Journal of Strength and Conditioning Research, 2006, 20(4), 843–850
? 2006 National Strength & Conditioning Association
KINEMATICAL ANALYSIS OF THE SNATCH IN ELITE
MALE JUNIOR WEIGHTLIFTERS OF DIFFERENT
JOSE´CAMPOS,1PETR POLETAEV,2ANDRE´S CUESTA,3CARLOS PABLOS,1AND
1Laboratory of Biomechanics (Faculty of Physical Education and Sport Sciences), Department of Sport and
Physical Education, University of Valencia, Valencia, Spain;2Russian Weightlifting Federation, Moscow, Russia;
3Department of Nursery, University of Valencia, Valencia, and Spanish Weightlifting Federation, Madrid, Spain.
ABSTRACT. Campos, J., P. Poletaev, A. Cuesta, C. Pablos, and
V. Carratala ´. Kinematical analysis of the snatch in elite male
junior weightlifters of different weight categories. J. Strength
Cond. Res. 20(4):843–850. 2006.—The purpose of this study was
to analyze the differences in the technical pattern of the snatch
in elite junior weightlifters of different weight categories. The
sample was a group of 33 men weightlifters from different
weight categories. The comparative study included 2 groups,
taking into account weight categories. Group A included 17
weightlifters from the lightest categories, 56 and 62 kg; group
B included 16 weightlifters from the heaviest categories, 85 and
105 kg. Three-dimensional photogrammetry technique was uti-
lized. Regarding group differences, we can conclude that lifters
belonging to heavier categories are more efficient, as they man-
age to have longer barbell propulsion trajectories, which allows
them to exert actions on the barbell for a longer period, espe-
cially in the initial lifting phase. They attain greater barbell ver-
tical velocity (p ? 0.029), a longer vertical bar trajectory nor-
malized on first pull (p ? 0.011), and a greater, although limited,
bar height loss on the catch (p ? 0.008). Besides, intergroup
differences evidence that heavier category lifters observe a dif-
ferent temporal organization of the movement based on a longer
first pull (p ? 0.000), a shorter transition (p ? 0.030), and a
longer turnover (p ? 0.049). No significant differences were
found in the analyzed angular parameters during the first and
second pull. We believe the intergroup differences found not to
be determining enough to consider a technical model adapted to
the characteristics of each body weight category. This confirms
that a successful lift is multifactor based and individual depen-
dent. Given its transcendence, this evidence should be taken into
account in the technical training of young lifters.
KEY WORDS. biomechanics, weightlifting, power, technique
high level of muscular power during the lifting as well as
reaching an effective transference of that power to the bar
in a short period of time. The snatch is a characteristic
movement of maximum power in which speed and coor-
dination play a decisive role. As an acyclic movement, the
snatch demands high coordination, as per Bernstein’s (4)
idea of a kinematical chain in which each link adds to the
creation of reactive or reflex forces transferred by one an-
other, ultimately making up an ideal pattern of time or-
ganization. To a great extent, the final result of the
snatch is conditioned by the actions taken in the first and
ne of the most important aims of weightlifting
sports is to develop a technique that enables
athletes to lift heavy weights. Coaching con-
sists of training weightlifters to generate a
second pull. During such phases, action is exerted on the
barbell with a view to attaining the maximum vertical
From the biomechanical point of view, several studies
have described the movements of the bar and the lifter.
Some of them have been used as reference in our study,
such as those carried out by Lukashev (22), Vorobyev (28,
29), Gue (19), Bartonietz (2), Bauman et al. (3), Isaka et
al. (21), Stone et al. (27), Gourgoulis et al. (18), Schilling
et al. (25), and Campos and Poletaev (5). These papers
provide detailed information on the behavior of the pa-
rameters accounting for the lifters’ maximum perfor-
mance. One of the most interesting areas of analysis is
that of kinematical and dynamic parameters used by lift-
ers under competitive conditions. In this respect, the
works by Ono et al. (24), Enoka (8, 9), Garhammer (12–
14, 17), Baumann et al. (3), and Isaka et al. (21) are note-
worthy. Furthermore, some research lines in this field
have focused on differences between lifters with different
skill, weight, and performance levels (3, 10, 12, 19, 22).
However, papers focusing on differences between differ-
ent body-weight category lifters are most scarce.
The aim of the study is to analyze and compare the
biomechanical profiles of lifters of different weight cate-
gories in competitive conditions. The techniques of the
snatch of male weightlifters were analyzed during the
2003 European Junior Championships in Valencia
(Spain). We believe the analysis of the technique of junior
lifters to be extremely interesting due to the fact that they
are in an initial phase of high performance. In fact, some
studies stated that the young lifters are already at a high
level of snatch technique (19).
Experimental Approach to the Problem
There is general agreement that highly skilled athletes
employ an optimum sequential pattern of intersegmental
coordination and produce longer barbell positive acceler-
ation phases when compared to the less skilled (10). In
weightlifting particularly, where body weight determines
competence categories, a possibility is suggested that dif-
ferent structures and models adapted to the lifters’ mor-
phological features might exist, establishing frontiers in
timing and motion fluidity. In fact, Vorobyev (29) and
Bartonietz (2) argue that from the dynamic viewpoint lift-
ers with shorter levers are under worse conditions, since
844CAMPOS, POLETAEV, CUESTA ET AL.
TABLE 1. Sample characteristics.*
(Cat. 56 and 62 kg)
(n ? 17)
Ivan Herna ´ndez
(Cat. 85 and 105 kg)
(n ? 16)
* Official Results European Junior Championships (2003) (European Weightlifting Federation).
TABLE 2. Phases and time instants for analyzing the snatch.
T1: Barbell liftoff
T2: First maximum knee exten-
T3: Maximum knee flexion
T4: Peak maximum vertical ve-
locity of the barbell (bar
T5: Second maximum knee ex-
T6: Peak maximum height of the
T7: Instant of the ‘‘catch’’ of the
T8: Instant of the maximum
squat in snatch
T1–T2: First pull
T3–T5: Second pull
this factor shortens the barbell’s lifting trajectory and
subsequently the time of action on it.
The structure of motor patterns has been established
and consists of a number of factors of a variant and in-
variant nature (7). Such factors are dependent on differ-
ent circumstances. Velocity usually behaves as a variant
factor accounting for the dynamics of uniarticular and
pluriarticular movements. As far as weightlifting is con-
cerned, the barbell’s vertical velocity is one of the most
relevant parameters when evaluating the lifting tech-
nique. Yet, its control is conditioned by other determining
parameters such as the amplitude or the trajectory of the
actions involved. In fact, Garhammer (15) already sug-
gested that the timing and the length of the phases could
play important roles as kinematical variables.
Based on such evidences, we intend to verify whether
lifters of different weight categories display different
technical execution patterns when performing the snatch
in situations of maximum effort in competition. The find-
ings could help explain the reasons justifying the attain-
ment of high performance and could also guide coaches
in the technical preparation of their lifters.
We studied a sample consisting of 33 male weightlifters.
The comparative study included 2 groups, taking into ac-
count weight categories. Group A included 17 weightlif-
ters from the lightest categories, 56 and 62 kg; group B
included 16 weightlifters from the heaviest categories, 85
and 105 kg. The 8 best results of each category were an-
alyzed. The lifts analyzed were the heaviest successful
snatches made by selected lifters. Descriptive data on
weight category, body mass, height, and best result of the
subjects are shown in Table 1. As shown, groups A and
B are different in their morphological features, and so
group A lifters, if compared to group B lifters, have dif-
ferent mean values for height and weight (0.23 m and 32
kg lower, respectively).
A 3-dimensional photogrammetry technique was used,
based on 2 synchronized video cameras SVHS, Panasonic
AGDP 800 (50 fields per second; Panasonic, Barcelona,
Spain). The cameras were positioned in front of the plat-
form, on a horizontal plane, approximately 10 m away
from the subjects and with their optical axis at 90?. The
digitizing process was performed by Kinescan Digital 1.1,
from the Institute of Biomechanics of Valencia (IBV; Va-
lencia, Spain); 3-dimensional data were constructed using
the direct linear transformation (DLT) method (1); A cal-
ibration system (3 ? 3 ? 1.5 m) was positioned on the
platform and recorded prior to lifts for each film session.
The analysis covers from barbell lift-off to the instant
of the maximum squat after catching the bar. Conse-
quently, movement was divided into 8 different time in-
stants and 6 phases based on changes in the knee angle
and the position of the barbell (Table 2).
ANALYSIS OF SNATCH IN ELITE MALE JUNIOR WEIGHTLIFTERS
TABLE 3. Time analysis.*
(Categories 56 and 65 kg)
(n ? 17)
(Categories 85 and 105 kg)
(n ? 16)
? (T1–T6)?(T1–T4) (s)
* T1–T2 ? first pull; T2–T3 ? transition; T3–T5 ? second pull; T5–T6 ? turnover; T6–T7 ? catching; T7–T8 ? absorption; ?T1–T6
? from liftoff to maximum bar height; ?T1–T8 ? from liftoff to maximum squat; ?T1–T4 ? from liftoff to peak bar velocity; ?(T–
T6) ? (T1–T4) ? time between instant of peak bar velocity and instant of maximum bar height.
Phase distribution for the snatch (%).
TABLE 4. Intergroup differences in time variables (t-test).*
* T?zmaxbar ? time to maximum bar height; T?vmax ? time to
peak bar velocity; T?1pull ? duration of first pull; T?trans ?
duration of transition; T?turnov ? duration of turnover.
Thirty-seven variables in the following areas were ana-
lyzed: phase timing, kinematics of the bar, and kinemat-
ics of the body.
Descriptive statistics, standard deviation, and coeffi-
cient of variation were used for the statistical treatment
of the data. The ? level to indicate statistical significance
is p ? 0.05. Coefficient of variation values are expressed
in percentages. Additionally, t-test for independent sam-
ples was used to analyze intergroup differences between
weightlifters of different weight categories (groups A and
B). The assumption of the homogeneity of variance was
tested using the Levene test.
Table 3 shows the duration of each lifting phase for the
lifters from each individual group. On average, to lift the
barbell to the highest point (T1–T6) the lifters of group A
take 1.001 seconds and the lifters of group B take 1.089
seconds. For taking the bar to the catch instant (T1–T7)
the lifters in groups A and B take 1.145 and 1.186 sec-
onds, respectively, while the mean time to reach maxi-
mum absorption (T1–T8) is 1.281 and 1.337 seconds.
Moreover, the barbell peak velocity is reached 0.241 and
0.259 second before maximum height for groups A and B,
Even so, the resulting time structure for the snatch
movement (T1–T8) shows a distribution profile in which
the first pull (T1–T2) accounts for 37.7 and 40.1% of the
time total for groups A and B, respectively; transition
(T2–T3) takes up 11 and 9.1%; second pull (T3–T5) 12.3%
for both groups; turnover (T5–T6) 17.1 and 17.2%; catch
(T6–T7) 11.2 and 10.1%; and absorption (T7–T8) 10.6 and
11.3%, respectively, of the total time (Figure 1).
This time pattern is evidence that lifters use a wide
initial phase to impel the bar, which accounts for more
than a third of the total time, whereas the second pull is
the shortest phase. Regarding variability, the most vari-
able phase by far is absorption, followed by transition and
second pull. On the contrary, the least variable phases
are first pull and turnover.
The results obtained by the best-performing lifters in
groups A and B have been included in Table 3. Their val-
ues show time patterns slightly different from their re-
spective group means. The best-performing lifter in group
A takes longer in completing the snatch, for both the
snatch total time (T1–T8) and the different snatch phas-
es, except for transition (T2–T3). On the contrary, the lift-
er with the best results in group B takes shorter in per-
forming the snatch for all execution phases. These lifters’
most relevant differences with respect to the means of
their groups are found in the time up to the barbell’s max-
imum velocity peak (T1–T4). In this case, both lifters take
shorter than their groups (0.74 and 0.72 second, respec-
Regarding group differences (Table 4), t-test showed
differences in the time the bar takes to reach peak veloc-
ity (p ? 0.038), the time the bar takes to reach maximum
height (p ? 0.006), the time used for the first pull (p ?
0.000), the time for transition (p ? 0.020), and the time
for turnover (p ? 0.049). Such differences are evidence
846CAMPOS, POLETAEV, CUESTA ET AL.
TABLE 5. Kinematics of the bar.*
(Categories 56 and 65 kg)
(n ? 17)
(Categories 85 and 105 kg)
(n ? 16)
* Vv ? vertical velocity of the bar; Vv?T2 ? vertical velocity of the bar at instant T2; Vv?trans ? vertical velocity of the bar at instant
T3; Acc?1pull ? acceleration of the bar at first pull; Acc?2pull ? acceleration of the bar at second pull; Traj?bar ? vertical trajectory
of the bar; Trajbar?nor ? normalized vertical trajectory of the bar; Traj?1pull ? vertical trajectory of the bar during the first pull;
Traj?2pull ? vertical trajectory of the bar during the second pull; Lost?zbar ? height of the bar lost from T6 to T7.
TABLE 6. Intergroup differences in kinematical variables (t-
* Vv ? vertical velocity of the bar; Vv?T2 ? vertical velocity of
the bar at T2; Vv?trans ? vertical velocity of the bar at transi-
tion; Acc?1pull ? acceleration of the bar at first pull; Acc?2pull
? acceleration of the bar at second pull.
that group A lifters, if compared to those in group B, take
less time in reaching the barbell’s maximum height, less
time in the first pull, more in the transition, and less in
the turnover. The only phase that did not produce inter-
group differences was that of the second pull (T3–T5).
Kinematics of the Barbell
The barbell reaches its maximum vertical velocity (Vv)
during the second pull, the mean value being 1.7 m·s?1
for group A and 1.78 m·s?1for group B. At instant T2, at
the end of the first pull, the barbell’s vertical velocity is
1.17 m·s?1and 1.26 m·s?1respectively. These data show
that on completion of the first pull—instant T2—the bar-
bell has already reached 68.8 and 70.8% of its top speed
for groups A and B respectively, while during the tran-
sition phase (T2–T3) the barbell loses part of the speed
built up in the first pull. An analysis of the lifters’ indi-
vidual patterns shows evidence that 14 lifters lose from
0.01 to 0.20 m·s?1in the transition phase, 17 maintain
the speed reached at the end of the first pull (T2), and
only 2 lifters manage to increase it.
As to the acceleration of the barbell, the first pull
reaches 3.17 m·s?1for group A and 3.50 m·s?1for group
B, while the second one reaches 5.89 m·s?1and 5.35 m·s?1,
respectively, which proves that lifters apply maximum
power to lift the load on this second pull (T3–T5).
The analysis of the barbell vertical trajectory
(traj?bar) shows that a 0.97-m and 1.08-m maximum tra-
jectory is achieved by groups. Partial trajectories cover
0.34 m and 0.41 m for the first pull (traj?1pull) and 0.24
m and 0.26 m for the second pull (traj?2pull) respectively.
In the catching (T6–T7), the barbell mean height
(lost?zbar) goes down 7 and 6 cm for lifters of group A and
At maximum vertical position (T6), the barbell’s mean
height is 69.2% of the lifters’ height for group A and some-
what higher for group B (70.3%), but the difference is not
As to the values obtained by the best lifters in each
group (Table 5), they seem to follow the same general
trend: if compared to the lifter in group B, the group A
lifter has a lower vertical velocity in absolute terms (1.73
and 1.84 m·s?1) and for the instant corresponding to the
end of the first pull (1.30 and 1.51 m·s?1). In the intra-
group comparison, they both reach values above their
groups’ means for most parameters. However, attention
must be drawn to maximum vertical velocity. Neither of
the lifters have the best results in their groups; in fact 6
lifters in the first group and 3 in the second group attain
better values. On the contrary, as far as the barbell’s ver-
tical velocity at instant T2 is concerned, the winner in
group 1 is second best in his group, and the winner of
group 2 represents the highest value. Consequently, bar-
bell velocity at the end of the first pull is indeed a per-
formance parameter to bear in mind.
The t-test findings (Table 6) show significant differ-
ences between groups A and B with regard to Vv, barbell
maximum vertical acceleration (Acv), its total vertical
trajectory (TrajV), its first pull vertical trajectory for both
its absolute (Traj?1pull) and normalized (traj?norm?1pull)
values, and the height loss in the catch phase (Lost?zbar).
Unlike lifters in group B, group A lifters have less vertical
velocity, less vertical acceleration, a shorter total barbell
trajectory, a shorter first pull normalized trajectory, and
a greater height loss during the catch phase. As far as
the barbell vertical trajectory is concerned, group differ-
ences also remain when the normalized values are cal-
culated in relation with the lifters’ height. In this case,
group A lifters also show a shorter barbell vertical trajec-
ANALYSIS OF SNATCH IN ELITE MALE JUNIOR WEIGHTLIFTERS
TABLE 7. Joint angular kinematics.*
(Categories 56 and 65 kg)
(n ? 17)
(Categories 85 and 105 kg)
(n ? 16)
Vang?Trunk ext (rad·s?1)
* Knee?T2 ? knee joint angle at T2; Knee?T3 ? knee joint angle at T3; Knee?T5 ? knee joint angle at T5; Knee?T7 ? knee joint
angle at T7; Knee?T8 ? knee joint angle at T8; KneeVang?1pull ? knee joint angular velocity at first pull; KneeVang?2pull ? knee
joint angular velocity at second pull; HipVang?1pull ? hip joint angular velocity at first pull; HipVang?2pull ? hip joint angular
velocity at second pull; Vang?Trunk ext ? trunk peak angular velocity on extension.
TABLE 8. Intergroup differences in angular variables (t-
* Knee?T7 ? knee joint angle at T7; Knee?T8 ? knee joint angle
Joint Angular Kinematics
In general terms, all weightlifters apply a countermove-
ment action on the knee joint in the transition and the
second pull, to a greater or lesser extent. The mean values
of the flexion-extension are 140? for the first maximum
knee extension at instant T2 (knee?T2), 124? for the max-
imum knee flexion at instant T3 (knee?T3), and 170? for
the second maximum knee extension at instant T5
(Knee?T5). On the catch instant (T7) the knees are 43?
bent and, on maximum absorption (T8) the bending
reaches 35?. This also confirms that, after the catch, lift-
ers continue absorbing the load by bending their knees
The analysis of angular patterns for hip and knee ex-
tension shows that, for both cases, angular velocity values
are higher in the second pull than in the first one (8.74
versus 4.34 rad·s?1knee, and 8.60 versus 3.78 rad·s?1hip,
respectively). Regarding trunk extension, the angular ve-
locity peak on extension reaches a 4.63 rad·s?1mean val-
As to the values obtained by the best lifters in each
group (Table 7), opposed tendencies are seen. The best-
performing lifter in group A reaches higher knee exten-
sion values on T2, T3, T5, and T7 than the rest of his
group as a whole, while the best-performing athlete in
group B has lower values than those of his group for the
same time instants.
Table 8 shows that significant differences between
groups A and B are found in the knee flexion angle at
instant T7 (p ? 0.024), and on maximum knee flexion at
instant T8 (p ? 0.012). This is to say that group A lifters,
unlike those in group B, bend their knees more both on
the catch (40? versus 45.8?) and on maximum absorption
(32.1? versus 37.5?). It is worth mentioning that there are
no differences between groups in angular parameters
during the first and second pull.
The purpose of this study was to analyze the differences
between weightlifters from different weight categories.
Results show that their execution technical patterns do
differ for some of the analyzed parameters.
In general, the time structure observed by the lifters
is in line with that reported by previous studies. More
specifically, Gourgoulis et al. (18) reported 0.47 second for
the first pull, 0.16 second for the second one, 0.15 second
for the transition, and 0.23 second for turnover. These
times are very close to those in our study, which confirms
the adjustment to a time structure typical of high perfor-
mance (Table 3).
The analysis of the time sequence of the phases re-
veals that the lifters from both groups use different time
patterns, in line with findings by Bauman et al. (3) and
Vorobyev (29). In our study, such differences apply to the
duration of the first pull (T1–T2), the duration of transi-
tion (T2–T3), and the time elapsed until the barbell’s ver-
tical velocity peak is reached (T1–T4), in such a way that
the heaviest lifters in group B use a model based on a
longer initial phase (T1–T2), delaying the barbell’s max-
imum vertical velocity (T4). Therefore, it seems that the
heaviest lifters tend to increase the time for concentric
muscle activity during the first pull and the lightest lift-
ers tend to increase the time for eccentric muscle activity,
with no differences being found in the decisive phase of
the second pull.
There is a basic dynamic principle, namely the fact
that a longer propulsive trajectory allows lifters to act
upon the barbell longer, this resulting in better conditions
to apply force on it. However, not all lifters have the same
body structure. The discussion of the findings is in rela-
tive terms. In this respect, Vorobiev (29) and Bartonietz
(2) pointed out that shorter lifters move the barbell less
than taller ones, which is disadvantageous for driving the
barbell. Such approaches are confirmed in our study.
When vertical trajectory values are normalized by the lift-
ers’ height (Traj?Norm?1pull), significant differences (p ?
0.011) are found in the barbell’s trajectories in the first
848CAMPOS, POLETAEV, CUESTA ET AL.
height (h) during the snatch representing the lifter with best
result in group A (Miculesco, 125 kg).
Typical curves of barbell vertical velocity (Vv) and
pull (T1–T2). Group A lifters, of a lighter weight, draw a
shorter trajectory than those in group B, showing a dif-
ferentiated time pattern in the execution of the initial
phase of the lifting (T1–T2). That is, a phase described by
Garhammer (15) as strength-oriented in which maximum
strength requirements prevail.
With respect to barbell velocity, our study found out
that lifters reach an average vertical velocity which is
similar to that reported in studies of elite lifters ranging
between 1.68 and 1.93 m·s?1(3, 13, 21, 27). In addition,
group B lifters were found to impart greater vertical ve-
locity on the barbell, this being in line with findings by
Baumann et al. (3) and Bartonietz (2).
Different studies have reported barbell velocity to in-
crease continuously between the first and second pulls
(2). Nevertheless, some lifters are able to increase vertical
velocity during the transition phase (T2–T3), while others
are unable to do so due to the countermovement action of
the knees. Out of the 33 lifters studied, 15 lose velocity
in the transition, although such a loss only accounts for
1.7% of the velocity reached until instant T2. Even so,
this value is lower than that reported by Gourgoulis (18)
in his study on Greek elite lifters (2.7%). Therefore, this
is a group of young lifters whose mechanical efficiency
level is already notable.
Even so, by analyzing coincidences between the lifters
from both groups, data shows that, in line with Souza and
Shimada (26), the definitive forces occurred at the end of
the second pull. The timing for moving the barbell seems
to be supported by the following facts: on the one hand,
achieving a high maximum vertical velocity percentage
in the first pull and on the other, imparting maximum
acceleration on the barbell in the second pull. At the end
of the first pull (T2), the barbell reaches 68.3 and 70.8%
of its maximum velocity for groups A and B, while max-
imum acceleration is attained in the second pull (T3–T5)
with a 5.89 m·s?1and 5.35 m·s?1mean values respective-
ly. The use of this similar pattern for barbell acceleration
does not prevent intergroup differences. In fact, signifi-
cant barbell acceleration differences were found in the
second pull, thus proving that lighter lifters attained a
greater explosive component at this decisive snatch point.
As pointed out earlier, a 1.7% velocity loss takes place
during the transition (T2–T3). At the start of the second
pull (T3), the barbell has already reached a velocity that
is higher than two-thirds of its maximum. Figure 2 rep-
resents the vertical velocity and barbell height curves cor-
responding to the lifter with the best result in group A.
We can appreciate a loss of velocity in the transition
phase (T2–T3) and that the peak velocity is reached right
before maximum knee extension (T5). The loss of velocity
appears as an individual pattern, while the instant for
peak velocity is a general behavior, which coincides with
the results obtained by Bauman et al. (3). It is worth men-
tioning the fact that the barbell continues to elevate until
it reaches the highest point, without being affected by the
velocity loss in the transition phase–an indicator of high
ability of the lifters.
The barbell’s height loss (lost?zbar) in the catching
phase (T6–T7) is small. A 6-cm average loss was found
for the whole of the lifters, which is lower than the values
reported by Baumann et al. (3) and Gourgoulis et al. (18)
(10–14 cm and 13.5 cm, respectively). We believe these
differences to be relevant. Please note that for some cases
it means the figures are twofold, which would mean that
efficiency levels are clearly better in the junior category
lifters. Therefore, we think this could be due to the de-
termination of the catch instant. In our case, this is the
instant when the lifter stretches his arms completely. On
the other hand, the use of a 50-Hz sampling frequency
entails the acceptance of reduced error margins, if we
bear in mind that the velocity at which the barbell moves
during that lifting phase is approximately 0.6 m·s?1.
With regard to angular pattern, 2 styles have been
acknowledged for propelling the barbell in the snatch in-
termediate movement (11). On the one hand, the double
knee bend (DKB)—a bouncing action of the knees—and
on the other a style based on hip extension called frog-leg
pull (FLP). Depending on the use of each style, 2 vertical
velocity trajectories can be made: for example, in the DKB
style barbell velocity drops in the transition phase. Ac-
cording to Enoka (8), the bounce action has positive re-
percussions and facilitates the use of the elastic energy
stored on the lifter’s musculature.
The fact that no significant differences were found be-
tween the groups with regard to angular parameters dur-
ing first and second pull confirms the conviction that the
barbell’s kinematical parameter variations are modified
to a greater extent by the influence of the loads than by
that of body movements. As a confirmation of this fact,
none of the angular parameters analyzed was signifi-
cantly related to the barbell’s maximum vertical velocity.
Except for 2 lifters, our findings reveal the lifters use
a DKB variant. In other words, they bend their knees in
the transition phase regardless of their weight category.
This seems to be a generalized behavior in this lifter
group. Figure 3 illustrates the angular behavior corre-
sponding to the lifter with the best result in group B. His
ankle and knee joints display an almost time-parallel
flexion-extension tendency, especially in the transition
and the second pull. In any case, ankle, knee, and hip
reach their maximum extension at the end of the second
pull, as evidence of the contribution of lower segments to
the propulsion of the barbell in a motion similar to that
in the vertical jump, as argued by Garhammer (17) and
Canavan et al. (6). More specifically, the hip reaches max-
imum extension within the 40 milliseconds previous to
instant T5, which is in line with findings by Gourgoulis
et al. (18).
These actions are actually aimed at accelerating the
barbell and pushing it to the vertical to facilitate the
catch phase (T6–T7), which requires a sequential action
based on the coordination of partial impulses. The hip
ANALYSIS OF SNATCH IN ELITE MALE JUNIOR WEIGHTLIFTERS
joint angular displacements of hip, knee, ankle and angle be-
tween trunk and horizontal during the snatch, representing
the lifter with best result in group B (Klokov, 170 kg).
Typical curves of barbell vertical velocity (Vv) and
reaches its maximum extension before the knee and the
ankle do, whereas the trunk reaches maximum extension
during the turnover (T5–T6), right before the barbell
reaches maximum height (T6). Therefore, the sequential
action produced after the first pull is performed in an
orderly manner, starting with hip extension followed by
the extension of the ankle, the knee, and finally the
Finally, the analysis of the results obtained by the
best performers in groups A and B—although not directly
connected to the paper’s focus—help understand some of
the keys to a successful snatch. Based on the results, the
best lifters seem to stand out from the rest basically with
regard to barbell vertical velocity at the end of pull 1. The
remaining parameters studied show execution levels
more in line with the characteristics of individual pat-
terns than with generalizable trends for one or another
In their way to high performance, junior lifters must at-
tain high levels of mechanical efficiency. This paper has
verified this idea with a group of international weightlif-
ters. Despite the differentiated execution pattern imple-
mented by each lifter, in general, junior lifters use a pat-
tern similar to that of elite weightlifters with regard to
both time structure and kinematical and dynamic model.
This pattern is based on 3 major aspects: 1) the rapid
movements of the barbell on the first pull; 2) a small ve-
locity loss during the transition; and 3) a dynamic action
of an explosive nature on the second pull aimed at push-
ing the barbell up towards the vertical.
Furthermore, chronological structure seems to be an
important parameter in describing the lifters’ technical
level and ability to reach maximum efficiency along the
different phases. Among others factors, a successful lift
depends on the skill of the lifter to avoid a velocity loss
on the barbell during the transition. This requires the
implementation and coordination of 2 complementary ac-
tions: the countermovement of the legs during the first
and second pull and the stretching of the trunk. Given
their transcendence, these 2 aspects should be taken into
account in the technical training of young lifters, regard-
less of their body weight category.
Regarding group differences, we can conclude that lift-
ers belonging to heavier categories are more efficient, as
they manage to have longer barbell propulsion trajecto-
ries, which allows them to exert actions on the barbell for
a longer period, especially in the initial lifting phase cor-
responding to a strength-oriented action. This is possibly
one of the reasons accounting for the better situation of
the group when imparting more vertical velocity on the
barbell, as proved. In fact, the goal of the training must
be higher power values, such as using high pulls.
A successful lift seems to be the consequence of a mul-
tifactor basis that materializes individually. We did no-
tice a number of differences in the execution technical
pattern of the snatch of different weight category lifters
but also some similarities, namely imparting velocity on
the barbell especially at the initial lift phase and achiev-
ing optimal time coordination. Therefore, we believe the
intergroup differences found not to be determining
enough so as to assume the existence of a technical model
adapted to the characteristics of each weight category.
Training should be addressed from an open, individual-
ized perspective to help lifters build an efficient individ-
ual technical pattern.
1.ABDEL-AAZIZ, Y.I., AND Y. KARARA, H.M. Direct linear trans-
formation from comparator coordinates into object space coor-
dinates in close range photogrametry. In Symposium on Close
Range Photogrammetry, American Society of Photogrammetry.
Champaign, IL: ASP, 1971. pp. 1–18.
BARTONIETZ, K.E. Biomechanics of the snatch: Toward a higher
training efficiency. Strength Cond. 18:24–31. 1996.
BAUMAN, W., V. GROSS, K. QUADE, P. GALBIERZ, AND A.
SCHWIRZ. The snatch technique of world class weightlifters at
the 1985 World Championships. Int. J. Sport Biomechan. 4:68–
BERNSTEIN, N.A. The Coordination and Regulation of Move-
ment. Baltimore: Pergamon Press, 1967.
CAMPOS, J., AND P. POLETAEV. Biomechanicheskiy analiz in-
dividual’noy tekhniki ryvka tiazheloatletov v uslovijakh otvet-
stvennykh sorevnovaniy. Vestnik sportivnoy nauki, Q. Sovetsky
Sport Moskva 3(5):16–23. 2004.
CANAVAN, P.K., G.E. GARRET, AND L.E. ARMSTRONG. Kinematic
and kinetic relationships between an Olympic-style lift and the
vertical jump. J. Strength Cond. Res. 10:127–130. 1996.
CARTER, M.C., AND D.C. SHAPIRO. Control of sequential move-
ments: Evidence for generalized motor programs. J. Neurophy-
siol. 52:787–796. 1984.
ENOKA, R.M. The pull in Olympic weightlifting. Med. Sci. Sport
ENOKA, R.M. Muscular control of a learned movements: The
speed control system hypothesis. Exp. Brain Res. 51:135–145.
ENOKA, R.M. Load and skill-related changes in segmental con-
tributions to weightlifting movements. Med. Sci. Exerc. 20:178–
GARHAMMER, J. A dynamic rigid link model applied to the
Olympic snatch lift. Master’s thesis, University of California,
GARHAMMER, J. Biomechanical comparison of the U.S. Team
with divisional winners at the 1978 World WeightliftingCham-
pionships. Report to the U.S. National Weightlifting Commit-
GARHAMMER, J. Biomechanical characteristics of the 1978
world weightlifting champions. In Biomechanics VII-B. A. Mo-
recki, K. Fidelus, K. Kedzior, and A. With, eds. Baltimore: Uni-
versity Park Press, 1981. pp. 300–304.
GARHAMMER, J. Biomechanical profiles of Olympic weightlift-
ers. Int. J. Sports Biomechan. 1:122–130. 1985.
850CAMPOS, POLETAEV, CUESTA ET AL.
15.GARHAMMER, J. Weightlifting and training. In Biomechanics of
Sport. C.L. Vaughan, ed. Boca Raton, FL: CRC Press, 1989. pp.
GARHAMMER, J. A comparison of maximum power output be-
tween elite male and female weightlifters in competition. Int.
J. Sport Biomechan. 7:3–11. 1991.
GARHAMMER, J.A. A comparison of propulsive forces for weight-
lifting and vertical jumping. J. Appl. Sport Sci. Res. 6:129–134.
GOURGOULIS, V., N. AGGELOUSSIS, V. KALIVAS, P. TONIOU, AND
G. MAVROMATIS. Snatch lift kinematics and bar energetics in
male adolescent and adult weightlifters. J. Sports Med. Phys.
Fitness 44:126–131. 2004.
GOURGOULIS, V., N. AGGELOUSIS, G. MAVROMATIS, AND A. GAR-
AS. Three-dimensional kinematic analysis of the snatch of elite
Greek weightlifters. J. Sport Sci. 18:643–652. 2000.
GUE, N. Teaching methods for weightlifting exercises. Doctoral
thesis. State Central University of Physical Culture, Moscow,
ISAKA, T., J. OKADA, AND K. FUNATO. Kinematic analysis of the
barbell during the snatch movement of elite Asian weightlift-
ers. J. Appl. Biomech. 12:508–516. 1996.
KAUHANEN, H., K. HA¨KKINEN, AND P.V. KOMI. A biomechanical
analysis of the snatch and clean & jerk techniques of Finnish
elite and district level weightlifters. Scand. Sports. Sci. 6:47–
LUKASHEV, A.A. Analysis of the snatch technique in elite
weightlifters. Doctoral thesis. State Central University of
Physical Culture, Moscow, 1972.
24.ONO, M., M. KUBOTA, AND K. KATO. The analysis of weightlift-
ing movement at 3 kinds of events for weightlifting partici-
pants of the Tokyo Olympic Games. J. Sports Med. 9:263–281.
SCHILLING, B., M. STONE, H. O’BRYANT, A.C. FRY, R. COGLI-
ANESE, AND K. PIERCES. Snatch technique of collegiate national
level weightlifters. J. Strength Cond. Res. 16:551–555. 2002.
SOUZA, A.L., AND S.D. SHIMADA. Biomechanical analysis of the
knee during the power clean. J. Strength Cond. Res. 16:290–
STONE, M.H., H.S. O’BRYANT, F.E. WILLIAMS, R.L. JOHNSON,
AND K.C. PIERCE. Analysis of bar paths during the snatch in
elite male weightlifters. Strength Cond. 20(4):30–38. 1998.
VOROBYEV, A.N. Tiasheloatletichesky sport. Osnovy fisiologii y
spotivnoy trenirovki. (2 Edition). [Weightlifting Sport. Physio-
logical and training foundations]. Moskva. Ed. Fiskultura i
VOROBYEV, A.N. A Textbook on Weightlifting. Budapest: Inter-
national Weightlifting Federation, 1978.
WOLTRING, H.J. A Fortran package for generalized, cross val-
idatory spline smoothing and differentiation. Adv. Eng. Soft-
ware 6:104–113. 1986.
We would like to thank the Spanish Sport Council and the Span-
ish Weightlifting Federation for the additional support provided.
Address correspondence to Dr. Jose ´