1. Introduction

Water polo is a sporting situation, the tactical aspect plays a key role. This work, based on video analysis of sports performance, is an attempt to analyze the correct application of the tactics that a coach imparts to his team.

18 matches of the National Women's Water Polo Series A1 Championship were analyzed by a coach assisted by a statistician and a performance analyst.

The purpose of this research was to identify individual events to evaluate the correct application of the schema foreseen in these phases of the game and then to relate conformity to the outcome of the event.

2. Aim

The work objective was to verify if the different attack patterns in numeric superiority, when they were well executed, have led to the expected results in order to create a methodology of training model. The data were collected through the Kinovea system.

The results show a positive and statistically significant correlation coefficient between tactical compliance and outcome of the events.

3. Method

The research approach is integrated and consists of 3 distinct methods:

ü Case study (18 matches of the Italian Serie A1 Women's Championship, season, played by the Volturno s.c.) for the analysis of matches,

ü action research method for coach contribution,

ü and theoretical-argumentative method to deduce a theoretical framework in which define the data processing.

The data collection and analysis was entrusted to performance analysis, carried out with the help of a water polo trainer, statist and performance analyst.

The compliance assessment for tactical models is entrusted to the trainer, based on the comparison analyzed by video analysis of the attack model against the attack pattern actually implemented during the game.

The research team investigated matches with the Kinovea system, isolating single keyframes related to attack events, identifying the attack pattern implemented, so the trainer issued an assessment of the conformance of the attack model (Napolitano, S., Tursi, D. 2013).

It has been elaborated a spreadsheet containing the results for each single event if it has been well-executed and if it has led to the realization of the point or not. The coefficient of correlation between each attack pattern with the result of the event was calculated by elaborating basic descriptive statistics.

In total, 7 attack models were analyzed on 125 events during 18 games. The results obtained are discussed by the research team, conscious of the internal validity allowed by the method of search for action of this type of qualitative analysis.

Materials

Two cameras to film the games

One laptop for analyze end prepare video lessons to show by projector to players during all meetings. Kinovea software was used for select most important part of videos and select the frames.

4. Results

From the evaluation table that was built by combining, for each single event, the evaluation of the trainer on the conformance of the models with the final outcome of the event.

We reported below:

ü Comparison with pattern drawing with Dartfish screenshot of the model implemented during the game;

ü Basic descriptive statistics (presence of single models on total events, presence of "well patterned patterns" on total events, etc ...);

ü Straight linear regression plot for single models;

ü Comparison of the correlation coefficients of individual models. An evaluation table was constructed by combining, for each single event, the evaluation of the coach on the compliance of patterns with the event final outcome.

Table 1

Table 2

Table 3

Graphic 1

Graphic 2

Figure 1. Design of pattern named «schema 1» & Figure 2. Pattern «schema1»Screenshot

Figure 3. Design of pattern named «schema 7» & Figure 4. Pattern «schema7»Screenshot

Graphic 3

Graphic 4

Table 4

Table 5

5. Conclusions

The results obtained demonstrated a general effectiveness of tactical patterns when they are well executed.

There are, however, significant differences in the correlation coefficients of the various applied patterns.

In order to give greater weight to the study of the effectiveness of the proposed patterns, the research should be expanded by evaluating the adversary variable and thus having a more consistent data base in order to establish a direct, apparent and general relationship between the calculated coefficient and the effectiveness of patterns.

The research team is aware of the internal validity of this type of qualitative analysis but can be used by other teams with the necessary modifications.

The results of the analysis surely represent a valuable support for the coach, in order to perfect the team's specific training in the following season.