Content uploaded by Joel M Stager
Author content
All content in this area was uploaded by Joel M Stager on Jan 24, 2018
Content may be subject to copyright.
International Journal of Aquatic Research and Education
$41:2* ?:2'*7 79.(1*
Intra-Individual Head Depth Variability During the
Competitive Swim Start
Andrew C. Corne!
Eastern Michigan University&(473*9*2.(-*):
Hiroki Naganobori
Indiana University
Joel M. Stager
Indiana University
4114<9-.8&3)&)).9.43&1<4708&9 -A588(-41&7<4708',8:*):./&7*
@.8!*8*&7(-79.(1*.8'74:,-994=4:+47+7**&3)45*3&((*88'="(-41&7%4708"#9-&8'**3&((*59*)+47.3(1:8.43.339*73&9.43&14:73&14+
6:&9.(!*8*&7(-&3)):(&9.43'=&3&:9-47.>*)*).9474+"(-41&7%4708"#
!*(422*3)*).9&9.43
473*A3)7*<&,&34'47..740.&3)"9&,*74*1397&3).;.):&1*&)*59-$&7.&'.1.9=:7.3,9-*425*9.9.;*
"<.2"9&79 International Journal of Aquatic Research and Education$41479.(1*
./&7*
;&.1&'1*&9 -A588(-41&7<4708',8:*):./&7*;41.88
350
International Journal of Aquatic Research and Education, 2012, 6, 350-363
© 2012 Human Kinetics, Inc.
www.IJARE-Journal.com
original rESEarCH
Intra-Individual Head Depth Variability
During the Competitive Swim Start
Andrew C. Cornett, Hiroki Naganobori, and Joel M. Stager
The research on the competitive swim start primarily consists of group mean and
maximum depths with little attention given to individual variability. The purpose
of this study was to quantify intra-individual racing start depth variability and use
it to assess minimum water depth standards. Twenty-two competitive swimmers
executed ve racing starts into a water depth of 3.66 m. Intra-individual variability
was quantied by taking the standard deviation of the maximum depth of the center
of the head for the ve racing starts executed by each swimmer. The mean value
was 0.09 m with a standard deviation of 0.06 m. Analysis of means and standard
deviations showed that about one-third of swimmers would be expected to have a
head depth deeper than the current minimum water depth requirement (i.e., 1.22
m) for at least 10% of starts. Based on this research conducted in deep water, our
recommendation is that swimmers demonstrate both consistency and control of
racing start depth before being permitted to execute starts in shallow water.
Most competitive swimming races begin with the swimmers diving into the
water from a starting platform elevated above the water surface. The potential for
injury exists during the execution of this complex motor task if an athlete contacts
the pool bottom due to a dive that is too deep and/or a pool that is too shallow.
When a swimmer comes into contact with the pool bottom during the start, the
risk and severity of injury is inuenced by many factors such as (a) the swimmer’s
momentum; (b) the swimmer’s age and physical traits; (c) the orientation of the
head, neck, and torso; and (d) the location of impact and interface (Viano & Paren-
teau, 2008). Depending on the factors above, injuries may be as minor as bumps,
scrapes, or bruises and as severe as cervical vertebrae fractures or dislocations.
In an effort to prevent all such injuries from occurring during the execution
of racing starts, governing bodies in the sport mandate the minimum water depth
at the starting end of competitive swim pools. In the United States, for example,
USA Swimming (USAS), the National Federation of State High School Associa-
tions (NFHS), and the National Collegiate Athletic Association (NCAA) require a
minimum water depth of 1.22 m (4 ft) in order for swimmers to execute competitive
racing starts from a starting block (USA Swimming, 2012; National Federation of
Andrew Cornett is an assistant professor at Eastern Michigan University, School of Health Promotion
and Human Performance, Ypsilanti, MI. Hiroki Naganobori is at Indiana University in Bloomington.
Joel Stager is the Director of the Counsilman Center for the Science of Swimming at Indiana University
in Bloomington, IN, where the research was conducted.
1
Cornett et al.: Intra-Individual Head Depth Variability During the Competitive Sw
Published by ScholarWorks@BGSU, 2012
Intraindividual Head Depth Variability 351
State High School Associations, 2012; NCAA, 2011). The fact that all three of the
governing bodies in the United States have the same minimum water depth might
suggest that empirical evidence points toward this particular depth, but the extent
to which this water depth is supported by data derived through research is unclear.
The current literature on competitive swim start safety predominantly consists
of comparisons of group means for racing start depth as a method of stratifying
risk. Previous research has examined the effect of start type (Counsilman, Nomura,
Endo, & Counsilman, 1988; Gehlsen & Wingeld, 1998; Welch & Owens, 1986),
swimmer age (Cornett, White, Wright, Willmott, & Stager, 2010, 2011a, 2011b),
swimmer experience level (White, Cornett, Wright, Willmott, & Stager, 2011),
the competitive stroke performed during the rst length of competition (Cornett et
al., 2010, 2011a, 2011c), and pool depth (Blitvich, McElroy, Blanksby, Clothier,
& Pearson, 2000; Cornett et al., 2011b). Broad conclusions from this research
are that (a) older and more experienced swimmers have deeper racing starts than
their younger and less experienced counterparts; (b) starts involving steeper entry
angles, e.g., the pike entry, result in deeper starts than those utilizing a “atter”
entry type; (c) start depth varies with the competitive stroke performed during the
rst length of competition with the deepest starts associated with breaststroke and
the shallowest with the front crawl, or freestyle; and (d) start depth changes as
water depth changes.
There is value in the above research, but when it comes to establishing mini-
mum water depth rules, it seems more appropriate to consider the extreme cases,
or maximum depth values, rather than group means. The reason for this is simple:
the group mean describes the typical case and the swimmer who contacts the
pool bottom during execution of a racing start is not a typical case. Cornett et al.
(2010) demonstrated this when they reported that 14 out of 471 (3.0%) swimmers
that executed a racing start in a swimming competition in a 1.22 m (4 ft) water
depth contacted the pool bottom. A few studies have presented data on the deepest
maximum head depth observed in a variety of starting conditions and one study
(Gehlsen & Wingeld, 1998) used this maximum value to suggest revisions to
the minimum water depth standards. Gehlsen and Wingeld (1998) analyzed the
depths and speeds of collegiate swimmers performing different types of starts from
a variety of starting block heights and concluded that 1.22 m was not an adequate
minimum water depth. Because none of the swimmers attained a head depth deeper
than 1.4 m in the study, they recommended that minimum water depth be “greater
than 1.4 m for experienced divers” (p. 30).
Using the group maximum depth to establish minimum water depth stan-
dards—as Gehlsen and Wingeld (1998) did—is perhaps more appropriate than
utilizing group mean depths, but there are issues with this approach as well. The
problems might be the inherent assumptions that (a) swimmers dive to the same
depth every time they execute a racing start and (b) the swimmer who performs the
deepest observed racing start is at the greatest relative risk. Our current hypotheses
are that (a) there is intra-individual variability associated with racing start depth
and (b) a swimmer with high racing start depth variability might be at greater risk
of impacting the pool bottom than another swimmer that has a deeper mean racing
start depth but lower intra-individual variability (Figure 1). To date, our hypotheses
have gone untested because we have been unable to locate any studies that present
intra-individual racing start depth variability when the same swimmer executes
2
International Journal of Aquatic Research and Education, Vol. 6, No. 4 [2012], Art. 9
https://scholarworks.bgsu.edu/ijare/vol6/iss4/9
DOI: 10.25035/ijare.06.04.09
352 Cornett, Naganobori, and Stager
multiple racing starts from the same starting block into the same water depth using
the same start type.
Thus, the purposes of this study were (a) to quantify intra-individual racing start
depth variability when the same swimmer executes multiple starts under the same
experimental conditions; (b) to determine whether or not intra-individual racing
start depth variability is related to swimmer characteristics such as age, height,
mass, and competitive experience as a means of further stratifying relative risk;
and (c) to use individual mean and variability values to assess current minimum
pool depth regulations.
Method
Participants
The university’s Human Subjects Committee approved the project prior to the ini-
tiation of the study and informed consent was obtained from each participant and
his/her guardian (when younger than 18 years of age). The participants were all
members of a competitive swim club in the Midwestern United States. The swim-
mers’ age, height, mass, and competitive swimming experience were recorded prior
to data collection. Because a focus of the study was to determine which factors,
if any, were related to racing start depth variability during the competitive racing
start, a wide range for age, height, mass, and competitive swimming experience
was desirable.
Figure 1 — The racing start depth distribution for swimmer A (solid line) has a deeper
mean starting depth but lower variability than the distribution for swimmer B (dashed line).
In this hypothetical situation, the probability that swimmer B will contact the pool bottom
in a 1.22 m pool is greater than for swimmer A.
3
Cornett et al.: Intra-Individual Head Depth Variability During the Competitive Sw
Published by ScholarWorks@BGSU, 2012
Intraindividual Head Depth Variability 353
Procedures
The testing took place in a swimming venue that consisted of a six-lane competition
pool (22.86 m × 13.70 m) and a separate diving well (12.83 m × 10.96 m). Data
collection was conducted in the diving well and the water depth was 3.66 m (12 ft)
where the swimmers entered the water. No other activity took place in the facility
during testing. A portable starting block with a standard platform height of 0.76
m above the water surface was custom designed and built for this project (Adolph
Kiefer and Associates, Zion, IL). The block was mounted on a steel platform that
provided the ability to move the starting block to various locations on the pool
deck. The start platform was inclined at an angle of 10° from horizontal and had
a surface area of 0.39 m2.
All swimmers performed ve competitive starts from the standard starting
platform into the diving well. For all starts, swimmers were asked to complete a
racing start and a subsequent front crawl, or freestyle, sprint across the diving well.
In order to mimic a typical competitive situation, swimmers were asked to step onto
the block, to take their mark, and then the start was initiated with an audio signal
from a commercial starting system (Daktronics, Omnisport HS 100, Brookings, SD).
Video Recording, Calibration, and Video Analysis
The video recording, calibration, and video analysis procedures have been previ-
ously described (Cornett et al., 2011c). Briey, the underwater portion of the dive
start was lmed using a Canon GL2 digital video camcorder (Canon Inc., Tokyo,
Japan) enclosed in a sealed housing unit (Ikelite Underwater Systems, Indianapolis,
IN) and mounted on a tripod on the bottom of the diving well. A Canon WD-58
wide-angle adapter (Canon Inc., Tokyo, Japan) was used to ensure that the eld
of view included the subjects’ underwater motions from entry until farther than
the deepest point of the dive. The video signal was captured using SIMI Motion
software (zFlo Inc., Quincy, MA).
The dive area in front of each block location was calibrated using a custom-
built 1 m × 3 m aluminum frame was placed vertically in line with the center of the
starting block, perpendicular to the side of the pool, and with the top of the frame
about 0.1 m below the surface of the water. The origin of the coordinate system
was at water level directly below the center of the starting block and the axes were
oriented such that the x-axis pointed horizontally and perpendicular to the wall
and the y-axis pointed vertically upward.
Following the calibration of the dive area, the competitive dives were recorded
and analyzed using SIMI Motion. In each dive, the external auditory meatus, an
anatomical landmark for the center of the subject’s head, was manually digitized
beginning with the frame in which it was rst visible. For more detail on the pro-
cedures, see Cornett et al. (2011c).
Data Analysis
The maximum depth of the center of the head (DEPTH; meters) was determined for
each trial. Intra-individual racing start depth variability was quantied in this study
by taking the standard deviation of the maximum depth of the center of the head
for the ve racing starts executed by each swimmer (SD_DEPTH; meters). Pearson
4
International Journal of Aquatic Research and Education, Vol. 6, No. 4 [2012], Art. 9
https://scholarworks.bgsu.edu/ijare/vol6/iss4/9
DOI: 10.25035/ijare.06.04.09
354 Cornett, Naganobori, and Stager
product moment correlation coefcients were computed between SD_DEPTH and
ve variables: (a) age (AGE; years), (b) height (HEIGHT; meters), (c) mass (MASS;
kilograms), (d) competitive swimming experience (EXPERIENCE; years), and (e)
the mean value for the maximum depth of the center of the head for the ve racing
starts executed by each swimmer (MEAN_DEPTH; meters). All calculations and
analyses were performed using IBM SPSS Statistics (Version 19.0; IBM, Armonk,
NY), and an alpha level of 0.05 was used to determine statistical signicance for
all inferential procedures.
In order to assess current minimum water depth standards, we rst made an
assumption that DEPTH was normally distributed for each swimmer. An inspec-
tion of normal probability plots for DEPTH for the ve racing starts executed
by each swimmer indicated that this assumption was justied. Then, 0.15 m was
added to the MEAN_DEPTH for each swimmer to create an adjusted head depth
variable (ADJUSTED_MEAN_DEPTH) because the external auditory meatus, the
landmark for the center of the head, is approximately 0.15 m shallower than the
deepest point of the head (Blitvich et al., 2000). Once this adjustment was made,
we used ADJUSTED_MEAN_DEPTH and SD_DEPTH to determine the expected
percentage of starts for each swimmer deeper than 1.22 m. This depth was of par-
ticular interest because it is the current minimum water depth in which swimmers
are allowed to execute racing starts in the United States (USA Swimming, 2012;
National Federation of State High School Associations, 2012; NCAA, 2012). We
calculated the expected percentage of starts deeper than 1.22 m for each swimmer
by rst calculating a z score using the following equation:
z
XX
s
=−
where X was set to 1.22 m for each swimmer, X
– was the ADJUSTED_MEAN_
DEPTH, and s was SD_DEPTH. Next, the area under the curve beyond each z
score was determined using the Gaussian, or Normal, distribution and that value
was used as the estimated percentage of starts deeper than 1.22 m for each swim-
mer. A sample calculation with an ADJUSTED_MEAN_DEPTH of 1.15 m and
SD_DEPTH of 0.05 m is shown in Figure 2.
Results
Twenty-two swimmers were recruited to participate in this study and the means
and standard deviations for AGE, HEIGHT, MASS, and EXPERIENCE were 17.1
± 4.4 years, 168.6 ± 12.5 cm, 65.6 ± 16.3 kg, and 6.5 ± 4.9 years, respectively.
MEAN_DEPTH, ADJUSTED_MEAN_DEPTH, and SD_DEPTH for each subject
are shown in Table 1. One of the swimmers was identied as an outlier because the
standardized score, or z score, for SD_DEPTH for the swimmer 0.29 0.09
0.06 3.33
−=
was
greater than 3.29, which has been previously identied as a cutoff for univariate
outliers (Tabachnick & Fidell, 2007). As a result, this swimmer was excluded from
the data set during the correlational analysis but was included in the assessment
of pool depth regulations.
5
Cornett et al.: Intra-Individual Head Depth Variability During the Competitive Sw
Published by ScholarWorks@BGSU, 2012
Intraindividual Head Depth Variability 355
Pearson correlation coefcients revealed that SD_DEPTH was signicantly
correlated with MEAN_DEPTH (r = 0.66, p = 0.001) but not with any of the subject
characteristics (i.e., AGE, HEIGHT, MASS, and EXPERIENCE; p > .05).
Finally, when using ADJUSTED_MEAN_DEPTH and SD_DEPTH to assess
pool depth regulations for racing starts, the z scores for the swimmers for a depth of
1.22 m ranged from –2.29 to 17.00. The percentage of starts estimated to be deeper
than 1.22 m for each swimmer— determined using the z scores and the Gaussian,
or Normal, distribution— ranged from less than 0.01% to 98.90% (Table 1).
Discussion
Due to the potentially catastrophic nature of injuries resulting from swimmers
contacting the pool bottom when executing a racing start, it seems that minimum
water depth rules need to be reinforced by empirical data. The competitive swim
start safety literature primarily consists of reports of mean and maximum depths
for particular groups or experimental conditions. The present study suggests that
swimmers have unique distributions for racing start depth and, as a result, individual
Figure 2 — Sample calculation for the estimated percentage of starts deeper than 1.22 m
for each swimmer where X is the depth of interest, X
– is the mean value for the maximum
depth of the center of the head from ve racing starts with a 0.15 m adjustment added to it
to estimate the actual deepest point of the head, s is the standard deviation for the maximum
depth of the center of the head from ve racing starts, and z is the z score. The area shaded
in black is the area under the curve beyond a z score of 1.40 and corresponds to 8.08% of
starts deeper than 1.22 m for this swimmer.
6
International Journal of Aquatic Research and Education, Vol. 6, No. 4 [2012], Art. 9
https://scholarworks.bgsu.edu/ijare/vol6/iss4/9
DOI: 10.25035/ijare.06.04.09
356
Table 1 Means, Standard Deviations, z Scores, and Percentage of Starts Deeper Than 1.22 m
for the Five Racing Starts Executed by Each Swimmer
Subject MEAN_DEPTH
ADJUSTED_
MEAN_DEPTH SD_DEPTH
z Score for
1.22 m
% of Starts Deeper
Than 1.22 m
1 1.55 1.70 0.21 -2.29 98.90
2 1.20 1.35 0.18 -0.72 76.42
3 1.01 1.16 0.09 0.67 25.14
4 0.96 1.11 0.12 0.92 17.88
5 0.79 0.94 0.29 0.97 16.60
6 0.98 1.13 0.08 1.13 12.92
7 1.02 1.17 0.04 1.25 10.56
8 0.95 1.10 0.09 1.33 9.18
9 1.00 1.15 0.05 1.40 8.08
10 0.93 1.08 0.07 2.00 2.28
11 0.86 1.01 0.10 2.10 1.79
12 0.86 1.01 0.09 2.33 0.99
(continued)
7
Cornett et al.: Intra-Individual Head Depth Variability During the Competitive Sw
Published by ScholarWorks@BGSU, 2012
357
Subject MEAN_DEPTH
ADJUSTED_
MEAN_DEPTH SD_DEPTH
z Score for
1.22 m
% of Starts Deeper
Than 1.22 m
13 0.91 1.06 0.05 3.20 0.07
14 0.65 0.80 0.10 4.20 <0.01
15 0.75 0.90 0.06 5.33 <0.01
16 0.78 0.93 0.05 5.80 <0.01
17 0.65 0.80 0.07 6.00 <0.01
18 0.55 0.70 0.08 6.50 <0.01
19 0.55 0.70 0.07 7.43 <0.01
20 0.54 0.69 0.07 7.57 <0.01
21 0.53 0.68 0.05 10.80 <0.01
22 0.73 0.88 0.02 17.00 <0.01
Note. MEAN_DEPTH and SD_DEPTH are the mean and standard deviation for the maximum depth of the center of the head for ve
racing starts. ADJUSTED_MEAN_DEPTH is the MEAN_DEPTH plus 0.15 m to account for the distance from the center of the head to
the deepest part of the head.
Table 1 (continued)
8
International Journal of Aquatic Research and Education, Vol. 6, No. 4 [2012], Art. 9
https://scholarworks.bgsu.edu/ijare/vol6/iss4/9
DOI: 10.25035/ijare.06.04.09
358 Cornett, Naganobori, and Stager
variability in racing start depth should be considered, if appropriate, data-driven
minimum water depth standards are to be developed.
To briey review, there were three purposes of this study relating to intra-
individual variability: (a) to quantify it; (b) to determine which factors, if any, were
correlated with it; and (c) to use it to assess current minimum water depth regula-
tions. The major ndings regarding the rst two purposes were that (a) the range
of values for intra-individual variability is quite large in a competitive sample and
(b) intra-individual variability is positively correlated with the average depth of
the swimmer’s racing start. Two of the relevant results related to the third purpose,
the assessment of minimum water depth regulations, were that when executing
racing starts in deep water: (a) approximately one-third of swimmers are expected
to execute at least 1 in 10 starts deeper than 1.22 m, the current minimum water
depth for racing starts, and (b) the deepest part of the head for 2 of the 22 swim-
mers was, on average, deeper than 1.22 m. The results for the three purposes are
discussed independently in this section.
Intra-Individual Racing Start Depth Variability
Three racing start safety studies (Counsilman et al., 1988; Gehlsen & Wingeld,
1998; Welch & Owens, 1986) asked swimmers to execute two or three racing starts
under the same experimental conditions, but these previous studies did not report
individual means or standard deviations. Instead, in these early studies, the data
were pooled and presented as group means for each starting condition. As a result,
they do not provide values to which to we can compare our data.
In this regard, the standard deviation for the maximum depth of the center of
the head (SD_DEPTH) for the ve racing starts executed by the swimmers lmed
in the present study had a mean value of 0.09 m, a standard deviation of 0.06 m,
and ranged from 0.02 to 0.29 m. The range of values for SD_DEPTH seems large
as SD_DEPTH was nearly 15 times greater for the most variable swimmer than
for the least variable one. The practical signicance of the difference between
SD_DEPTH values of 0.02 and 0.29 m can be demonstrated with an example.
Although MEAN_DEPTH differed by 6 cm between the two swimmers with the
highest and lowest SD_DEPTH values in this study, consider the hypothetical
situation if the two swimmers, Swimmer A and Swimmer B, had had the same
MEAN_DEPTH value (e.g., 1.0 m) but SD_DEPTH values of 0.02 and 0.29 m,
respectively. Assuming that DEPTH is normally distributed for each swimmer,
we can use the standard normal distribution to determine the percentage of starts
within a range of depths. Swimmer A, with a MEAN_DEPTH of 1.0 m and an
SD_DEPTH of 0.02 m, would be expected to have a DEPTH between 0.98 and 1.02
m for about two-thirds of starts and between 0.96 and 1.04 m for approximately
95% of starts. This means that DEPTH for 19 out of 20 starts for Swimmer A would
fall within an 8 cm (3.15 in) window. On the other hand, Swimmer B, with the
same MEAN_DEPTH of 1.0 m and a SD_DEPTH of 0.29 m, would be expected to
have a DEPTH between 0.71 and 1.29 m for about two-thirds of starts and between
0.42 and 1.58 m for approximately 95% of starts. For Swimmer B, then, DEPTH
for 19 out of 20 starts would fall within a 1.16 m (45.67 in) window. The range of
DEPTH values expected for Swimmer B seems substantial considering that the
mandated minimum water depth for racing starts is 1.22 m (USA Swimming, 2012;
9
Cornett et al.: Intra-Individual Head Depth Variability During the Competitive Sw
Published by ScholarWorks@BGSU, 2012
Intraindividual Head Depth Variability 359
National Federation of State High School Associations, 2012; NCAA, 2011). This
hypothetical example demonstrates that it is not sufcient to assess swimmer risk
during the execution of the racing start by evaluating a single head depth value,
or even the average depth to which a swimmer dives; intra-individual racing start
depth variability must be considered as well.
Factors Correlated with Intra-Individual Racing Start Depth
Variability
Because SD_DEPTH varies from swimmer to swimmer, and variability may be
interpreted to add to swimmer risk, it was decided to examine the factors that
maybe causal to or coincident with this intra-individual variability. The signicant
positive correlation between SD_DEPTH and MEAN_DEPTH suggests that intra-
individual variability increases as the average depth of the start increases. While
we cannot know for certain if a causal relationship exists between these variables,
we originally hypothesized that SD_DEPTH would increase with MEAN_DEPTH.
We expected this relationship simply because there is more “room” for start depth
to vary for swimmers that, on average, dive deeper. Since this study was conducted
in a diving well with a water depth of 3.66 m, it would seem that SD_DEPTH was
not constrained by the pool bottom. That is, the water was likely deep enough that
the pool bottom did not prevent racing start depth from varying in that direction.
The same cannot necessarily be said of the water surface though. It seems that the
water surface would place a greater constraint on SD_DEPTH for shallower starts
than deeper ones. We recognize that this is speculation and recommend that future
research replicate the present study in different water depths as a means to conrm
or reject this hypothesis.
The signicant positive correlation between SD_DEPTH and MEAN_DEPTH
is interesting but not necessarily practically useful. MEAN_DEPTH is measured
after a swimmer has executed a racing start so it cannot be used to identify the most
variable swimmers a priori. Of the subject characteristics that could be measured
before the execution of a racing start (i.e., AGE, HEIGHT, MASS, and EXPERI-
ENCE), none were correlated with SD_DEPTH.
Common sense would argue that as swimmers gain competitive swimming
experience, racing start depth variability would decrease. Our previous research
demonstrated that experienced competitive swimmers perform deeper racing starts
than those of inexperienced competitive swimmers (White et al., 2011). Because
more experienced swimmers tend to execute deeper starts and racing start depth
variability increases with start depth, we hypothesized that a correlation might exist
between SD_DEPTH and EXPERIENCE if we controlled for MEAN_DEPTH.
However, no relationship was found when we conducted a partial correlation and
we can only speculate as to why our hypothesis concerning this relationship proved
unconrmed.
It may be that our original hypothesis differed from the hypothesis we actually
tested. We initially expected that as a swimmer accumulated practice time executing
racing starts the swimmer’s racing start consistency would improve (as reected by
a decrease in SD_DEPTH). However, we did not measure accumulated racing start
practice time; we measured competitive swimming experience. We did this because
while it is reasonable to expect swimmers to recall the number of years they have
10
International Journal of Aquatic Research and Education, Vol. 6, No. 4 [2012], Art. 9
https://scholarworks.bgsu.edu/ijare/vol6/iss4/9
DOI: 10.25035/ijare.06.04.09
360 Cornett, Naganobori, and Stager
been involved in competitive swimming, it seemed unlikely that swimmers could
precisely estimate the amount of practice time spent on racing starts during the
course of their careers. Although we assume that the two variables are correlated,
we could nd no literature describing the nature of this relationship. Future research
should focus on quantifying the amount of daily, weekly, monthly, or seasonal
practice time coaches and swimmers specically dedicate to the racing start.
Assessment of Minimum Water Depth Regulations
This is the rst study to use intra-individual variability when assessing racing start
depth requirements. The results of this analysis demonstrate that when performing
racing starts in relatively deep water (in this study, 3.66 m), competitive swimmers
regularly execute starts with head depths close to or greater than 1.22 m, the cur-
rent minimum water depth requirement. The deepest part of the head for 6 of 22
(27%) swimmers was, on average, within one-tenth of a meter of 1.22 m, which
provides a small margin for error. For two of the swimmers in the study, the deepest
part of the head was, on average, deeper than 1.22 m. And nally, one-half of the
swimmers in the study are expected to have a head depth deeper than 1.22 m in a
water depth of 3.66 m for at least 1 in 100 starts. Based on these numbers alone,
it seems that 1.22 m is not an appropriate minimum water depth for the execution
of competitive swim starts.
There have been two other studies in the literature that assessed minimum
water depth requirements by having swimmers execute racing starts in relatively
deep water (Gehlsen & Wingeld, 1998; Welch & Owens, 1986). As occurred in
the current study, both of these previous studies observed swimmers with head
depths deeper than 1.22 m when performing starts in deep water. Welch and Owens
(1986) lmed collegiate swimmers executing racing starts from starting blocks of
varying heights in a water depth of 3.81 m (12.5 ft). They concluded that minimum
water depth should be increased to 1.37 m (4.5 ft) to “provide an increase margin of
safety” (p. 6) to the swimmers. Gehlsen and Wingeld (1998) were also interested
in the effect of start type (pike and at) and starting platform height (0.46, 0.56,
0.66, and 0.76 m) on racing start parameters when collegiate swimmers execute
racing starts in a water depth of 4 m (13.1 ft). Because of the observation that none
of the swimmers “ascended” to a depth greater than 1.4 m, Gehlsen and Wingeld
(1998) recommended that pool water depth should be at least 1.4 m for “experience
divers.” Based solely on the head depths (and not depth variability) of the swim-
mers in our study, we would agree with Welch and Owens (1986) and Gehlsen and
Wingeld (1998) in that 1.22 m is not an appropriate minimum water depth for
the execution of racing starts; however, the maximum head depth value recorded
in deep water is not the only factor that requires consideration.
It is important to consider the effect of water depth on racing start depth when
evaluating minimum water depth regulations. Blitvich et al. (2000) analyzed the
racing start parameters of 36 junior elite swimmers when executing starts in two
different water depths: 1.2 m and 2.0 m. They found that mean maximum head
depth was signicantly deeper in a 2.0 m water depth (0.88 m) than in a 1.2 m
water depth (0.79 m). Subsequent studies have conrmed the ndings of Blitvich
et al. for racing starts in competition (Cornett et al., 2011a) and in different water
depths (Cornett et al., 2011d). From their study on the effect of water depth on
11
Cornett et al.: Intra-Individual Head Depth Variability During the Competitive Sw
Published by ScholarWorks@BGSU, 2012
Intraindividual Head Depth Variability 361
racing start depth, Blitvich et al. concluded that minimum water depth require-
ments “appear safe for use by skilled competitive swimmers” (p. 38) but added
that “diving skills instruction should be provided for swimmers prior to performing
dives in water of 1.2 m” (p. 38), and coaches should be certain their swimmers
are aware of water depth in the competition venue. Thus, Blitvich et al. came to
a different conclusion concerning the minimum water depth for racing starts than
Welch and Owens (1986) and Gehlsen and Wingeld (1998) by suggesting that
1.2 m is a sufcient minimum water depth provided appropriate instructional and
educational practices are in place. The major difference between the studies is that
the former study considered the effect of water depth on racing start parameters
while the latter studies did not.
Because of this relationship between racing start parameters and water depth
and because the starts in this study were lmed in deep water, Table 1 may over-
estimate the risk of swimmers contacting the pool bottom during the execution
of racing starts in a 1.22 m water depth. We can adjust the individual means and
standard deviations obtained in this study to account for the effect of water depth
by using values from a previous study on racing start safety. Cornett et al. (2011d)
reported, among other things, the means and standard deviations for maximum head
depth for competitive swimmers executing starts in water depths of 1.53 m and
3.66 m. In their study, mean racing start depth was 0.17 m shallower in a 1.53 m
water depth than in a 3.66 m water depth and the standard deviation for maximum
head depth was 38% lower in 1.53 m water depth as compared to a 3.66 m water
depth. We used these values to “correct” our ADJUSTED_MEAN_DEPTH and
SD_DEPTH and then recalculated the z scores and the expected percentage of
starts deeper than 1.22 m. After the water depth correction, swimmers, as expected,
performed fewer racing starts deeper than 1.22 m. The majority of swimmers (19
out of 22, or 86%) seemed to have minimal risk as they were expected to perform
fewer than 1 in 10,000 starts deeper than 1.22 m. Two of the swimmers, however,
were expected to perform more than 1 in 3 starts deeper than 1.22 m and one of
these two still had a mean head depth in excess of 1.22 m.
When we adjusted the mean and variability values in this study to account for
the effect of water depth, swimmers were expected to perform fewer racing start
deeper than 1.22 m, but two of the swimmers still seemed to be at risk. The problem
is that the average water depth correction applied to all swimmers in this study may
have under-corrected for some swimmers and these two swimmers specically.
The two swimmers who executed the deepest racing starts in deep water are also
the ones who will need to make the largest depth adjustments when moving to a
shallower water depth. The important point is that swimmers must recognize pool
water depth differences and modify start depth appropriately if they are to perform
racing starts in 1.22 m safely. If a swimmer lacks the ability to consistently control
and modify racing start depth, that swimmer could be at a great risk of injury when
performing racing starts in shallow water.
Conclusions and Recommendations
This is the rst study to quantify intra-individual racing start depth variability by
having the same swimmer execute multiple racing starts in the same experimental
conditions. From this analysis, we conclude that it is not sufcient to assess only
12
International Journal of Aquatic Research and Education, Vol. 6, No. 4 [2012], Art. 9
https://scholarworks.bgsu.edu/ijare/vol6/iss4/9
DOI: 10.25035/ijare.06.04.09
362 Cornett, Naganobori, and Stager
the maximum head depth from a single racing start, or even the mean value from
multiple starts; racing start depth variability must be considered as well. The risk
that the swimmer will contact the pool bottom increases as racing start depth vari-
ability increases; however, consistent racing start depth alone does not mean that
a swimmer is safe executing starts in shallow water. If the swimmer consistently
performs deep racing starts, he or she is in danger of contacting the pool bottom. It
is also critical that swimmers can accurately perceive water depth—or take steps to
become aware of it—and make the appropriate modications to racing start depth.
As a result, we recommend that swimmers demonstrate racing start depth
consistency and control before being permitted to execute competitive racing starts
in shallow water. For example, USA Swimming currently requires that swimmers
11 years of age and older with at least one year of previous competitive experi-
ence demonstrate that they are capable of safely controlling racing start depth.
We recommend additional requirements before swimmers are permitted to start in
shallow water. First, swimmers should be observed to execute racing starts with
depth consistency by performing multiple racing starts in deep water with minimal
racing start depth variability. Once this has been accomplished, the swimmers
should demonstrate depth control by displaying the ability to consistently execute
shallow racing starts.
Acknowledgments
Financial support for the project was received from USA Swimming. Ron Van Pool and Carol
Zaleski were fundamental in providing support and important input. Paul Sigfusson, DDS,
deserves special recognition as liaison with USA Swimming and project advisor. Finally, we
would like to thank John Petersen of Risk Management Services, Inc. and Murray Stevens
of North Baltimore Aquatic Club for encouragement and constructive feedback.
References
Blitvich J.D., McElroy, G.K., Blanksby, B.A., Clothier, P.J., & Pearson, C.T. (2000). Dive
depth and water depth in competitive swim starts. Journal of Swimming Research,
14, 33-39.
Cornett, A.C., White, J.C., Wright, B.V., Willmott, A.P., & Stager, J.M. (2010). Racing start
safety: Head depth and head speed during competitive swim starts into a water depth
of 1.22 m. International Journal of Aquatic Research and Education, 4, 365-378.
Cornett, A.C., White, J.C., Wright, B.V., Willmott, A.P., & Stager, J.M. (2011a). Racing
start safety: Head depth and head speed during competitive swim starts into a water
depth of 2.29 m. International Journal of Aquatic Research and Education, 5, 14-31.
Cornett, A.C., White, J.C., Wright, B.V., Willmott, A.P., & Stager, J.M. (2011b). Water depth
inuences the head depth of competitive racing starts. International Journal of Aquatic
Research and Education, 5, 32-41.
Cornett, A.C., White, J.C., Wright, B.V., Willmott, A.P., & Stager, J.M. (2011c). Block
height inuences the head depth of competitive racing starts. International Journal of
Aquatic Research and Education, 5, 174-186.
Cornett, A.C., White, J.C., Wright, B.V., Willmott, A.P., & Stager, J.M. (2011d). Racing start
safety: Head depth and head speed during competitive backstroke starts. International
Journal of Aquatic Research and Education, 5, 389-401.
Counsilman, J., Nomura, T., Endo, M., & Counsilman, B. (1988). A study of three types of
grab start for competitive swimming. National Aquatics Journal, 2-6.
13
Cornett et al.: Intra-Individual Head Depth Variability During the Competitive Sw
Published by ScholarWorks@BGSU, 2012
Intraindividual Head Depth Variability 363
Gehlsen, G.M., & Wingeld, J. (1998). Biomechanical analysis of competitive swimming
starts and spinal cord injuries. Journal of Swimming Research, 13, 23-30.
National Federation of State High School Associations. (2012). 2012-13 NFHS swimming
and diving and water polo rules book. Indianapolis, IN: National Federation of State
High School Associations.
National Collegiate Athletic Association. (2011). 2012 and 2013 NCAA men’s and women’s
swimming and diving rules. Indianapolis, IN: National Collegiate Athletic Association.
Tabachnick, B.G., and Fidell, L.S. (2007). Using multivariate statistics, 5th ed.Boston:
Allyn and Bacon.
USA Swimming. (2009). Racing start certication checklist. Colorado Springs: USA
Swimming.
USA Swimming. (2012). 2012 USA swimming rules and regulations. Colorado Springs:
USA Swimming.
Viano, D.C., & Parenteau, C.S. (2008). Analysis of head impacts causing neck compression
injury. Trafc Injury Prevention, 9, 144-152.
Welch, J., & Owens, W. (1986). Water depth requirements of competitive racing starts.
Swimming Research, 2(3), 5-7.
White, J.C., Cornett, A.C., Wright, B.V., Willmott, A.P., & Stager, J.M. (2011). Competitive
swimmers modify racing start depth upon request. International Journal of Aquatic
Research and Education, 5, 187-198.
14
International Journal of Aquatic Research and Education, Vol. 6, No. 4 [2012], Art. 9
https://scholarworks.bgsu.edu/ijare/vol6/iss4/9
DOI: 10.25035/ijare.06.04.09
