Ice hockey, like many intermittent sports, is characterized by participants repeatedly performing high-intensity, short-duration bursts of maximal power (3), requiring acceleration and changes of direction (7). A physically demanding contact sport, ice hockey requires participants to have well-rounded physical capabilities to be successful at the elite level, such as anaerobic fitness (36), aerobic endurance (8), and muscular strength and endurance (31). Because of these demands, physical testing of players regularly occurs at the preseason stage of training and often continues throughout the season. The results of these tests could influence a coach’s decision on aspects such as individual playing time for the season (17) and could also be an indicator of game success (21) or future career success (8). The National Hockey League (NHL), played in the United States and Canada, is widely regarded as the best professional league in the world. Each year, the NHL invites young unsigned players to take part in a battery of physical tests called the National Hockey League Entry Draft (NHLED) combine, to assign them a draft round entry number. Professional teams then use these ratings, which are based on test results alongside playing assessments, to offer contracts to players.
The NHLED combine uses a variety of testing protocols, all of which take place off-ice, to assess on-ice playing potential. Some of the tests used in the yearly event include the Wingate 30-second anaerobic power test, a cycle ergometer V[Combining Dot Above]O2max test, muscular endurance tests (push-ups and sit-ups), grip strength, bench press, standing long jump and vertical jumps (8). However, inconsistencies in surrounding literature suggest the combine results may hold little predictive value. Studies that have investigated the results of the combine testing have found little or no direct correlations to individual draft round entry (8,9,34), which leads to questions over the validity of these tests. Therefore, the aims of this study were to review ice hockey testing protocols, considering the reliability of off-ice and on-ice testing protocols for ice hockey players, and to suggest a justified testing battery for the sport.
Off-Ice Testing Measures and Links to Future Success
Several studies have attempted to ascertain whether any of the tests in the NHLED combine can provide worthwhile fitness measures specific to ice hockey, or information to the potential success of players in the future. Vescovi et al. (34) found no correlation with any physical test in the NHLED combine and future career success (draft entry). In a 3-year study between 2001 and 2003 (n = 252), all off-ice elements of the NHLED combine testing protocol were assessed against draft round entry. Maximum 68.2-kg bench press repetitions, maximum push-ups, push strength, pull strength, long jump, vertical jump, maximum curl-ups, sit-and-reach, absolute peak power, relative peak power, fatigue index, and V[Combining Dot Above]O2max all proved to be nonsignificantly correlated to draft round entry. Burr et al. (9) correlated lower-body power against the 2006 draft round entry (n = 95) and found significant moderate correlations for countermovement jump and draft position (r = 0.42) and squat jumps and draft position (r = 0.47). A larger (n = 853) study by Burr et al. (8) concluded that peak anaerobic power, measured with a 30-second Wingate test, was a statistically significant (p = 0.000) variable in a multivariate model for future success in terms of draft entry position; however, the variance explained by this model was <10%. Although the NHLED combine results may produce contrasting data for predicting career success in terms of draft selection, they may predict playing success. Peyer et al. (28) investigated off-ice measures in relation to an ice hockey–specific general performance indictor, termed the plus/minus (+/−) system. This simple scoring system allocates a “plus” to each player on the ice when a goal is scored for the team, or a “minus” if the goal is scored against the team. The running total can be kept over a game or season as a measure of a player’s overall “success”—a higher plus/minus score may indicate the player has been influential in scoring goals and defending the goal. This season long study of National Collegiate Athletic Association (NCAA) Division I players (elite junior level, n = 24) found off-ice tests of repeated sprints (r = −0.6), chin-ups (r = 0.5), bench press (r = 0.5), and leg press (r = 0.6) to be correlated to plus/minus performance. Green et al. (21) conducted a 3-year-long study of the aerobic endurance of players in the NCAA Division I (n = 29), which demonstrated a significant moderate correlation (r = 0.41) between V[Combining Dot Above]O2max and players’ net scoring chances. Green et al. (21) also reported a significant correlation (r = 0.41) between blood lactate at the fourth stage of an incremental treadmill test and playing time. However, Lau et al. (25) found that lower blood lactate levels did not significantly improve the amount of work a player can perform, defined by skating distance.
Relationships Between Off-Ice and On-Ice Testing
The NHLED combine uses off-ice testing protocols, despite the majority of studies that use off-ice performance as a predictor of on-ice future success and performance displaying little to no correlation (8,9,17,21,34). The validity of such tests has been argued because of the lack of specificity to on-ice demands, with some studies suggesting that off-ice tests are of limited use (34).
Durocher et al. (15) showed that off-ice (cycle ergometry) measures of V[Combining Dot Above]O2max and lactate threshold were not accurate predictors of the same measures produced from an on-ice graded skating test. They reported no correlation of V[Combining Dot Above]O2max between on-ice and off-ice measures and found that off-ice measures of V[Combining Dot Above]O2max significantly underpredicted on-ice recorded values by 3.3 ml·kg−1·min−1. On-ice V[Combining Dot Above]O2max values closely matched those of a previous study by Durocher et al. (16). Durocher et al. (15) also showed that lactate threshold was significantly higher (p = 0.01) on-ice than off-ice. When expressed as a percentage of V[Combining Dot Above]O2max, lactate threshold occurred on-ice at 86%, 15% higher than off-ice. Comtois et al. (11) also found that selected off-ice measures could not predict on-ice performance. They reported nonsignificant relationships between an on-ice sprint (40 m) and off-ice measures of power, using broad jump and vertical jump protocols. Similarly, Gilenstam et al. (20) also reported no significant correlations between an on-ice sprint (47.85 m) and off-ice measures of peak torque of the quadriceps, using an isokinetic dynamometer, and V[Combining Dot Above]O2max relative to body weight, for men or women. Supporting this, Bracko and George (6) reported no correlation between an on-ice sprint (44.80 m) and off-ice power, measured with vertical jump height. Finally, Behm et al. (2) reported that off-ice power and strength tests could not predict on-ice performance. They found that maximum skating speed (measured as a sprint with a flying start between the marked blue lines) was not significantly correlated to drop jump index, 1RM leg press, or squat jump height.
However, Potteiger et al. (29) found that on-ice sprint (54 m) time could be accurately predicted by peak power, derived from a Wingate test using a cycle ergometer (r = −0.43, p < 0.05). Additionally, Farlinger et al. (18) reported significant (p < 0.01) correlations between on-ice sprints (30 m) and off-ice measures of broad jump (r = −0.74), vertical jump (r = −0.71), and relative absolute and mean Wingate peak power (r = −0.71 and −0.73). It is worth noting that the testing protocols for skating speed vary across studies, which may influence reported correlations, as reliability coefficients vary between tests, from 0.6 (6) to as high as 0.99 (4).
Off-Ice Tests With Biomechanical Similarities to On-Ice Tests
Several studies have criticized off-ice testing protocols, suggesting many of the testing methods do not match the biomechanical movements of on-ice tests (2). Other studies have investigated off-ice tests that share biomechanical similarities to their on-ice equivalents.
One such test is for off-ice sprint performance, which has demonstrated high correlation to on-ice sprint performance, and maximum skating speed. Behm et al. (2) reported a significant (r = 0.51, p < 0.05) correlation between off-ice 40-yd sprint and maximum on-ice skating speed in male players, and Bracko and George (6) reported similar findings using a 40-yd sprint with female players (r = 0.72, p < 0.05). Farlinger et al. (18) reported similar findings between 30-m off-ice and 35-m on-ice sprints, reporting a correlation of r = 0.78 (p < 0.01).
Buchheit et al. (7) investigated a novel on-ice intermittent endurance test, modified from a previously valid and reliable off-ice test. The “30-15 Intermittent Ice Test” consists of 30-second shuttle skates of 40 m, interspersed with 15-second passive recovery periods. Participants are required to keep up with an audible signal and are classed as finished when they fail to maintain the required skating speed; that is, they do not make the required distance by the audio signal on 3 consecutive occasions. They found the new test to be highly reliable, with an intraclass correlation coefficient of 0.96, and found high correlations between maximal velocity (r = 0.72) and maximum heart rate (r = 0.61) when compared with the off-ice testing protocol.
However, not all studies have reported meaningful correlations between biomechanically similar off-ice and on-ice tests. Koepp and Janot (24) found that a running treadmill protocol elicited a significantly higher V[Combining Dot Above]O2max than a skating treadmill protocol. The running protocol reported a mean V[Combining Dot Above]O2max value of over 6 ml·kg−1·min−1 greater than the skating protocol (p < 0.05). Additionally, Knous et al. (23) reported similar results, finding on-ice measures of V[Combining Dot Above]O2max to be lower than off-ice treadmill measures (p < 0.05). However, Knous et al. (23) reported lactate responses similar to those reported in Durocher et al. (15), with lactate threshold occurring at 94% of V[Combining Dot Above]O2max on-ice, 20% higher than off-ice.
There are several reasons suggested for the variety in the data. Off-ice tests may not recruit the same muscle mass as skating (15), and although jumping protocols elicit the same underlying power production as skating, the movements are biomechanically different (26). Similarly, discrepancies exist with cycle testing. Although some studies report significant correlations between maximum skating speed and peak power derived from Wingate testing (18,29), other studies have found the 2 variables to not show correlation (15), suggesting that this mode of testing has yet to be fully validated. Especially at top speed, skating requires horizontal leg power, which is more easily replicated with a sprinting motion, than vertical leg power, which is replicated with jump protocols (18). Also, skating mechanics rely more on impulse than stretch-shortening cycle (2), so testing protocols such as countermovement jumps and drop jumps may lack specificity. Variances in the V[Combining Dot Above]O2max data may be explained by different methodological approaches. Although some studies (15,16) tested their participants on a normal ice rink, others used a specially designed ice treadmill (24). Although considered biomechanically similar, skating on an ice treadmill forced subjects to take a lowered stance, possibly restricting the aerobic potential of working muscles (24). An extensive summary of ice hockey tests and related outcomes can be found in Table 1.
Further Limitations in Testing Protocol
Ice hockey is characterized by high-intensity, short-duration sprints (3). With the established demand on both the aerobic and the anaerobic energy systems that intermittent sports require, there has been a recent increase in using repeated-sprint ability (RSA) tests for assessing the physical capabilities of intermittent sport athletes. The specificity of “traditional” anaerobic power measures (e.g., the Wingate anaerobic test) has been questioned for intermittent sport athletes (1) and in particular ice skaters (13), and as many of these tests require specific laboratory equipment, the number of teams able to perform these tests is limited. Therefore, RSA tests have gained popularity with athletic trainers (27). Several versions of on-ice repeated-sprint tests have been tested in current literature. The Reed Repeat Sprint Skate (RRS) test consists of players performing six 91-m sprints with rest periods of 30 seconds. Although Reed et al. (32) have reported fairly high test-retest correlations (r = 0.78) for the RRS test, it has been criticized for being too exhausting, often leading to increased variability in response scores as a result of players being unable to perform to their maximum ability (30). Carey et al. (10) used an RSA test of 5 laps of an ice pad, with 30-second rest periods. However, this repeated-sprint method was not measured for reliability as the test was only performed once and methodological problems are presented in the study, including the use of a stopwatch for timing and researchers judging when players crossed the finish line; this decreases the usefulness of this test. Power et al. (30) recently reported test-retest correlations of a new type of repeated-sprint test, named the Repeat Ice Skating Test. This test consists of participants completing 3 half laps of an ice pad, with a 10-second rest period in between. Total reported distance of the skate was 49 m. Average peak power was derived with a formula from previous RSA literature (35), and an intraclass correlation coefficient value was reported at 0.99, suggesting very high reliability for the test. However, several methodological issues exist with the test. Although total distance for each skate was defined at 49 m, the skating path was not determined in the article. Indeed, the skate path chosen by the individual player was highlighted as an issue, which could affect the test. Also, the distance recorded was a measurement from the halfway point (the red line) of an ice pad, around the goal, and back to the red line. Although this may be 49 m at the rink where the study was conducted, ice pads are not required to be one standard size, and therefore, caution is advised if using the test data for comparative studies.
Despite studies suggesting the importance of on-ice agility (2,4,6), it has rarely been tested in ice hockey players. Many studies have focused on sprint times and V[Combining Dot Above]O2max; however, studies (8,9,21,34) have suggested that these test protocols may not be appropriate measures to predict success. In the past, the NHLED combine previously used the hexagon agility test, but this has subsequently been dropped from their testing battery. Farlinger et al. (18) support this decision, as they found nonsignificant correlation of on-ice agility with the hexagon test. Several studies (5,6,14) have reported high correlations of on-ice acceleration and jump height, and others have reported similarly high correlations of on-ice acceleration to overall speed (6,14); however, acceleration and speed are not the only components of agility. According to Turner et al. (33), agility also includes the ability to change direction quickly, without losing balance, and incorporates strength and power. Cooke et al. (12) stressed the importance of another component—namely, deceleration—on the ability to change direction quickly. It can be concluded therefore that agility tests require both physical ability and skill (20), and so it is reasonable to suggest that agility should be tested specifically, as current off-ice tests have been found to not correlate with on-ice agility (18). One method of testing on-ice agility regularly used is the cornering S test, first introduced by Greer et al. (22). This test requires participants to skate an S-shaped course at full speed, requiring acceleration, deceleration, changes of direction, balance, crossover, steps, and lateral push-off components. Studies (18,22) have reported a high test-retest reliability (r = 0.95 and 0.96) of this test, using a variety of athletes playing at a high standard; however, this test is not used in current NHL combine testing protocols because the omittance of any on-ice tests.
Proposed Testing Battery
With such equivocal testing protocols, questions remain over the most appropriate testing battery for ice hockey players. It is the opinion of the authors that an appropriate testing protocol should include on-ice and off-ice tests, and should measure acceleration, speed, anaerobic and aerobic power, upper-body and lower-body strength, change-of-direction ability, and body composition. Acceleration, speed, and change-of-direction tests have been included to assess the skating ability of the athletes. Upper-body and lower-body strength tests have been included as Ransdell and Murray (31) suggest they are linked with the likelihood of injury, given the physical nature of ice hockey. Aerobic power has been included and tested using the 30-15 Intermittent Ice Test (7). This test requires players to skate for repeated 30-second bouts with 15 seconds of active recovery in between. The players must keep up with an audio signal, and the speed at which the players skate increases each stage. Aerobic conditioning is considered an important fitness component by coaches, who may choose to base playing time on how “fit” they consider a player to be. The protocol described avoids retesting similar components (e.g., off-ice sprints and on-ice sprints), which will avoid overtesting the athletes and will reduce overall testing time. The chosen tests have been used in previous ice hockey studies and can be considered highly reliable (Table 2). Finally, the tests chosen in the battery have been chosen to be as practical as possible. The tests require little or inexpensive equipment, and because of their short duration, it is realistic to test a whole team over 2 sessions. To be as specific and applicable as possible, we feel that acceleration, speed, aerobic power, and change-of-direction ability should be tested with on-ice tests, whereas anaerobic power, anthropometry, and upper-body and lower-body strength should be tested using off-ice protocols.
An attempt has been made to highlight important limitations of current ice hockey fitness testing protocols, by investigating the reliability of current tests in relation to their on-ice counterparts, or their ability to predict future success. With current testing protocols showing limited or equivocal practical results, strength and conditioning coaches should ensure that they are fully aware of the strengths and weakness of both on-ice and off-ice testing when working with a team. It is suggested that ice hockey testing protocol should give consideration to using more specific testing measures, testing players on the ice in full kit. Finally, a testing battery has been suggested that covers the major fitness components of ice hockey, with consideration to the budget and expertise of the majority of hockey teams.
The majority of studies in this review suggest that off-ice testing has limited use when testing ice hockey players. Because many of the traditional tests used by coaches are performed in an off-ice setting, it is suggested that new testing protocols be considered, and these should include testing players with on-ice tests and off-ice to allow for the most applicable results. The concept of “fitness” for ice hockey is difficult to define because of the complex demands of the game, and the range of new fitness tests presented (Table 2) attempts to cover all the elements involved in ice hockey. Also, it has been designed to work in conjunction with regular on-ice and off-ice training to assess initial fitness levels and monitor any changes over the season. Similar studies to those reviewed could be conducted to investigate the relationship between on-ice measures and future success. Also, there is scope for the development and testing of reliable on-ice repeated sprint tests and agility tests.
1. Aziz A, Chuan T. Correlation between tests of running repeated sprint ability and anaerobic capacity by Wingate cycling in multi-sprint sports athletes. Int J Appl Sports Sci 16: 14–22, 2004.
2. Behm DG, Wahl MJ, Button DC, Power KE, Anderson KG. Relationship between hockey skating speed and selected performance
measures. J Strength Cond Res 19: 326–331, 2005.
3. Bishop D, Lawrence S, Spencer M. Predictors of repeated-sprint ability in elite female hockey players. J Sci Med Sport 6: 199–209, 2003.
4. Bracko MR. On-ice performance
characteristics of elite and non-elite women’s ice hockey players. J Strength Cond Res 15: 42–47, 2001.
5. Bracko MR, Fellingham GW. Prediction of ice skating performance
with off-ice testing in youth hockey players. Med Sci Sports Exerc 29: S172, 1997.
6. Bracko MR, George JD. Prediction of ice skating performance
with off-ice testing in women’s ice hockey players. J Strength Cond Res 15: 116–122, 2001.
7. Buchheit M, Lefebvre B, Laursen P, Ahmaidi S. Reliability, usefulness, and validity of the 30-15 intermittent ice test in young elite ice hockey players. J Strength Cond Res 25: 1457–1464, 2011.
8. Burr JF, Jamnik RK, Baker J, Macpherson A, Gledhill N, McGuire EJ. Relationship of physical fitness test results and hockey playing potential in elite-level ice hockey players. J Strength Cond Res 22: 1535–1543, 2008.
9. Burr JF, Jamnik RK, Dogra S, Gledhill N. Evaluation of jump protocols to assess leg power and predict hockey playing potential. J Strength Cond Res 21: 1139–1145, 2007.
10. Carey D, Drake M, Pliego G, Raymond R. Do hockey players need aerobic fitness? Relation between VO2max and fatigue during high-intensity intermittent ice skating. J Strength Cond Res 21: 963–966, 2007.
11. Comtois AS, Savage J, Briand M, Whittom F, McCrossim J, Sercia P, Leone M. Repeat ice-skating sprints in relation to off-ice tests. Med Sci Sports Exerc 43: S143, 1998.
12. Cooke K, Quinn A, Sibte N. Testing speed and agility in elite tennis players. Strength Cond J 33: 69–72, 2011.
13. De Boer RW, Van Ingen Schenau GJ, De Groot G. Specificity in training in speed skating. In: Physiological, Biomechanical and Technical Aspects of Speed Skating. Rispens P., Lamberts R., eds. Groningen: Private Press, 1985. pp. 39–48.
14. Diakoumis K, Bracko MR. Prediction of skating performance
with off-ice testing in deaf ice hockey players. Med Sci Sports Exerc 30: S272, 1998.
15. Durocher JJ, Guisfredi AJ, Leetun DT, Carter JR. Comparison of on-ice and off-ice graded exercise testing in collegiate hockey players. Appl Physiol Nutr Metab 35: 35–39, 2010.
16. Durocher JJ, Jensen DD, Arredondo AG, Leetun DT, Carter JR. Gender differences in hockey players during on-ice graded exercise. J Strength Cond Res 22: 1327–1331, 2008.
17. Durocher JJ, Leetun DT, Carter JR. Sport-specific assessment of lactate threshold and aerobic capacity throughout a collegiate hockey season. Appl Physiol Nutr Metab 33: 1165–1171, 2008.
18. Farlinger CM, Kruisselbrink LD, Fowles JR. Relationships to skating performance
in competitive hockey players. J Strength Cond Res 21: 915–922, 2007.
19. Geithner CA, Lee AM, Bracko MR. Physical and performance
differences among forwards, defensemen, and goalies in elite women’s ice hockey. J Strength Cond Res 20: 500–505, 2006.
20. Gilenstam KM, Thorsen K, Henriksson-Larsen KB. Physiological correlates of skating performance
in women’s and men’s ice hockey. J Strength Cond Res 25: 2133–2142, 2011.
21. Green MR, Pivarnik JM, Carrier DP, Womack CJ. Relationship between physiological profiles and on-ice performance
of a National Collegiate Athletic Association Division I hockey team. J Strength Cond Res 20: 43–46, 2006.
22. Greer N, Serfass R, Picconatto W, Blatherwick J. The effects of a hockey-specific training program on performance
of Bantam players. Can J Sport Sci 17: 65–69, 1992.
23. Knous J, Pivarnik JM, Womack CJ, Carrier DP. Comparison between on and off ice physiological testing of collegiate hockey players. Med Sci Sports Exerc 37: S83, 2005.
24. Koepp KK, Janot JM. A comparison of VO2max and metabolic variables between treadmill running and treadmill skating. J Strength Cond Res 22: 497–502, 2008.
25. Lau S, Berg K, Latin RW, Noble J. Comparison of active and passive recovery of blood lactate and subsequent performance
of repeated work bouts in ice hockey players. J Strength Cond Res 15: 367–371, 2001.
26. Matthews MJ, Comfort P, Crebin R. Complex training in ice hockey: The effects of a heavy resisted sprint on subsequent ice hockey sprint performance
. J Strength Cond Res 24: 2883–2887, 2010.
27. Meckel Y, Machnai O, Eliakim A. Relationship among repeated sprint tests, aerobic fitness, and anaerobic fitness in elite adolescent soccer players. J Strength Cond Res 23: 163–169, 2009.
28. Peyer KL, Pivarnik JM, Eisenmann JC, Vorkapich M. Physiological characteristics of National Collegiate Athletic Association Division I ice hockey players and their relation to game performance
. J Strength Cond Res 25: 1183–1192, 2011.
29. Potteiger JA, Smith DL, Maier ML, Foster TS. Relationship between body composition, leg strength, anaerobic power, and on-ice skating performance
in division I men's ice hockey athletes. J Strength Cond Res 24: 1755–1762, 2010.
30. Power A, Faught BE, Przysucha E, McPherson M, Montelpare W. Establishing the test-retest reliability and concurrent validity for the Repeat Ice Skating Test (RIST) in adolescent male ice hockey players. Meas Phys Educ Exerc Sci 16: 69–80, 2012.
31. Ransdell LB, Murray T. A physical profile of elite female ice hockey players from the USA. J Strength Cond Res 25: 2358–2363, 2011.
32. Reed A, Hansen H, Cotton C, Gauthier R, Jette M, Thoden J, Wenger H. Development and validation of an on-ice hockey fitness test [abstract]. Can J Appl Sport Sci 5: 245, 1980.
33. Turner A, Walker S, Stembridge M, Coneyworth P, Reed G, Birdsey L, Barter P, Moody J. A testing battery for the assessment of fitness in soccer players. Strength Cond J 33: 29–39, 2011.
34. Vescovi JD, Murray TM, Fiala KA, Van Heest JL. Off-ice performance
and draft status of elite ice hockey players. Int J Sports Physiol Perform 1: 207–221, 2006.
35. Watson RC, Sargeant TL. Laboratory and on-ice test comparisons of anaerobic power of ice hockey players. Can J Appl Sport Sci 11: 218–224, 1986.
36. Wilson K, Snydmiller G, Game A, Quinney A, Bell G. The development and reliability of a repeated anaerobic cycling test in female ice hockey players. J Strength Cond Res 24: 580–584, 2010.
Keywords:Copyright © 2013 by the National Strength & Conditioning Association.
performance; measurement; intermittent sport