Pacing question

[joeevans8]

Starting erging in earnest this fall and have entered a 30 minute row at Mid Atlantic Sprints in Virginia first weekend in Feb.
PB's at this time (all recent) are
2000m- 7:30
5000m 19:35
10000m 40:45
15000m 1:03:40
1/2 marthon 1:30:45
Haven't rowed in competition-what should I aim for for a 30 minute competitive piece
47 years old, 173 pounds
thanks in advance

[Mike Caviston]

A little over 200W, or about 1:59. Good luck.

[arakawa]

[quote="Mike Caviston"]A little over 200W, or about 1:59. Good luck.
Where do you get or how do you generate such a profile?

[Mike Caviston]

Make a table in Excel, create a JPEG, upload to Photobucket.com, paste a link here. Almost that simple, though it took some practice & false starts to figure out each step.

[toby]

I calculated the pace a little differently than Mike (I think). In Excel I plotted split time as a function of total time spent erging each piece (all expressed in seconds). For example, the 2K split time is 1:52.5 = 112.5s, while it took 7:30 = 450s to row it. The whole data table:

distance(m) total time(sec) split(sec)
2000 450 112.5
5000 1175 117.5
10000 2445 122.3
15000 3820 127.3
21097 5445 129.0

Split time is expected to scale as a logarithmic function of the length of the piece. So I made a scatter plot of the data, had Excel fit a logarithmic function, and used that function to predict the split time for a 30min (1800sec) row. The best fit function is:

split time = 6.826*ln(total time) + 70.1

Plugging in 30min=1800sec for total time, the calculated split for the 30min piece = 121.2sec = 2:01.2.

Of course, all this assumes that the pieces in the dataset are "all-out" efforts (or at least equal efforts, comparable to the effort planned for the 30min). It also assumes that your fitness is not increasing rapidly. The latter assumption may be particularly suspect, since you have been training for just a few months.

You can also estimate using "Paul's Law," which states that the split time increases 5sec with every doubling of distance. Your 2K of 7:30 makes this calculation convenient for a 30min piece, since 30min is exactly 4 times your 2K. Paul's Law predicts that you'd have to add 5sec to your 1:52.5 for a 15min row, and another 5sec for a 30min row, for a projected 30min split time of 2:02.5. The nice thing about Paul's Law is that you can use your most recent, most all-out effort (of any distance/time) to estimate your current split for any other distance/time.

You could even just split the difference between your 5K and 10K split times, since the average of the total times is right at 30min. That would give a split time very close to 2:00.

Now you have a range of estimates from 1:59-2:02.5 to choose from.

With all that said, I use Mike Caviston's Wolverine Plan and have never known Mike to give anything but thoroughly sound advice, so take the above with a grain of salt.

[joeevans8]

Thank ou all.
I do know my fitness level has improved quickly. I was figuring on 1:58 or 1:59 for this event, but knowing how I used to run road races ("Oh I feel great, I'll go out fast) I don't want to spend the last 5 minutes dying a withering death because I rowed the first ten minutes at 1:56 or thereabouts.

[tomhz]

You can also estimate using "Paul's Law," which states that the split time increases 5sec with every doubling of distance. Your 2K of 7:30 makes this calculation convenient for a 30min piece, since 30min is exactly 4 times your 2K. Paul's Law predicts that you'd have to add 5sec to your 1:52.5 for a 15min row, and another 5sec for a 30min row, for a projected 30min split time of 2:02.5.

Paul's Law is correct in the way that doubling the distance adds a number of seconds to your split. But the amount of seconds is a personal matter depending on your physique, strength, age, fitness. For 90% of ergers it is between 3 and 5 seconds, I think. For me it is 3.5sec.

For joeevans8 it seems to rise for longer distances. This will change as you will learn to pace the longer distances.
2:00 seems a fair starting point for a 30min race.

Tom

[arakawa]

Where do you get or how do you generate such a profile?

What I meant to ask was, how did you arrive at those paces?

Going from 2K to 5K added 5 seconds to the pace. Doubling the distance from 5K to 10K added 4.8 seconds. Doubling (roughly) the distance from 10K to HM added 6.7 seconds. My conclusion is that the pace (i.e. time per 500 m) is not a logarithmic function of distance.

Seconding what others have said, I have never known you to give anything but thoroughly sound advice, so I'm curious as to how you arrived at the target paces given one's 2K pace.

[PaulS]

Your PB's having one that is just a bit under 20 minutes and another that is a bit over 40 minutes make an estimate of 7500m for 30 minutes (2:00) a very reasonable starting pace, then proceed to pick it up after the first half is out of the way. Choosing at easy "halfway point" to either "move" or "hold".

For example:
15:00 @ 2:00, then go to pulling 1:59's if you are feeling good.
7:30 @ 1:59, then go to 1:58 if feeling good.
3:45 @ 1:58, then go to 1:57 if feeling good.
and so on, the final minute you can empty whatever is left.

At any point you do not feel accelerating is an option, "hold" the current pace target, except for the last minute when you have 30 or less strokes to go, may as well burn the anaerobic stores at that point.

[toby]

Where do you get or how do you generate such a profile?

Going from 2K to 5K added 5 seconds to the pace. Doubling the distance from 5K to 10K added 4.8 seconds. Doubling (roughly) the distance from 10K to HM added 6.7 seconds. My conclusion is that the pace (i.e. time per 500 m) is not a logarithmic function of distance.

The logarithmic function relating split time as a function of distance in the dataset provided is:

split time (sec) = 7.228*ln(distance in m) + 56.8

This logarithmic equation explains ~98% of the variance in the data, compared to a linear equation that explains ~94% of the variance. Objectively, the data are (a bit) better explained by a logarithmic curve. Physiologically, I don't know why much longer erg pieces are possible with very modest reductions in power, which is what produces the logarithmic relationship between pace and distance.

Paul, I would be curious to know what factors you think affect the "constant" in Paul's Law. Why are some people able to double distance with only a 3sec/500m increase in split time, while others need to increase splits by 5sec/500m? It would be intuitive that someone who is very fast, and thus for whom each sec/500m equates to a big power difference, would be closer to 5sec/500m than to 3sec/500m. However, I suspect that it in reality it might be the faster folks who are better able to maintain pace across a wide range of distance. I haven't come across a detailed discussion of this, but maybe I just missed it.

[Mike Caviston]

What I meant to ask was, how did you arrive at those paces? Going from 2K to 5K added 5 seconds to the pace. Doubling the distance from 5K to 10K added 4.8 seconds. Doubling (roughly) the distance from 10K to HM added 6.7 seconds. My conclusion is that the pace (i.e. time per 500 m) is not a logarithmic function of distance. Seconding what others have said, I have never known you to give anything but thoroughly sound advice, so I'm curious as to how you arrived at the target paces given one's 2K pace.

Those aren’t target paces, and I didn’t arrive at them. I simply calculated the per-500m paces and Watts for the times Joe listed for each distance, and plotted Watts against distance for a quick visual approximation of how power output decreases as distance increases. Is the overall relationship between distance and pace linear or logarithmic? Who knows? There’s no data to analyze (we need lots of subjects doing tests at various distances under standardized, controlled conditions). For the few times Joe lists, the relationship is fairly linear. Since 30’ falls roughly half way between Joe’s times for 5K and 10K, I noted the midpoint (in Watts) between his 5K & 10K values, and suggested the corresponding pace. Of course, anyone who’s read my posts about race strategies and effective pacing knows I would suggest starting a little slower than the target pace and gradually picking up the pace throughout the race (i.e., negative split).

Mike Caviston

[toby]

Is the overall relationship between distance and pace linear or logarithmic? Who knows?Mike Caviston

Mike, don't you think it's pretty clear that the relationship between distance rowed and pace (expressed in sec/m) isn't linear? If it were, then the difference in pace between a 500m and a 2K piece would be the same as the difference between a 50,000m and a 51,500m piece.

Pace (sec/m) is a logarithmic function of power (watts), which could explain why the power vs. distance graph you posted is fairly linear.

If someone wanted to know generally how power scales with distance the C2 rankings at various distances could be used. Since this whole thing revolves around a second or two @ 2:00 I don't think I will investigate further. It is intriguing to think of a model where the total work (power*time) is constant, though. I reckon someone smarter than I am has figured this out already.

I apologize if this greasy spot that used to be a dead horse is wasting too much of everyone's time.

[arakawa]

Those aren’t target paces, and I didn’t arrive at them. I simply calculated the per-500m paces and Watts for the times Joe listed for each distance, and plotted Watts against distance for a quick visual approximation of how power output decreases as distance increases.

Thanks for suffering this fool - I didn't bother to compare the numbers.

[PaulS]

Paul, I would be curious to know what factors you think affect the "constant" in Paul's Law. Why are some people able to double distance with only a 3sec/500m increase in split time, while others need to increase splits by 5sec/500m? It would be intuitive that someone who is very fast, and thus for whom each sec/500m equates to a big power difference, would be closer to 5sec/500m than to 3sec/500m. However, I suspect that it in reality it might be the faster folks who are better able to maintain pace across a wide range of distance. I haven't come across a detailed discussion of this, but maybe I just missed it.

The "constant" is the Erg and PM, measuring the Athlete and their balance of fitness in terms of Strength and Endurance (The variable.). (The exponential power requirement of the Ergs Pace calculation takes care of the rest.)

Paul's Law is not a performance predictor (though it can be loosely used that way), but a way of determining in which direction training focus should go. If I had an athlete that fit the model then they keep training as they are. Otherwise it will show that there needs to be a Strength or Endurance focus.

If an athlete doubles the distance and goes only 3 seconds slower, they need to improve their strength, especially if they want to get faster at the short distances. Sorry if that sounds completely obvious, but since the question was being asked it must not be.

Happy to clarify further if needed.