My freestyle technique, what needs work?

Any labchart freestyle files available to zip up and attach here?:)

I find that little bump interesting, is it typical of other elite swimmers? If it is the catch then it is interesting that she decelerates afterward. I wonder if she is sculling there or just not able to apply enough power while in that position, or if that is a spike in drag, or if it is kick related, or... Lindsay - We have seen in swimmers elite and non elite that momentary "bump" in velocity you referred to in your post. Because the telemetry is collected at 1,000 data points a second, it reveals very subtle variations in the stroke cycle that might not be evident with just regular vision. As you pointed out even with that momentary bump in velocity in general terms, this swimmer is progressively decelerating through this phase of the stroke cycle. Your questions about what might be causing this effect is a good one, and let me put together some general ideas from our experience and I will post them. Bud I would really like to fool around with that software. Geochuck - Here is the link to the "reader" program for the software. One of the greatest features of this technology is the reader program allows you to use all of its features. Having the ability to use the files long after a testing session is completed we feel is really important to the improvement process, and have been told it really it adds additional value to the experience. The program is designed so that the analysis can be done right on your computer screen by simply scanning over areas of the telemetry using your mouse. When you do that, the video scene automatically updates to that position, so you can visually see what your doing at that moment in time. It makes the software very easy and fun to use. For anyone interested, I would be happy to forward some sample files. I should warn you that I have been told after a testing session, spouses of Masters swimmers have had difficulty prying them away from the computer screen for weeks!! www.adinstruments.com/.../

Thanks Bud for the warning and the link. Even though my wife happens to be in the same room with me, I have not had a conversation with her or seen her for over a week. After I get into this it may be 2 weeks.

Budd I have noticed that many swimmers, when their hands exit water, they are shooting water forward off their arms. I do not like to see this. Also max pressure should be applied from the start of the catch phase to finish.

I find that little bump interesting, is it typical of other elite swimmers? If it is the catch then it is interesting that she decelerates afterward. I wonder if she is sculling there or just not able to apply enough power while in that position, or if that is a spike in drag, or if it is kick related, or... Lindsay – Here are some of our findings after testing many swimmers. Let’s use the telemetry I posted earlier on the elite swimmer, and let me say again, these are general findings. As the stroke cycle begins, and the hand/arm enters the water and starts the pulling pattern, velocity tends to progressively decline to the minimal point in the stroke cycle somewhere under the shoulder. (As the tracing I posted demonstrated) Sometimes we do see a small velocity increase “bump” about halfway through this pulling phase, and then another decline. But generally, as the arm moves somewhere under the shoulder, this is the point where minimal velocity is attained. We believe this is caused by the drag of the arm as it moves underneath the body. As the stroke cycle begins, the hand/arm progressively moves down, and under the body. This movement under the body, progressively creates increasing drag, and as the telemetry shows, the velocity decreases to the minimal velocity position somewhere under the shoulder. It happens to almost everyone. Using the telemetry I posted of the elite sprinter, notice the difference between the maximum velocity and minimal velocity points for each stroke cycle. Better swimmers have a smaller max/min variation in velocity than swimmers that are slower, and that might be obvious. However, many slower swimmers we have tested can generate similar peak velocity values, but their max/minimum velocity difference is much greater. In addition, that difference between swimmers can be very small when looking at one stroke cycle. But during a race where swimmers are using stroke rates between 50 and 60 stroke cycles per minute, that small difference becomes cumulative, and can define from a swimming perspective differences in performance. The “bump” in velocity that you referred to that occurs about halfway through the initial pulling phase, is really up for debate, and I think is hard to precisely define. We suspect that it is not caused by the arm, but possibly the body of the swimmer moves momentarily into a position of decreased drag in the water, and thus the velocity makes a momentary increase. The other reason why we feel drag is the major player is if you look at the image of the telemetry when maximum velocity is attained, there is much less arm surface area under the body. One arm is outstretched just getting ready to start the next pull, and the other arm is partially out of the water, with only a small portion of the hand/arm still in the water, So the stroke correction would be to try to get the arm though the pulling phase while minimizing the amount of max/min velocity difference, being especially focused where the minimum velocity phase occurs. (Somewhere under the shoulder) Obviously we believe any swimmer can achieve optimal results by incorporating Velocity/Video Telemetry technology into their daily training plan. Budd

Budd I have noticed that many swimmers, when their hands exit water, they are shooting water forward off their arms. I do not like to see this. George - I agree. A clean drag free exit from the back of stroke is very important. Also max pressure should be applied from the start of the catch phase to finish. I agree - Pulling hard is important, as well as pulling in a manner that does not create a big variance between the peak and the minimum velocity during a stroke cycle.

More really interesting stuff Budd! The testing with the pull buoy suggests an interesting experiment, if you have a swimmer push off in a streamline, and then push off and start kicking immediately you should be able to get a handle on how much propulsion the kick provides at different velocities. How much propulsion the kick provides has been a hot topic for many years and I think you have the tools to answer it. At least for a given swimmer's kick. Have you done any testing along these lines? Assuming that the sort of descending sine wave in the sprinter graph is the result of dolphin kicking it would be interesting to see what the velocity curve looks like with just streamlining, and with flutter kicking, given the magnitude of the accelerations and decelerations one wonders whether with a flutter kick the two legs wouldn't largely cancel out. It is eye opening to see the magnitude of the drag during the leg recovery! Depending whether each dolphin kick spans one or two of those waves would tell us whether the upward portion of the kick is propulsive or not. Would I be correct that your values for acceleration, power, and force are derived from the velocity data? If so, the force would be net force (propulsive - resistant drag) on the swimmer rather than propulsive force? I also assume that the Aquanex measures pressure rather than force, i.e. it doesn't tell you what direction the force is in, and the pressure has to be multiplied by the surface area to determine the total force? If so, someone with a windmill style pull might be exerting force near the top of the stroke but much of that force would be downward and not be contributing to forward propulsion. Also, when dealing with pressure the total force is proportional to the surface area (actually projected backward facing surface area would be the effective area) so someone with a bigger arm (or better oriented arm) could exert a larger propulsive force than a smaller arm (or less well oriented arm) even though the pressure they are generating is smaller. Almost by definition the faster swimmer is exerting more net productive force than the slower swimmer, i.e. the average force is greater even if the peak force is lower, and again the pressure at the hand is only half the equation, you have to multiply it by the effective surface area. It may well be that the age group swimmer is experiencing more drag, but controlling a few more variables would more conclusively prove your hypothesis. It seems to me that the velocity meter inherently measures net force so you need something more to tease apart propulsive and resisting drag forces. The pressure sensor helps in this regard but requires effective surface area to determine propulsive force. Video data taken from the front could help determine area, although an outline taken from the front angle and combined with out of the water measurements could do the trick at the expense of a lot of calculations, although I guess computers can take care of those. Have you tried attaching the pressure sensor to the upper arm to see how much drag pressure there is there? That would give some very informative data with regard to your drag hypothesis.

Preferably the swimmer would use a pull buoy and not be kicking. Have you already tried something like this? Lindsay - We have tested with a pull buoy ( arms only) to subtract out the possibility of what the kick might add. Without the legs, there is a difference, but we have found generally nothing significant. The misleading element when testing with a pull buoy is that is when you float the lower body, for some people this reduces drag. Some people we have tested actually swim faster with a pull buoy in, so the telemetry in that configuration can be misleading. It would be very interesting to see how much of the drop in velocity can be accounted for just by lack of propulsion, if you were to have a swimmer swim a few strokes and then just stop with one hand out front and one hand at the side we could subtract out the decline in speed there from the decline in speed when pulling and get an idea of how much is left to be accounted for. Your question about propulsion is a good one, and we have recently conducted some very interesting measurements in that regard. Our Velocity Meter/Video Telemetry can also measure acceleration, power (watts) and force (lbs and kg.) as well as velocity all at the same time. In regard to propulsive force, we recently tested two high level swimmers one of college age and one of age group age for freestyle, with their best event being the 500 freestyle. Both subjects were simultaneously measured using our Velocity Meter and a device developed by Dr. Rod Havriluk, called the “Aquanex” that places pressure sensors on the hands between the center two fingers. The device measures propulsive force, and also collects underwater video at the same time. Since Dr. Havriluk has a validated device that measures propulsive force through a stroke cycle, one the goals was to determine using two independent means of measurement, how the propulsive force data would compare between the two devices. We had the swimmers do progressive trials from slow speeds to an all out sprint with both devices collecting force and video telemetry at the same time for both subjects for all trials. To our surprise, both devices collected very similar average peak force data within a couple of lbs. on both subjects for all of the trials. In general terms we found with these two subjects as the stroke cycle begins, and the hand/arm enters the water and starts the pulling pattern, propulsive force progressively increased and peaked in the stroke cycle at a location basically under the shoulder. This location was verified by both devices using the telemetry data and the synchronized video record. The findings below for force are from the Aquanex, because it specializes in measuring propulsive force, and the corresponding velocity measurements are from the Velocity Meter, since measuring velocity is its specialty. As I said above, both devices were collecting at the same time for each trial. For me, there were two really interesting findings during this testing. First, on the fastest trial average peak force measured by the Aquanex for the college swimmer at 1.90 meters/sec. averaged 22 to 23 lbs. of peak propulsive force, while the age group swimmer at 1.75 meters/sec averaged 27 to 28 lbs of peak propulsive force. From a performance standpoint, the college swimmer had much faster times, (15 sec. difference) in 500 freestyle. Secondly, readings between 22 and 27 lbs. of propulsive force do not seem high, but might be for swimming. A search of the literature on this subject revealed a study about Alexander Popov by a Russian scientist that claimed during the race at the Olympics when Popov broke the world record for the 50 meter freestyle, he used 24 watts of power. (17.70 lbs. of propulsive force converted) Even though both of these swimmers have pretty good times for their ages, the college swimmer has faster performance times and the average velocity of the fastest trial was also significantly higher. With that being said, I would have expected the college swimmer to produce significantly higher force values than the age group swimmer, but that was not the case. Even though the age group swimmer produced higher force values, he also had significantly slower performance times and average velocity on the fastest trial compared to the college swimmer. So to me, the age group swimmer has to have greater drag during the stroke cycle. An interesting outcome to say the least. Budd

Wow...some very introspective thought from the last post. This is why I enjoy posting some of our findings on a public forum like this because all of our work for the last 20 years has gone through extensive peer review by some of the best in the business, and I enjoy the feedback. I actually have forwarded your post to one of my aerospace engineer alumni to evaluate your comments. I believe all the comments you mentioned can be measured with our technology. Just a matter of getting to the right equations. Let me try to answer some of your observations that we have tried to examine. if you have a swimmer push off in a streamline, and then push off and start kicking immediately you should be able to get a handle on how much propulsion the kick provides at different velocities. How much propulsion the kick provides has been a hot topic for many years and I think you have the tools to answer it. At least for a given swimmer's kick. Have you done any testing along these lines? We have conducted extensive testing doing exactly what you described above, and in general terms here is what we found. Starting with push and glide from the wall, you get an exponential velocity deceleration curve like the one you posted earlier of my swimmer. When you push and flutter kick immediately it also produces an exponential deceleration curve with small “bumps” of acceleration and deceleration variance. The variations in the telemetry are caused by the opening and closing of the feet/legs. However, when you fit a mean line though the push and kick telemetry, and plot it against the push and glide line, they plot right on top of each other. (No difference) Again, I think you assume kicking from the wall provides propulsion that would create a deceleration curve that would be greater than the push and glide curve, but the act of separating the feet/legs creates drag, and the net effect is the same deceleration curve. Generally the curves separate somewhere around the 1.5 meters/sec point, where the push and glide curve continues down exponentially, and the push with kick velocity starts to level off. Again referring to thousands of push off and glide verses push off and kick trials, the push with flutter kick rarely ever produces a deceleration curve line that is better than the push and glide line, until you get to really slow velocities, not associated with regular swimming speeds. In addition, for most of the swimmers we have tested using the push and dolphin kick immediately from the wall typically produces a line that is slower than push and glide or the push and flutter kick to somewhere around 1.5 meters/sec. because most swimmers are really not as good at the dolphin technique as they think, and stay underwater to long and breakout well below their regular swimming velocity. But for that particular application, our technology can specifically define that intersection precisely without guesswork. Would I be correct that your values for acceleration, power, and force are derived from the velocity data? If so, the force would be net force (propulsive - resistant drag) on the swimmer rather than propulsive force?They are derived from the velocity data, and at this time, I would agree with you, because our values from the velocity meter were slightly higher than the hand sensors. That is why on the post, I used the values from the hand sensors and not the Velocity Meter. I also assume that the Aquanex measures pressure rather than force, i.e. it doesn't tell you what direction the force is in, and the pressure has to be multiplied by the surface area to determine the total force? If so, someone with a windmill style pull might be exerting force near the top of the stroke but much of that force would be downward and not be contributing to forward propulsion.I believe from the synchronized video, we can make some assumptions about the direction of the force/pressure through any part of the stroke cycle, and we have developed a very effective method of demonstrating that effect. The force telemetry I have seen with freestyle up to this point shows max force or pressure as you call it, occurred in general terms somewhere under the shoulder. To date, I have not seen any telemetry where the max force/pressure value for a stroke cycle occurring anywhere during the initial phase of a stroke cycle for freestyle. And if our comparable numbers are anywhere close to correct 20 to 30 something pounds of propulsive force does not seem very high to me, but as I said earlier, may be those are high numbers for free swimming. Also, when dealing with pressure the total force is proportional to the surface area (actually projected backward facing surface area would be the effective area) so someone with a bigger arm (or better oriented arm) could exert a larger propulsive force than a smaller arm (or less well oriented arm) even though the pressure they are generating is smaller.This is where I think we differ. When you speak about how one person could produce more or less propulsive force, what I think you leave out is during the stroke cycle as propulsive force is progressively being generated to its peak during a stroke cycle, what is also concurrently happening with the instantaneous velocity of swimmer? I believe I have already given you evidence to the answer of that question. It seems to me that the velocity meter inherently measures net force so you need something more to tease apart propulsive and resisting drag forces. The pressure sensor helps in this regard but requires effective surface area to determine propulsive force. Video data taken from the front could help determine area, although an outline taken from the front angle and combined with out of the water measurements could do the trick at the expense of a lot of calculations, although I guess computers can take care of those.At this time, the Velocity Meter could be measuring net force, and more work needs to be done in this area. This was the first time where two devices collected telemetry at the same time, so it's never good to jump to any conclusions without careful review, but the initial results were encouraging. Obviously these kind of measurements are not as easy like with other sports because the physics of moving through the water are not analogous to movements or measurements on dry land. Have you tried attaching the pressure sensor to the upper arm to see how much drag pressure there is there? That would give some very informative data with regard to your drag hypothesis.Not yet, the first step was to see how close those aerospace engineering alumni of mine could come up with equations that would output values from the Velocity Meter close to the hand sensors telemetry.

Again referring to thousands of push off and glide verses push off and kick trials, the push with flutter kick rarely ever produces a deceleration curve line that is better than the push and glide line, until you get to really slow velocities, not associated with regular swimming speeds. Aha! As I and many others have long suspected, flutter kicking is not actually propulsive at regular swimming speeds! Toss those kick sets into the trash can! Well, probably not, but this would give serious weight to the argument that the kick is not used for propulsion but rather to maintain body position, or to set rhythm, or as a rotational anchor, or some combination. If its rhythm we should be training for maximum cadence rather than maximum power. If its as a rotational anchor we may be stuck with the need for a more powerful kick. In either case, you have apparently settled a debate that has been going on for decades! Now we can move on to debating what role the kick does play. This is where I think we differ. When you speak about how one person could produce more or less propulsive force, what I think you leave out is during the stroke cycle as propulsive force is progressively being generated to its peak during a stroke cycle, what is also concurrently happening with the instantaneous velocity of swimmer? I believe I have already given you evidence to the answer of that question. I think we are inferring force from velocity in different ways. Here's my line of reasoning, sorry if it is overly detailed and seems pedantic, that's just the way I reason when trying to be careful: force = mass * acceleration or acceleration = force / mass since mass is constant we can largely ignore it and just say that acceleration is directly proportional to force. If you apply a constant force on a mass it will produce constant acceleration which will give you a sloped straight line on a graph of the velocity, with the slope being proportional to the force. A positive force will produce a positive slope, i.e. increasing velocity, and a negative force will produce a negative slope, i.e. declining velocity. The minima and maxima of the velocity graph correspond to the points where the slope of the graph is zero, i.e. the force is zero. In our case the points where propulsive and drag forces are equal. A minimum in the velocity graph therefore represents a shift from drag outweighing propulsion to propulsion outweighing drag. A maximum represents a shift from propulsion outweighing drag to drag outweighing propulsion. We can observe that the minimum occurs when the hand is below the shoulder which is the point where propulsion equals drag and that from there to the maximum, where the hand is finishing the pull, there is a steady increase up to the maximum. The path is pretty close to a straight line in this segment, with just a little curvature at the ends, which tells us that the net force is close to constant in this phase. So, by my reasoning the steady upward rise from minimum to maximum in the graph represents a period of constant force rather than progressively rising force. It seems to me that you are saying that force is proportional to velocity and therefore force is rising during this period, where I am asserting that force is proportional to acceleration and therefore relatively constant during this period. Returning to my rambling, we reach the maximum as the hand finishes the pull and propulsive forces drop to whatever the kick provides, which is less than drag so, with drag predominating the slope of the line changes to downward. With the absolute amounts of drag and propulsion in the front quadrant not yet determined it is hazardous to draw too many conclusions but I think we can safely say that we're slowing down all the way through the front quadrant (except for the little bump) and we're speeding up all the way through the rear underwater quadrant. So you absolutely do not want to cut short the back quadrant in order to get your hand back in the front quadrant as quickly as possible, which was what was asserted in one of the other threads. If we were able to overlay the Aquanex force data we would be able to separate out propulsive force from drag force and see whether it is lack of propulsion, or a surplus of drag that leads to the deceleration in the front quadrant. My hypothesis is that there is not a lot of propulsion in most of the front quadrant.