Quantitative Analysis of High Goal Polo: A Pilot Study

Abstract

Introduction

Academic literature on Polo players is limited. Polo is a global sport played professionally by males and females, but the dearth of literature remains. We aimed to quantify the characteristics of the winning team of a High Goal Polo tournament, through group and knockout stages (seven games; five wins, two losses).

Methods

Footage from the 2017 Jaeger le Coultre Gold Cup was viewed via Pololine, a dedicated online platform (www.pololine.tv/fullmatches), and analysed using a customised matrix devised in consultation with recreational and professional Polo players. Two raters (R) performed quantitative analysis for each game. Percentage success rates were calculated for forehand and backhand shots of differing lengths and ride-offs (two players contesting for the right to play a shot). Raw differences for melees (coming together of two or more players from each team), penalties awarded (short and long) and turnovers conceded were recorded. Means and 90% Confidence Intervals are shown.

Results

Intra-rater reliability was large to very large (R1 ICC: 0.69 90% C.I.: 0.52 to 0.89. R2 0.72; 0.52 to 0.89), with inter-rater reliability displaying similar values (0.71; 0.62 to 0.78). In games won, fewer turnovers conceded (-3.00; -5.69 to -0.31) and more penalties were awarded (1.43; -0.32 to 3.18) than the opposition. Long penalty (0.86; 0.02 to 1.69) and short penalty (0.57; -0.13 to 1.28) conversion values suggest penalties present advantageous attacking and goal scoring opportunities. Counterintuitively, more melees were won in games lost (4.00; 3.12 to 4.88) than in games won (-0.57; -3.28 to 2.14). Forehand success rates were similar between wins and losses, as were ride-offs. Backhands demonstrated variability in games won (Middle: -1.71; -10.59 to 7.16 and Long: 0.29; -20.28 to 20.85), but were more unsuccessful in lost games (Middle: -6.00; -12.16 to 0.16 and Long: -17.5; -24.09 to -10.91).

Conclusion

We recommend Polo players and teams aim to develop their backhand success rate, whilst conceding fewer penalties than the opposition. The role of melees and ride-offs in the present analysis was unclear. We suggest future research scrutinises these contested actions, especially ride-offs that take place off the ball as these may facilitate attacking play by other team members, much like blocking plays in grid-iron football.

Full text

Quantitative Analysis of High Goal
Polo: A Pilot Study
Russ Best1,2 & Regan Standing2
1School of Social Sciences, Humanities & Law, Teesside University, UK
2WCentre for Sports Science & Human Performance, WINTEC, NZ

Introduction

Methods: Performance Analysis
•Devised a performance matrix
–What would we like to measure?
–What can we measure?
–Define success
•How do you quantify play like
this…?
Polocam Polo (2015)

Action Definition Successful Unsuccessful
Forehand Long >10 horse lengths with elbow flexion Player or player’s team retains
possession.
Possession is lost, including ball
hitting horse/player or out of bounds
Forehand Middle >2 <10 horse lengths As above As above
Backhand Long >10 horse lengths played with elbow
extension As above As above
Backhand Middle >2 <10 horse lengths As above As above
Dribble Possession maintained within
approximately two horse lengths;
two or more contacts
Possession is maintained Possession is lost
Penalty Long An attacking penalty taken on the
halfway line or 60 yards from the
goal
As above or Goal scored As above or Goal is not scored
Penalty Short An attacking penalty taken 30 or 40
yards from the goal As above or Goal scored As above or Goal is not scored
Penalty As action Penalty awarded Penalty conceded
Turnover Unforced errors (missing the ball)
resulting in a change of possession Possession is obtained Possession is lost
Melee Coming together of two or more
horses from each team
Possession is retained ‘in traffic’ or
obtained Possession is lost
Ride off Fair contest for the ball between two
players, in line with the last shot hit
Possession is retained by player or
player’s team Possession is lost; penalty conceded

Methods: Performance Analysis
•1 team; 1 tournament
–(7 games)
•Footage from Pololine.tv
•% success rate and %
difference to opposition
•Penalties, Melees and
Turnovers as raw values
•Intra and Inter-rater reliability
calculated via ICC

Results: Games Won
•More penalties awarded than opposition in 4 out of 5 wins
•Long penalty (1 penalty; 0 to 1.7) and short penalty (0.5; 0 to 1.3)
conversion values suggest penalties present advantageous
attacking and goal scoring opportunities.
•Backhands demonstrated variability in games won (Middle:
-1.71%; -10.59% to 7.16% and Long: 0.29%; -20.28% to 20.85%)
•Fewer turn overs means more of the ball; 100% agreement with
outcome

Results: Games Lost
•Counterintuitively, more melees were won in
games lost (4 melees; 3.1 to 4.9)
•Backhands were more unsuccessful in lost
games
•Middle: -6.00%; -12.16% to 0.16%
•Long: -17.5%; -24.09% to -10.91%

Results: Other observations
•Small % difference in dribble and forehand
middle success rates in majority of games
•No trends were observed in forehand long
success rates but greater % difference than
forehand middle

Results: Other observations
•Ride offs demonstrated similar success rates
between wins and losses
•Reliability is satisfactory
–Intra: R1 ICC: 0.69 90% C.I.: 0.52 to 0.89. Large
–R2 0.72; 0.52 to 0.89 Very Large
–Inter: 0.71; 0.62 to 0.78 Very Large

Discussion/ Practical Implications
•Polo players should aim to develop their backhand
success rate(s)
•Concede fewer penalties than the opposition
•The role of melees and ride-offs in the present
analysis was unclear

Future directions
Case study (under review)
•What does a handicap mean?
•Which actions are associated with
handicap
–Higher
–Lower
•Is there ‘wiggle room’ within a
handicap?
•What’s a horse worth?
Whole tournament
•Do the findings hold up across a
tournament?
•Can we identify styles of play?
•Are there winning and or losing
behaviours?
Translate this into meaningful
actions and work-ons for players

Thanks for listening
@SimplyRussBest Russell.Best@wintec.ac.nz