Produced, edited, and copyrighted by
Professor Brent S. Rushall, San Diego State University
Volume 5, Number 2: February 2, 1999


Several questions have arisen in discussions of movement principles underlying body techniques in the crawl and backstrokes. Below are some points to consider.


Several swimming enthusiasts have promoted the idea that being as long and thin as possible in the water is the major technique principle for swimming faster. My published papers have described minimizing frontal resistance as being more important than reducing the wetted surface area but less important than creating waves (e.g., Rushall, Holt et al. 1994, Rushall, Sprigings et al. 1994). Some swimming theorists turn to some yacht/boat design principles and analogize them to swimming. Various forms of resistance are considered in boat designs. There is more to designing a sculling boat or a yacht's hull than being long and thin.

With sculling/rowing shells, a common estimate is that 82% of resistance is caused by the wetted surface and the remaining 18% by form resistance (the correct name instead of frontal resistance). Based on those figures the following can be deduced for rowing shells.

  1. Keeping the wetted surface area constant, reducing the form resistance (the cross-sectional profile) would improve performance but only marginally because of the large 82 to 18 ratio of surface to form resistance.
  2. Hull designs have one position in which they minimize drag resistance. If they rock or are off "balance" the wetted surface is increased and progress is impeded because of greater resistance.
  3. Keeping hull length constant, any increase in form resistance will impede progress.

For rowing shell designs there are tables on these factors (e.g., Hay, 1993).

There were two design alternatives contemplated for America's Cup yachts for the competitions in San Diego. One was flat and narrow which is good for rough water (e.g., America cubed, New Zealand's Black Magic) because the shape cuts through waves while only increasing resistance slightly. The other is faintly saucer-shaped and wide for very smooth water (e.g., Stars and Stripes). In rough water this shape rocks longitudinally ("hobby-horses"). In good conditions that suited its hull shape Stars and Stripes was unbeatable (the faster it went, the more it rose out of the water reducing both form and surface resistance). In slightly rough conditions Stars and Stripes was a "dog" mainly because it would hobbyhorse. In the America's Cup races the conditions more often than not favored the long narrow hulls. Dennis Connor gambled with Stars and Stripes' shape, and just beat all the US long narrow hulls. However, in finals he switched to a longer narrower boat of another contender, which he had never sailed, and lost terribly. New Zealand was clearly the better boat in the finals but that was not a fair test of the two design philosophies.

On numerous occasions it has been written and implied that a swimmer should remain long, on the side, and with an arm outstretched for a period in crawl/back stroke swimming. But a swimmer is neither like a racing shell (single, double scull, or whatever) nor a yacht. Here are the problems with such a concept in swimming.

  1. Virtually the total swimmer is underwater because of the position of natural floatation, and thus, the wetted surface is nearly maximized.
  2. No propulsive forces can be created in a long, side, stretched position and consequently, if that position is held, a swimmer will slow down due to the presence of only resistive forces.
  3. If maximally streamlined, a swimmer will slow down less but nevertheless, will slow down because of the state of non-propulsion. [While such a swimmer is slowing, other more efficient swimmers will be propelling themselves past the individual.]
  4. When a swimmer slows and then finally applies forces there are stages of slowing down (the correct term is negative acceleration not deceleration) and speeding up (the correct term is positive acceleration not acceleration). These changes violate Newton's First Law, the Law of Inertia which implies that more energy is needed to change inertia (to overcome frictional forces and inertia) than to maintain inertia (to overcome only frictional forces). Thus, such slowing-down/speeding-up swimming will be more exhausting and slower. Extreme negative and positive acceleration phases are not a characteristic of swimmers who must be efficient (e.g., 200 m plus) who usually demonstrate high rates, almost continuous propulsive force production, and minimum periods of time in long "sculling boat" positions.

Thus, the best streamlined position without any propulsive force is undesirable.

  1. It is not correct to assert that streamlining is more important than propulsion.
  2. A swimmer who can apply propulsive forces, with the least amount of non-propulsive time between applications, will approach more closely an inertial swimmer, which will be the best form of swimming for energy conservation and allocation.
  3. The hierarchy of preferences for streamlining, because of the severity of their resistance production is as follows:

There are many other principles of mechanics that have to be considered in concert with these features when determining what is good and bad for a swimmer. The principles that are correct for emphasis are likely to vary between swimmers.

When contemplating techniques, movement principles must be applied relative to each other. A technique item taken out of context without regard to its relationship to other features is usually a wrong procedure. The best streamlined shape possible but one which supports the most efficient production of propulsive forces is a more complete concept for contemplation positions in the water for speed swimming.

Streamlining is very important but should not be emphasized at the expense of continuous force application!


It has been proposed that a laterally flat body position is bad for swimming because places a swimmer in a position that acts like a barge. As an alternative, it has been proposed that swimmers should turn to close to 90 degrees to each side when swimming crawl and back strokes, that position supposedly reducing resistance.

Holding the body on the side produces no more or no less drag resistance than holding it prone in crawl or backstroke. The frontal cross-sectional area, the velocity of the swimmer, the wetted surface area of the swimmer, the density of the fluid, and the coefficient of drag will all be the same whether the body is on the front or side. Why? Because the buoyancy force remains constant no matter what the body's orientation. These are the factors that govern the various forms of drag resistance and are independent of orientation to gravitational pull. The body is no more streamlined on the side than on the front.

There actually could be a deficiency in a deep and narrow orientation. On the side the hydrostatic forces that cause a floating body to "rise" will be less. At some speeds, but I doubt if it would be significant in swimming, a deep and narrow body would sink as opposed to rise. This is because of the "slump" phenomenon and has to do with the lessening of water density around a moving floating object.

A "barge analogy" is incorrect for swimming. A barge's major deficiency is in the profile presented to the oncoming fluid that causes deflection and separation not the shape of the body following it. If a barge were rotated 90 degrees it still would present the same profile but would generate less beneficial hydrostatic forces.

There is a further problem with advocating holding a position on the side while swimming crawl/back stroke. The arm actions in crawl/back strokes are very complex but generally consist of three phases of sequenced movements, which are dependent upon the degree of body rotation. Those phases are:

  1. An elevation of the shoulder, medial rotation of the upper arm, and late flexion at the elbow as the body rotates to accommodate an "elbows-up" position and to re-position the forearm/hand drag surface to where forward propulsion can be created. The arm entry should conserve momentum from the out-of-water movement pattern. No propulsion occurs at this time.
  2. Adduction of the upper arm to maximize forward propulsion. This is to be completed in the time it takes the body to roll to, decelerate, and stop at the maximum rotation angle and then to accelerate and roll in the opposite direction. The latter part of this propulsive phase employs rapid extension at the elbow while adduction ceases.
  3. The last phase is to extract the arm once it has become non-propulsive and to round-out the movement path to conserve momentum. No propulsion occurs at this time although in the early phase of extraction momentum could be maintained.

This sequential action requires reciprocal even movements with the change of direction at each extreme of movement range requiring negative acceleration, stopping, and acceleration in the body roll.

To hold the body on the side for any length of time would require exaggerated body movements (faster roll and longer stops) which would increase the amount of wave resistance created by rolling. Also the delayed position of holding on the side would not accommodate placing the muscles in the correct position to achieve efficient propulsive movements in the sequence described above. It is impossible to produce an effective crawl/back stroke pull if the body was rotated 90 degrees to the side.

Performing the complex sequence of movements that positions an effective propulsive surface requires the body to be in particular positions at certain times for the musculature of the shoulders and arms to be correctly employed. If one tried to extend the elbow joint while on the side then the force would be applied to the side and not backward. The subtleties of movements to maintain propulsion in the most economical manner would be lost if one adopted an exaggerated side position. Body roll occurs largely to position sets of muscles to move joints to produce longitudinal forward propulsion and coupled with the arm pull produces a more direct propulsive force off the hand/forearm-propelling surface. In simpler terms, it is kinesiologically incorrect to attempt to develop maximal propulsive forces with the body rolled beyond 45 degrees. The roll should be smooth and consistent and minimize any "held" position.

In the articles by Rushall, Holt et al. (1994) and Rushall, Sprigings, et al. (1994) data from Larry Holt's laboratory in Nova Scotia illustrated when it is that a swimmer accelerates during the crawl stroke. That occurs when the arm is pushing predominantly backward. If an arm is stretched forward under water resistance is increased and a reaction force backward is developed, both of which would contribute to negative acceleration. Velocity curves in Cappaert and Rushall (1994) showed this very clearly.

In swimming, the frontal cross-sectional area does not change whether the posture is flat or vertical. What seems to be ignored is that a good crawl stroke swimmer is flat only for a moment in time as the body oscillates 90 degrees from 45 degrees on the left to 45 on right (about the average of today's great swimmers; see Rushall et al, 1998). A flat position is not held but simply a stage through which rotation passes. The only time the body should stop in a correctly rhythmed and force-applied stroke is when rotational direction is changed, that is, when it is at its greatest angle on the side which is rarely more than 45 degrees. However, a common technique error is a late breath which often causes the body to stop too long on the breathing side causing the arm position to move past the opportunity to attain an "elbow up" position. That is an error and should not be copied let alone advocated.

To summarize:

  1. objects travelling through water incur the same surface friction/resistance no matter how they lie, and
  2. objects incur the same frontal resistance no matter how they lie.

Flat floating objects incur greater hydrostatic forces under them when they move through a fluid than do deep and narrow objects. Those forces cause the objects to rise and thus diminish both surface and frontal resistance. This is why speedboats rise up, and why most boats have flat or semicircular cross-sectional profiles. Are there any ship/boat designs where the hull's shape is as narrow and deep as possible?

Swimmers rise up as they swim. That is why in a 50 m sprint, you can see more of the back, swim suit, and shoulders of a swimmer than in a 1500 m. A sprinter's greater visibility is also assisted by the bow wave being bigger and the following trough being deeper.

The whole cyberspace discussion about these matters has often been ridiculous. Crawl/back strokers do not turn on their sides to "reduce resistance." That is nonsense. They usually roll to better position limb segments to produce a more direct propulsive force, which is usually under them, resulting in better movement economy. It has nothing to do with resistance, water molecules, and other pseudo-academic concepts that are discussed out of context or incorrectly. Body roll also cancels out lateral movements that occur relative to the shoulder joint (Sanders, 1997a, 1997b).

Champion swimmers do not illustrate swimming in an exaggerated long, thin, on the side position in either crawl or backstroke. At most, body rotation is in the vicinity of 45 degrees to either side and the length of stretch forward is determined by the amount of time available to execute the stretch. A long stretch is more feasible for very tall long-limbed male distance swimmers but not exhibited by most of the world's best female distance swimmers.

In the following illustration the top picture depicts the long-side position that is frequently advocated. Below that are four world champions at their position of maximum body rotation. It should be noted at that position each champion is applying propulsive force. In the long-side position there is no propulsion being effected. None of these champions remotely resembles the long-side illustration.

It would be prudent to follow the form of the champions rather than what appears to be a staged picture.


There are several laws of physics and mechanics that cannot be ignored in swimming but seem to be dismissed by some of today's "gurus" in the sport.

The first involves Newton's First Law. If a swimmer were to stretch forward, or be in any position for that matter, and not apply propulsive force, that swimmer would negatively accelerate because of resistive forces in the fluid. To halt a loss in velocity it is desirable for a swimmer to attempt to apply force as continuously as possible. That means as one arm in crawl/back stroke ceases to accelerate or maintain the swimmer's momentum it is in the swimmer's best interests to have the other arm commence propulsive forces. The skill aspect is to have as little time as possible between propulsive phases.

Another principle is that greater speed can be obtained at the end of force application from a long stroke than a short stroke. There is more time to accelerate the load (the head, trunk, a good deal of the legs, and the non-propelling arm). Thus, one has to have the longest "effective pull" possible to achieve the fastest movement of the center of mass at the end of an arm pull. An effective pull is he propulsive phase of a stroke and now where the hand enters and exits the water. It should be noted that it is not the arm that accelerates through the water but the load that accelerates past a negatively accelerating arm.

A further principle is that if force production can be maintained by a circular movement at extremes and direction change stages, rather than stopped, less energy will be used. This is called "rounding out" and it conserves momentum. Because of this one should not straighten the arm at the end of the push back or stop it forward in a fully extended position. In crawl/back the recovery forward should be over the water. Some have advocated entering in crawl near the head and sliding forward under the water. That action would produce drag forces that would contribute to slowing the swimmer. Cappaert and Rushall (1994) also showed that phenomenon. With a reach forward over the water the inertia of the arm led by the hand finally dropping into the water can be used to shorten the period of time for repositioning until a propulsive force can be created. This is what very successful swimmers do. There are more world-record holding and champion crawl stroke swimmers who do this than do not (Evgenyi Sadovyi did not).

One of the major factors governing individual style is a swimmer's physical attributes. Technique differences between champions are usually determined by the necessity to accommodate peculiar physical dimensions while adhering to basic mechanical principles. In the case of Grant Hackett and Ian Thorpe, both are long-limbed and exceptionally tall, and the world's best in distance events. A problem that confronts tall, thin distance swimmers is the difference in duration between arm recovery and propulsion, the former being much shorter than the latter. Hackett and Thorpe demonstrate a fast recovery much in the same manner as did Sadovyi. That probably reduces the time its vertical forces are disruptive in the stroke. In Grant Hackett's case, the duration of both arm recoveries is in the vicinity of .4 seconds. However, the time spent by both arms in the propulsive phase is longer and so the recovered arm "has to be put somewhere." The end of the recovery consists of stretching forward long and straight under water while the other arm completes its propulsion. This movement is not made to "reduce resistance" or to "enhance streamline." It simply is done to accommodate the length of time this shape of swimmer takes to complete propulsion. The stretched arm is positioned in the water to minimize the transitory added resistance resulting from an increased wetted surface and the reactionary backward force. Swimmers of different shapes and physical proportions might not be required to perform with this restriction. The illustration above shows the two tall male swimmers performing differently to the two females even though Claudia Poll is very tall for a female swimmer.

It is interesting to note that in analyses of World and Olympic Champions (Cappaert & Rushall, 1994) the best swimmers exhibited a combination of good streamline and effective propulsion. The following were found to be true.

  1. The amount of force on a pull is relatively small compared to the amount of lesser swimmers' force production. Champions are rarely the strongest force-exerters.
  2. The superior streamlining of champions means less resistance so a smaller propulsive force is required for a particular speed.
  3. The efficiency ratio (total force / resistance) is highest in champions even though the total force is often less than in non-champions.
  4. The proportion of total work that produces propulsive forces is highest in champions (not all the effort in swimming contributes to propulsion).
  5. The hand/arm does not accelerate in a pull. Rather, it slows down relative to the water and the center of mass is propelled (accelerated) past it. Thus, it is very important to anchor the hand-forearm in the water and feel that one is swimming past that anchor.
  6. Even in champions, at slow speeds there is less propulsive force generated than at racing speeds.
  7. There is an optimal rate and length of effective pull that accompanies a swimmer's fastest swimming. To exaggerate either by lengthening, shortening, easing on force, or accentuating force will reduce efficiency and possibly lead to slower swimming.

Finally, the sensations of slow swimming rarely duplicate those of fast swimming. So to "feel" the same way at fast speeds as when stroking for fewest strokes per lap at training speeds would be a poor instructional strategy.

I hesitate to argue these facts against others' opinions and perceptions of what a stroke should be like. Swimming is such a complicated activity that no factor can be considered in isolation. Unfortunately, this is how many "gurus" talk about swimming techniques. There is a simple explanation for this advocacy. Swimmers are suspended in fluid. Newton's Third Law, which cannot be denied, implies to every action there is an equal and opposite reaction. Consequently, when one aspect of technique is changed there will be at least one other aspect that will also be altered as a counter-balancing reaction. Sometimes, a change for the good also produces another change for the good and of course some change for a "proposed good" also produces a bad change in some other stroke aspect.

The best index of swimming that I use in practical situations is strokes per length at intended race pace swimming velocities. When that number is fewest, and that least number can be persistently maintained while consistently swimming repetitions at race pace then you are in the "ball park" of having the stroke length which is best.

There are principles of mechanics that cannot be ignored when analyzing swimming techniques. It is folly to do so.

I hope these explanations clarify the matters of concern.


  1. Cappaert, J. M., & Rushall, B. S. (1994). Biomechanical analyses of champion swimmers. Spring Valley, CA: Sports Science Associates.
  2. Hay, J. G. (1993). The biomechanics of sports techniques (4th edition). Englewood Cliffs, NJ: Prentice Hall.
  3. Rushall, B. S., Holt, L. E., Sprigings, E. J., & Cappaert, J. M. (1994). A re-evaluation of the forces in swimming. Journal of Swimming Research, 10, 6-30.
  4. Rushall, B. S., Sprigings, E. J., Holt, L. E., & Francis, P. R. (1994). Forces in swimming--current status. NSWIMMING Coaching Science Bulletin, 2(4), 1-25.
  5. Rushall, B. S., Sprigings, E. J., Cappaert, J. M., & King, H. A. (1998). Crawl stroke body dynamics in male champions. Swimming Science Bulletin, 22, [ table.htm]
  6. Sanders, R. H. (1997a). Extending the 'Schleihauf' model for estimating forces produced by a swimmer's hand. In B. O. Eriksson & L. Gullstrand, (Eds), Proceedings of the XII FINA World Congress on Sports Medicine (pp. 421-428). Goteborg, Sweden: Chalmers Reproservice.
  7. Sanders, R. H. (1997b). Hydrodynamic characteristics of a swimmer's hand with adducted thumb: Implications for technique. In B. O. Eriksson & L. Gullstrand, (Eds), Proceedings of the XII FINA World Congress on Sports Medicine (pp. 429-434). Goteborg, Sweden: Chalmers Reproservice.

Return to Table of Contents for Carlile Coaches' Forum.