 3:11 pm

Post by Jim Barry » April 24th, 2006, 5:38 pm

Thanks for sharing those tables and the work Mike.

Just a thought:

Seems the talented racers in the Olympics have so much control over what pace is run where the less talented are at their mercy. The winners sort of force a bad pace on the losers. Just keeping up with any eventual winner will crush any lesser talented athlete but since they can keep pace for a little while it does seem incredibly logical (emotionally perhaps) to stay with them just in case they never do pick up the pace or faulter. I recall racing on the water with more talented crews and we always tried to keep up with them and we always got blown away. We were so hurting at the end, our pacing would have been called bad and I guess we lost by more than we needed to, but since it seemed such a foregone conclusion that we were going to lose we must have figured we were not going down without a fight. It is so human nature to try and win or medal. It's great to lay out the reality of what all that costs really.

Mike Caviston
1k Poster
Posts: 167
Joined: April 20th, 2006, 10:37 pm
Location: Coronado, CA

Post by Mike Caviston » April 24th, 2006, 6:06 pm

The psychology of racing is very curious and fascinating. In my youth I was definitely one of the ones who would go out hard (running, biking, rowing, whatever) and try to hang on as long as I could, I guess maybe thinking I could simply will myself to be better than I was. Or maybe the favorite would be struck by lightning and I’d cruise on to victory. Unfortunately, it never worked that way. I wonder what those second-tier elite athletes feel years later. Are they satisfied they tried to hang in there with the best, even though they got creamed? Or do they wonder how fast they could have been, even if they couldn’t win? These days, when I’m preparing myself or others for competition, I want to focus only on my own or my athletes’ performance – just do the best I can and let the winning and losing take care of itself. When I come to the indoor rowing data, it will be even clearer that the more unrealistic one is at the start, the further from their potential they will finish. A great many people are destined to finish in the middle of the pack, and just want to get a personal best. I’m hoping to convince them that next to actual training, pacing is the most important consideration. And bad pacing can neutralize the best training.

Mike Caviston

Post Reply