Overarm throwing seems to play an important role in many team sports, for example handball, water polo and of course canoe polo. One study examining an international canoe polo team found players to pass the ball 45 +/- 22 times during the first half and 23 +/- 10 times during the second half. Shots taken were 4 +/- 2 during the first half and 4 +/- 1 during the second half (Forbes et al., 2013). This was with ball possession for 50% of the time, which roughly translates to 7.6 throws per minute. Analysis of my own team revealed an average of 12 shots at goal per game during a whole tournament (in a later post more on this).
There are several components which make a pass or a shot successful. In my opinion the main ones are accuracy, distance, a feint/deceptiveness and velocity. Of course there are many more components, like game awareness, technique, strength, body size, etcetera, but these result in the aforementioned components. In this blog post I will focus on the velocity component, and specifically which training methods may be effective in improving it.
As far as I know no study has shown the importance of throwing velocity, although intuitively it feels important. In theory having a higher throwing velocity would contribute to teams playing at a higher tempo, having greater passing distance and making it more difficult for the goalkeepers plus defenders to stop a shot at goal. From unpublished research in elite teams we know that senior men throw faster than U21 men and that for both these teams playing squad players throw faster than the extended playing squad players. It has to be noted that a similar pattern is present for lean body mass, with the senior men playing squad players having the highest lean body mass and the U21 extended playing squad players having the lowest. This influences throwing velocity as well (van den Tillaar & Ettema, 2004).
Therefore it seems beneficial to try and improve throwing velocity. This could be done by improving technique, but I will focus on the physical training programs.
For this I performed a meta-analysis. In this method you combine studies to get to a more robust conclusion of the actual effect you want to examine and express it in a standardized value so you can compare methods despite different tests/outcome measures/number of participants. For example, it would be possible to compare the effectiveness of a training method for elite sprinters and recreational sprinters. Elite sprinters would probably improve far less than the recreational sprinters, but as their performances differ very little the smallest improvement has a relatively large effect. Similarly, the recreational sprinters would probably improve more, but because their times and the improvements will vary a lot the relative effect will not be as big. Going from 10.2 to 9.8 in a 100-meter race with the competition running between 9.8-10.4 transforms you from a sub-top sprinter to a potential world champion, while going from 13.2 to 12.8 in a field that ranges from 11.6-14.0 hardly makes any difference performance wise.
I'll keep the technical stuff relatively short. I performed a systematic search using an electronic database (Pubmed) with results up to June 2nd 2016. Key words used were “throw AND training”. Articles were included if they had throwing velocity as an outcome measure, used subjects from throwing sports between 18 and 40 years old, were intended to improve throwing velocity and the training program lasted more than three weeks. This eventually resulted in 22 included articles. Interventions were categorized on training method. If the training program used multiple methods these were categorized under a combination category. If they could not really be categorized they were categorized under "other". To compare interventions, effect sizes were calculated between pre- and post tests using Hedges and Olkin’s formula. To determine the effect size of a category the weighted average was calculated using Hedges’s g. Interventions were excluded if the 95% confidence interval of the ES fell outside of weighted average’s 95% confidence interval of the category. If categories existed of one intervention they were put together with their most similar category. 39 interventions were included in the final analysis. See figure 1. An excel file with the exact numbers can be found here.
An explanation of the categories with the number of interventions involved:
- "Throwing (n=5)": a program using throwing exercises with a regular to the sport specific ball. 2.8 training sessions per week for 8.9 weeks, total 26.4 sessions.
- "Weighted throwing (n=4)": a program using throwing exercises using light/regular/heavy balls. 3 sessions per week for 10 weeks, total 30 sessions.
- "Rotator cuff+traditional resistance (n=2)": a program specifically targeting the rotator cuff muscles, complemented by traditional resistance training (6-15 repetitions, no continuous acceleration). 2 sessions per week for 8 weeks, total 16 sessions.
- "Traditional Resistance+plyometric (n=2)": a program using traditional resistance (6-15 repetitions, no continuous acceleration) and plyometric exercises (e.g. clap push-ups, drop jumps). 2 sessions per week for 8 weeks, total 16 sessions.
- "Traditional resistance(+throw) (n=2)": a program using either traditional resistance (6-15 repetitions, no continuous acceleration) or traditional resistance supplemented by regular throwing exercises. 3 sessions per week for 12 weeks, total 36 sessions.
- "Explosive strength (+throw) (n=3)": a program using either explosive resistance exercises (olympic lifts/jump squats/bench throw/etcetera, low percentage of 1RM for all except olympic lifts) or explosive resistance exercises complemented by regular throwing exercises. 1.8 sessions per week for 8.9 weeks, total 16 sessions.
- "Maximal resistance (n=7)": a program using maximal strength exercises with no continuous acceleration (weight <=6 repetitions maximum). 3 sessions per week for 8.9 weeks, 25 sessions in total.
- "Maximal resistance+throw (n=3)": a program using maximal strength exercises with no continuous acceleration (weight <=6 repetitions maximum) complemented with regular throwing exercises. 2.8 sessions per week for 7.5 weeks, total 21.7 sessions.
- "plyometric (n=3)": a program using plyometric exercises (e.g. clap push-ups, drop jumps). 1.6 sessions per week for 6.7 weeks, total 17.3 sessions.
- "plyometric+throw (n=2)": a program using plyometric exercises (e.g. clap push-ups, drop jumps) complemented with regular throwing exercises. 2 sessions per week for 8 weeks, total 16 sessions.
- "Core training (n=3)": A program using exercises that teach to stabilize the core/one's posture while resisting the force generated in the exercise. 2.2 sessions per week for 7.5 weeks, total 14.9 sessions.
- "Multiple (n=3)": A program involving plyometric, explosive resistance, traditional resistance and resistance increased sport specific exercises. 3 sessions per week for 6 weeks, 18 sessions in total.
- "AllInt (n=39)": All interventions combined.
From the results it seems that complementing a maximal resistance program with throwing exercises seems a lot more effective than only doing the maximal resistance program. A regular throwing program does not seem to improve performance in already well trained individuals.
A great website on how to interpret effect sizes like the ones I used can be found here.
In conclusion, if you are an experienced throwing athlete you should do more than throwing if you want to improve your throwing velocity. The most effective methods for this are weighted throwing, traditional resistance combined with plyometric training, explosive strength training possibly complemented with throwing exercises, and doing a combination of three or more of the above mentioned methods.
You can find all references used for the meta-analysis with their effect sizes per category in this document.
Other references used
FORBES SC, KENNEDY MD and BELL GJ, 2013. Time-motion analysis, heart rate, and physiological characteristics of international canoe polo athletes. Journal of strength and conditioning research / National Strength & Conditioning Association, 27(10), pp. 2816-22.
VAN DEN TILLAAR, R. and ETTEMA, G., 2004. Effect of body size and gender in overarm throwing performance. European journal of applied physiology, 91(4), pp. 413-418.