Checking the Blueprint: A Narrative Review of Bob Bowman’s 2002 Butterfly Breakdown of Michael Phelps

The 2002 session in which Bob Bowman takes Michael Phelps’ butterfly apart, drill by drill, is still on YouTube and still worth watching. I wanted to know how much of it holds up more than twenty years later. This review checks the drills against Bowman’s own footage and later interviews, places the session in the history of the stroke, and asks whether other leading coaches, and the swimmers who trained under them, say anything that supports or qualifies what he did.

Picture: the author swimming butterfly. Photograph: Thomas Anderson Photography.

Why look at this again

On 30 July 2025, at the World Championships in Singapore, Léon Marchand lowered Ryan Lochte’s 200 m individual medley world record from 1:54.00, set in 2011, to 1:52.69 in the semi-final (Koos, 2025). A record in a medley event is a good moment to go back to one of the clearest public records of how an elite butterfly stroke was built.

How this review was done

This is a narrative review. I did not run a systematic search. The starting point was my own blog post from 2025 (Locteau, 2025). I checked each drill in it against the chapter list and on-screen captions of the 2002 video, and I checked every figure and every claim about other coaches against its source. Where I could not trace a claim to a source, I left it out.

I then went through Bowman’s later material: his MP Swim Tips videos, his 2016 clinic talk and his book. For the comparison with other coaches I went back to those in Architects of the Podium, searched for published interviews and video coach by coach, and then did the same for swimmers who trained under them. I only cite remarks I could read in full, and I checked direct quotations twice. For the history I used governing-body and historian sources wherever they existed. AI tools helped with searching and drafting, as noted at the end.

A short history: the stroke and the water around it

Butterfly began as a loophole in breaststroke. In the early 1930s swimmers found that bringing both arms forward over the water was faster, and Henry Myers is credited with using the arm recovery over a full length in 1933 (Swimming.org, 2016). The dolphin kick came in 1935, developed by the Iowa coach David Armbruster and his swimmer Jack Sieg (Buchanan, 2017; Swimming.org, 2016). Volney Wilson, a physicist who used the same kick, was disqualified at the 1938 Olympic Trials because the breaststroke rules did not allow it (Buchanan, 2017). By the 1952 Olympics in Helsinki every breaststroke finalist was using the butterfly arms, and FINA voted afterwards to split the event in two (Wigo, 2023). The rules changed in 1953, so that breaststrokers had to pull at or below the surface and butterfly became a stroke of its own (SwimSwam Contributors, 2020). It made its Olympic debut in Melbourne in 1956 (Wigo, 2023).

The next change came from underneath. At the 1984 Olympics Daichi Suzuki was covering around 25 metres underwater off the 100 m backstroke start, and by Seoul in 1988 five of the eight backstroke finalists were kicking 20 metres or more, David Berkoff’s “blastoff” being the best-known example (Mortenson, 2023; Poirier-Leroy, 2015). FINA responded in two steps. A 15 m surfacing requirement entered the backstroke rules in 1991 to 1992, after swimmers began covering whole lengths underwater, and the 1998 rule book added the same requirement for butterfly and freestyle (Zaleski, 2016). Misty Hyman was winning butterfly races with 35 m underwaters by 1997 and won the Olympic 200 m butterfly in Sydney in 2000, and by Beijing in 2008 Phelps was taking eight dolphin kicks off almost every wall (Mortenson, 2023).

The water changed too. Before the 1970s lane lines were little more than floats on a rope. Wave-reducing lane lines reached the Olympic stage in Montreal in 1976, and a spinning-disc design followed in Barcelona in 1992 (Koos, 2023). Pool depth and overflow gutters matter as well, because they absorb waves that would otherwise bounce back into the outside lanes (SwimSwam, 2016). The suit changed last. Polyurethane suits arrived in 2008 and were credited with nearly 200 world records, 43 of them at the 2009 World Championships in Rome, before the ban took effect on 1 January 2010 (Wong, 2009).

The 2002 session sits between those changes, after the underwater limit and before the polyurethane suits, in a stroke that had been separate for about fifty years. That matters for how it is read. Body line, catch, kick timing and breathing do not depend on the suit, the lane rope or the length of the underwater phase. A time or a stroke count does.

What the 2002 footage shows

Bowman’s analysis centres on a small number of checkpoints that can be repeated. Phelps holds a high, horizontal body line with minimal head lift and elevated hips, which reduces frontal drag through the glide. His recovery is nearly straight-armed and relaxed, and the hands enter the water cleanly with little splash, so momentum carries into the catch. The catch is high-elbow and accelerates beneath the torso, and the arm leaves the water just before full extension. He breathes on every stroke except at starts and turns, with a small chin lift so that the body line is not broken (Bowman, 2002).

The dolphin kick stays inside the body’s own wake and is timed to the arm cycle, so it stabilises the stroke as much as it drives it (Bowman, 2002). The research points the same way in general terms. A 2022 systematic review of 25 studies found that underwater kick velocity was most closely tied to foot acceleration and kick frequency, both of which were consistently higher in better swimmers (West et al., 2022). That fits the compact kick seen on the footage.

The drill progression

The footage shows a sequence of drills. Bowman starts with side kick and reverse body dolphin to build core control and body awareness. Underwater and vertical body dolphin isolate the kick. Single-arm drills, bent and straight, work on entry and recovery on their own. A drill he calls “2 & 2 & 2” then joins partial strokes back into the full cycle. Later in the session he adds a triple-kick drill to lift the hips and steady the posture, and a “best balance and distance” drill for body position between strokes (Bowman, 2002). The same combination drill turns up years later in his MP Swim Tips series, with Phelps demonstrating it as “Butterfly Drill, 2-2 and 2” (Bowman, n.d.-b).

Counting strokes

In the Counting Strokes episode of MP Swim Tips, Bowman says the point is efficiency: as few strokes as possible at the fastest speed the swimmer can hold, with the right number differing from swimmer to swimmer (Bowman, n.d.-a). His example is the 200 m butterfly final at the 2008 Beijing Olympics, when Phelps’ goggles filled with water. Phelps swam the race blind and counted his strokes to the wall, because he knew the count he usually held, 19 to 20 strokes, and he won gold in a world record time (Bowman, n.d.-a). The count belonged to Phelps. He had practised it until he could race on it without seeing.

Resistance work and volume

The 2002 video shows resistance work in the water. Stretch cords are used for assisted speed work, described on screen as giving the swimmer “a feel of” race-pace tempo, and pulling is done without paddles so that the arms and shoulders are not overloaded (Bowman, 2002). The volume behind it is documented elsewhere. A 2004 USA Swimming record of Phelps’ sets shows him swimming ten 400 m freestyle repeats, with the interval dropping from 4:45 to a final 3:40, in a practice of roughly 9,500 yards (USA Swimming, 2004). I cover that kind of work across nine coaches, Bowman included, in Architects of the Podium (Locteau, 2026).

Starts, turns and finishes

The analysis ends at the wall. The track start is built for a narrow, fast entry. The turn uses a quick head throw into an immediate dolphin kick. The finish is taken at full speed into the wall without gliding. None of this is unique to Phelps, but training it as carefully as the stroke itself is part of what the video documents (Bowman, 2002).

Do other coaches agree?

Bowman is one voice, and a fair review should ask whether anyone else at the top of the sport says something similar. The coaches in Architects of the Podium are a useful test. None of the remarks below is about butterfly mechanics, so they cannot confirm his technical checkpoints. They speak to how he coaches.

The clearest overlap is on adversity. Gregg Troy told SwimSwam in 2018 that, with developing racers, he tries to put them in difficult positions in practice, because performing well while uncomfortable carries over to difficult moments at international level (Grace, 2018). Bowman describes the same logic in The Golden Rules, where adversity is something to train for (Bowman, 2016a), and it sits behind the goggles story in Beijing.

Troy also speaks about technique in a way that fits the drill-by-drill structure of the 2002 session. In the same interview he said that learning technique at a young age saves a swimmer later, because corrections made down the line are steps back (Grace, 2018). In a summary of one of his mentoring conversations he is reported to argue that technique should not be left to the start and end of the season, with drills used even in warm-down (Fish, 2023).

On efficiency, David Marsh made a related point in 2015 about Ryan Lochte’s experiments with a new underwater technique: everything has an energy cost, and the question is what the extra speed is worth in energy (Carroll, 2015). That is the trade-off Bowman describes when he talks about counting strokes (Bowman, n.d.-a).

The one real caution comes from Jon Urbanchek. He described trying to even out Tom Dolan’s lopsided, galloping freestyle. After about six months Dolan lost at a meet, went back to his old stroke and, as Urbanchek tells it, became the best IM swimmer in the world. Urbanchek’s conclusion was that a swimmer’s efficient, rhythmic movement is something to leave alone (Stott, 2019). Nothing in Bowman’s material contradicts this. He says himself that the right stroke count varies from swimmer to swimmer (Bowman, n.d.-a), and Dean Boxall makes a similar point from the other side, saying he does not read textbooks, he reads his athletes (Race, 2022).

What their swimmers say

Coaches describe their own methods, whereas swimmers show what reached the water. Three of the coaches in Architects of the Podium worked with butterfly swimmers at the highest level, which makes those swimmers a fairer test than the coaches’ own words.

Mark Spitz had already set butterfly world records in 1967 when he joined Counsilman’s team at Indiana. In 1970 he told Sports Illustrated that Counsilman had not done much for his strokes, and that what Counsilman gave him was psychology, which he called fantastic (“Swimming isn’t everything,” 1970). Counsilman himself said that a great coach needs two abilities, to be a good organiser and a good psychologist, and his obituary in Swimming World also calls The Science of Swimming his classic book (Colwin, 2004). Counsilman was famous for the science, yet the best-known butterfly swimmer on his team remembered the mental side. Bowman’s record covers both, the technical breakdown on camera in 2002 and the mental side in his own writing (Bowman, 2002, 2016a).

Caeleb Dressel, who set the 100 m butterfly world record at the Tokyo Olympics, spent seven years at Florida with Gregg Troy (Swimming World, n.d.-b). Poirier-Leroy (2018) describes Dressel’s up-kick as unusually powerful and quotes coach Jason Calanog saying that Dressel worked on his kick for years, with slow, perfect kicks in practice. That is close to the logic of Bowman’s session, where the kick is isolated and drilled apart from the full stroke before it goes back into it (Bowman, 2002), and it fits the finding that kick performance depends on how well the foot accelerates and how fast the kick can be repeated (West et al., 2022).

Urbanchek’s own Olympic butterfly champion was Tom Malchow, who chose Michigan partly because he liked Urbanchek. After silver in the 200 m butterfly at Atlanta in 1996, Malchow worked with Urbanchek to change his approach, and he won the event in Sydney in 2000 (Swimming World, n.d.-a).

None of this proves Bowman’s checkpoints right. The parts of his approach that can be compared (isolating the kick, treating the mental side as trainable, keeping the swimmer’s own rhythm in view) turn up in other coaches’ work too.

A drill progression builds a stroke. It should not be forced onto every swimmer.

The mental side

Bowman’s wider coaching record keeps coming back to process over outcome. In The Golden Rules he sets out adversity as something to train deliberately (Bowman, 2016a), which is the idea behind the goggles story. His MP Swim Tips series includes an episode simply called Setting Goals, where short- and long-term goal setting is treated as a skill to practise (Bowman, n.d.-c). His 2016 address to the FINA Golden Coaches Clinic makes the case for the long view across an athlete’s career (Bowman, 2016b). The 2002 footage fits that picture, because its last chapter is on goal setting.

Motion capture and AI-based stroke analysis are now an active area of research in swimming. A 2026 narrative review found most of that work still narrow and at research stage, with limited validation and few systems a coach could use day to day (Fernandes & Goethel, 2026). My own view, and it is only a view, is that these tools will measure the discipline Bowman documented, and the work itself will stay the same. You still isolate the fault, drill it in parts, rebuild the whole stroke and hold the swimmer to the same standard when tired, under pressure or with flooded goggles.

Limits of this review

There are four limits. Most of the sources are coaching media and trade press, and the biomechanics rests mainly on one systematic review (West et al., 2022), so the review is stronger on what was said and shown than on measured effects. What is known of the 2002 session comes from the footage and from Bowman himself, and there is no independent measurement of Phelps’ stroke here. The comparison draws on published remarks from some of the other coaches, not all eight, and it stays at the level of coaching principles. Swimmers’ remarks reflect what each chose to say in a given interview, and the Spitz quotation dates from 1970. Finally, the history relies on secondary sources for several dates, and where they disagree I have followed the governing-body or historian account. A systematic search, and interviews with the coaches themselves, would be the next step.

Conclusion

What circulated for years about Phelps’ butterfly training turns out to rest on sources that can be named: Bowman’s 2002 session and his own later interviews. The drill structure holds up, and the checkpoints are not tied to an era. Other leading coaches, and the swimmers who trained under them, say much the same about adversity, early technical learning, the cost of efficiency and the mental side. The one caution they add, to leave a natural stroke alone, is something Bowman’s own approach allows for. As a coach, I would take two things from it. Go back to the footage and the interviews themselves, and use drills to build a stroke around the swimmer in front of you.

About this article: I write as Head Coach of Swimming, Water Polo and Life Saving at the University of Galway, and as a PhD candidate at Trinity College Dublin. This piece began as my 2025 blog post, Butterfly Blueprint, which remains online unchanged, and has been rebuilt as a narrative review and commentary, re-checked against the footage and interviews it describes. It draws in part on the same sources used in Taisalia’s wider review of elite swim coaching, Architects of the Podium, and on published interviews with other coaches covered there and with swimmers who trained under them. Research and drafting were supported by AI tools; all sources were checked by the author. A preprint of this article is archived on Zenodo: https://doi.org/10.5281/zenodo.23621723