spreadsheet/number crunching

[bloomp]

I'm a collegiate rower with a good background in math/science and I remember a guy posting on here or the UK forums about a good way to look at your erg splits over different distances and using a logarithmic trendline in excel to determine which scores were above or below your "ideal" performance.

I've been keeping pretty up to date numbers, with a variety of pieces and have found one interesting outlier. My 2k split is a 1:50.5, which falls well below the trendline (much faster) than my other pieces. Any opinions on why that's happening? And I highly recommend doing this with your times, it's a very good predictor of what split you should hold for X distance (just turn on the option in excel to show the formula and plug it in)!

Here's an image of it:


For reference, the first piece is a 250, the second is a 500, third is a 1k, fourth is a 2k, fifth is a 15 minute piece (3800m), 5k, 6k, 8k and then hour of power. If you think it makes a difference, I'm 5'9 and 146 pounds.[/img]

[RowtheRockies]

Could it be that 2K is the most common race distance (at least on erg, not sure about on the water) and your training has centered around this distance? Interval training done mostly in the 1' to 1K range?

Rich

[El Caballo]

I agree with RowtheRockies, the primary race distance for college is the 2K--your training is most likely based around that distance, and as a result, that distance should be one of your best. It is also possible the day you did your 2K PB was under near perfect conditions. The weather, your strength, and anything else that could have impacted the row.

Could it be that your other pieces were less than optimum? Did you keep a fairly even pace during all of your pieces? A fairly even pace with slightly negative splits (slightly faster at the end) seems to be the optimum (from my experience and from what Mike Caviston reports elsewhere in this forum). How consistent were your paces? I think your other pieces have lots of room for improvement. Over what length of time did you do these and has your conditioning improved during this period?

A suggestion: set your distance scale on the plot to logarithmic. Your plot would then be a straight line and your data points would be spread out more evenly (I think it makes it easier to visualize what distances need improvement).

Or, for Online World Ranking distances, try my spreadsheet at wakeley.us/rowing/pace_prediction6.xls (I would use the www and make this a link but as a "new member" of this forum I currently am not allowed to post URLs--maybe the full URL shows up in my signature). I have several formulas there (including a linear fit of pace vs. ln(distance)). I am in the process of updating it so other distances can also be entered. I have protected most cells so you do not accidentally overwrite important formulas, but if you want to unprotect anything to better understand the formulas and the math involved, the password is "c2erg".

[PaulG]

You might be referring to a thread I started in this forum. If you do a search on this forum for "pace curves" you will find it. I have a feeling I reinvented a well-discovered wheel, but it is still a good way to predict paces and monitor progress.
With regard to your curve, I notice that at the lower distances your paces are right on the curve with little error. As suggested, if you take the ln of the x-axis you will get a straight line and deviations from the line may be more apparent. Also limit the y axis to your range of paces and you will see more detail. Most people are significantly below the line at distances of 1000 m and less. The fact that you are right on it may indicate that you need some more strength and anaerobic work. Also pace curves evolve so if your training is improving, your most recent entries should be below the line. And as noted, entries for the distance you are training for will be lower also.

[bloomp]

Could it be that your other pieces were less than optimum? Did you keep a fairly even pace during all of your pieces? A fairly even pace with slightly negative splits (slightly faster at the end) seems to be the optimum (from my experience and from what Mike Caviston reports elsewhere in this forum). How consistent were your paces? I think your other pieces have lots of room for improvement. Over what length of time did you do these and has your conditioning improved during this period?

In response to that and what PaulG said, I pulled my 2k last November at the end of the head racing season. I'd normally just pulled 2k's before my team would get on the water in the spring (high school), or mid-spring season, so it probably has something to do with the distance training for head racing.

As for pacing, I aim for starting a distance piece (<15 minutes) at about 1-2 seconds above my goal split average and work down so that I pull the average I want or slightly below for about 75% of the piece. For example, my 6k that is listed there started at a 2:02, and worked the split down and the cadence up for the last 4000m. I think one of the possibilities is that many of my pieces are part of multiple part pieces (3x18', 2x5k, 2x6k, 8x500m). I guess I'll go back to try to PR on each piece then enter it.

The oldest one of those pieces is from October of last year - the hour of power. The most recent is a 1k from Monday. I do think that if I go and try to PR on a piece rather than work it into a set of workouts, the curve would accept the 2k data better. As for anaerobic and strength work, PaulG, do you suggest lifting (squats/clean and press/lunges) or just shorter, more intense erg pieces?

Caballo, thanks for the spreadsheet - as much as I work at times, being able to analyze them makes things easier to work towards.

[tomhz]

Bloomp,
the 6.0743 in your formula indicated your relative strength for shorter distance. Multiply it by ln(2)=0.693 to get X in "Double the distance - add X"
If it is high, your longer distance are weak, if it is low your shorter distances are weak.
In your case X=4.2. In my case (50yo light heavyweight) it is 3.6. On average one sees lower values for LW's than for HW's.
But as you mentioned your results are not consistent yet. Only 1 sec pace slower when going from 8K to 14.5K does not fit. This graph helps a lot to show you where you can go faster (the 250m and 500m and 1K and 8K for sure). Then make a new graph.

[PaulG]

Caballo: Excellent spreadsheet. More sophisticated than anything I came up with.

BloomP: I wouldn't presume to give training advice. Since you are in an organized rowing program, check with your coach.

TomHz: I knew there must be a way to relate the slope of pace curves to Pauls Law, but I did not know how. The slope (or coefficient ) of my pace curve is 9.04 yielding an x=6.26. This is compared to the WR holders in my age group (50-59) and weight class (HWT) which yields an x = 5.16. That is much closer to the standard of 5 in Pauls Law. I guess I need to work on distance. As a very light HWT (167 lbs), I am within 82-84% of the HWT world records from 500-2000 m (that means when they are finished I am 82-84% finished). I'm somewhat pleased I can generate that much speed at the shorter distances but I need much more endurance work.

[bloomp]

Ok so before practice today (we did a 10k), I did some shorter sprints - 250m in 50 seconds (1:40 split), 100m in 18.6 seconds (1:33 split), and I'm relatively sure I can replicate that sort of negative trend in the 500m and 1000m - I did both pieces twice at about the same pace, a few minutes rest in between. As for distance, I'm still working on it. Today's 10k was at a 2:06, but that was after those sprints pieces about 45 minutes earlier.

What is the opinion on weight normalization? As a lightweight I really like the idea, and I know that on the water I can pull a lot harder than on the erg, but I'm still happy with my almost sub 7 2k at my age/height/weight.

[Nosmo]

What is the opinion on weight normalization?

The standard weight normalization is that weight to the 2/9th power.

So two equally fit and equally good rowers on 165 lbs person and the other 200, you would expect the 200 lbs person to be

(200/165)^(2/9) =1.04 or 4% faster. (eg, 7:00 vs 6:42).

This is a rough formula that works for aerobic efforts so you would not expect it to work for 500m. (In theory the exponent is 1/3 rather then 2/9 for anaerobic races)

Now if you want to compare on the water times to erg times you have to add the weight of the boats to the body weight.

Final caveat is that this is theory and it makes the assumption that all other things are equal which they never are. Not sure who first said it but "the difference between theory and practice is much bigger in practice then in theory."