A Vision for Performance Analysis

  • 919 views
Uploaded on

Keynote Address at IACSS09

Keynote Address at IACSS09

More in: Sports , Technology
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Be the first to comment
No Downloads

Views

Total Views
919
On Slideshare
0
From Embeds
0
Number of Embeds
0

Actions

Shares
Downloads
0
Comments
0
Likes
2

Embeds 0

No embeds

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
    No notes for slide

Transcript

  • 1. A Vision for Performance Analysis in Elite Sports Prof. Dr. Martin Lames Institute for Sports Science, University of Augsburg IACSS09, Canberra, Australia September, 23 rd – 26 th 2009
  • 2.
    • Framework
    • Vision I: Better theories / models
    • Vision II: More support for practice
    • Vision III: Improved technologies
    • Final Remarks
    Program
  • 3. Framework Performance Analysis
  • 4. Tasks of performance analysis
    • Theoretical performance analysis
      • Explains structure of sports performances by general laws, e.g.
        • Relation between performance and performance indicators (parts of performance, skills & abilities)
        • Search for new models to explain internal functioning of sports performances
    • Practical performance analysis
      • Direct support for training and coaching, e.g.
        • Performance analysis in training and competition
        • Implementation of new technologies in practice
    Framework
  • 5. Why distinguish?
    • Theoretical and practical performance analysis are carefully to be distinguished
      • Require different research strategies, samples, designs, methods, approaches
      • Are both necessary to provide full scientific support
      • Depend on each other:
        • Practical PA is the most valuable source for hypotheses
        • Theoretical PA improves the impact of scientific support
    Framework
  • 6. Role of Computer Science
    • Theoretical performance analysis
      • Model-building and simulation
      • Knowledge generation by AI-Methods
    • Practical performance analysis
      • Technological support for performance analysis
      • Knowledge how to implement new tools in practice
    Framework
  • 7. Life cycle of technological innovations
      • Demonstrate potential for sport
    Framework
      • Invention
      • Routine procedure
      • Pilot implementation in practice
    Computer Science
      • Sport
  • 8. Vision I: Better Theories Problems
  • 9. Complexity and dynamics I Better theories Abilties and skills Perfor- mance Training
  • 10. Game Sports Better theories
  • 11. Complexity and dynamics II Better theories Cognition Motor Response Perception
  • 12. Problems we face
    • Nature of our subject:
      • Complex & dynamic
      • Non-linear phenomena
      • Games: Interaction
    • Need for adequate theories!
    Better theories
  • 13. Vision I: Better Theories Example: Relative Phase
  • 14. Relative phase
    • Players in net/wall games perform “cycles” from stroke to stroke
    • Relative phase describes the relation between those cycles, e.g. in-phase or anti-phase, it is thus a model of (positional) interaction during a rally
    • Hypotheses: Stable phases of rally = stable relative phase Instable phases of rally = instable relative phase
      • Stable: Normal game, no pressure
      • Instable: Create pressure, go for a winner, force an error, perturbation
    Better Theories
  • 15. Relative phase Better Theories
  • 16. Relative phase Better Theories
  • 17. Relative phase
    • In invasion games relative phase measures the degree of coupling of
      • teams
      • sub-units of a team (e.g. dyads)
      • player and opponent
    • It characterizes an important aspect of tactical behavior
    Better Theories
  • 18. Team’s centers in x-direction Relative Phase Football Better Theories
  • 19. Team’s centers in x-direction, time window Relative Phase Football Better Theories
  • 20. Summary Relative Phase
    • Research:
      • Squash: McGarry et al. (1999); McGarry & Walter (2007)
      • Tennis: Palut & Zanone (2005); Lames & Walter (2006)
      • Basketball: McGarry (subm.)
      • Football: Lames, Ertmer & Walter (2008); Lemmink & Frenken (subm.)
    • Result:
    • Relative Phase models
      • Positional interaction of players in net/wall games
      • Coupling of teams, parts of teams and dyads
    • Perspectives:
      • Perturbation analysis
      • New approach for performance analysis
      • System dynamic modelling of sports
    Better Theories
  • 21. Vision I: Better Theories Vision
  • 22. Vision
    • Greater awareness of problem
      • Complexity, dynamics, interaction are widely perceived as phenomena to be delt with
      • Insight in deficiency of linear models
    • New approaches
      • Dynamical systems theory
      • Game Sports: Field theory
    • Challenges
      • Not just fancy applications of new tools, but creation of substantial knowledge
      • Not just „fa çon de parler”, but substantial underpinnings
    Better Theories
  • 23. Recurrence analysis in Football Better Theories 00:00 15:00 30:00 45:00 00:00 15:00 30:00 45:00
  • 24. Vision II: More support for practice Qualitative Game Analysis
  • 25. Coupling of training and competition Description Training Analysis Transfer to training Competition More Support
  • 26. Steps of Qualitative Game Analysis More Support Transfer in training Video training Peer-debriefing Qualitative main analysis (Thick Description) Peer-briefing Competition Recording/ Realtime diagnostics
  • 27. Concept of social intervention Social intervention More Support Theory Practice Coach & athletes Scientific intervention Akt 1 Akt 2 Akt 3 Akt 4 Akt 5
  • 28. Practical experiences
    • Beach volleyball:
      • Sydney 2000
    • Handball:
      • U19 WC 2007
      • U17 EC 2008, WC 2009
      • German 1st league & 3rd league
    • Football:
      • German 1st league & 2nd league
      • Women: German Champion
      • Personal Tactics Advising
    More Support
  • 29. Vision II: More support for practice Tactical video training
  • 30. Data base
    • German national handball youth team (16-18)
    • Intervention time 2006-2009
    • Competitions and training camps covered: 27 in last three years
    • Highlights:
      • European Championships 2006  9th place
      • World Championships 2007  2nd place
      • European Championships 2008  Champion
      • World Championships 2009  5th place
    More Support
  • 31. Theoretical framework More Support Model of mass communication (Merten, 1994) M 1 M 2 M 3 … Messages, game philosophy, feedback Recipient Sender M x I 1 I 2 I 3 … Internal context Pre-knowledge, motivation I x E 1 E 2 E 3 … External context Home club game situation E x Information triple
  • 32. Variants of video training
    • Social configuration
      • team, small groups, single players
    • Teaching concept
      • cognitive-instructive, self-organised, mixture
    • Presentation concepts
      • e.g. selection of positive vs. negative scenes, slow-motion
    • Sequential study with succession of variants, qualitative assessment of effectiveness
    More Support
  • 33. Effectiveness of video training More Support
    • Assessment of effectiveness
      • communicative validation with stakeholders
      • strategy-tactics-comparison with external experts
      • qualitative interviews with players and coaches
      • assessment of recall and recognition (video-testing)
  • 34. Applications of video training
    • Video training with teams
    • Video training with groups of players
    • Individual video training
    • Training interventions
    • Motivation enhancing videos before match
    • Half-time interventions
    • Team scouting
    • Player scouting
    • Personal Tactics Advising
    • Video tests for efficiency of tactical instruction
    • Intranet-platform
    • Instruction of coaches and referees
    More Support
  • 35. Vision II: More support for practice Vision
  • 36. Vision
    • Improvement in practical support
      • Better understanding of interventions
      • Insight into the qualitative nature of diagnostics
      • Methodological flexibility
      • Assessment and control of effectiveness of interventions and learning processes
      • Towards a Theory of coaching
      • Towards a new profession: Sports Analyst
    More Support
  • 37. Vision III: Improved technologies
      • General development
  • 38. General developments
    • Technological development drives progress in sports
      • Miniaturisation
      • Cost reduction
      • Power increase
    • Things are now ubiquitious we didn‘t dare to dream of 15 years ago and …
    • this is going to continue!!!
    New Technologies
  • 39. Vision III: Improved technologies
      • Example: Image Recognition
  • 40. Image recognition in tennis Image Recognition
  • 41. Image recognition in canoeing Image Recognition
  • 42. Image recognition in swimming Image Recognition
  • 43. Image recognition in football Image Recognition
  • 44. Automated model recognition positions actions situations tactics assessments
    • Automated tracking of positions by image processing
    • Bottom-up inferences by methods of artificial intelligence Beetz et al. (2009). ASPOGAMO. IJCSS Vol. 8, Ed.1
    Image Recognition
  • 45. Heatmaps Heatmap Materazzi, FIFA-Final 2006 Image Recognition
  • 46. Heatmaps Heatmap Gallas, FIFA-Final 2006 Image Recognition
  • 47. Relative Positions Relative Positions Buffon, FIFA-Final 2006 Image Recognition
  • 48. Relative Positions Relative Positions Pirlo, FIFA-Final 2006 Image Recognition
  • 49. Middle Right Left Direction Toni Totti Camoranesi Gattuso Zambrotta Cannavaro Materazzi Grosso Perrotta Pirlo Buffon
  • 50. Middle Right Left Direction Toni Totti Camoranesi Gattuso Zambrotta Cannavaro Materazzi Grosso Perrotta Pirlo Buffon Perrotta Pirlo Gattuso Camoranesi Totti Toni Grosso Materazzi Cannavaro Zambrotta Buffon
  • 51. Vision III: Improved technologies
      • Vision
  • 52. Silhouette detection shot put Image Recognition
  • 53. Silhouette detection Gymnastics I Image Recognition
  • 54. Silhouette detection Gymnastics II Image Recognition
  • 55. Silhouette Detection
    • Silhouette detection
      • Based on video data (multiple perspectives)
      • Fits 3D-Model
      • 53df stick figure is underlying
      • No markers required
    • Perspectives
      • Full kinematic description of movements available to video recording
      • Training and competition
      • Highly automated, fast processing
    Image Recognition
  • 56. Final Remarks
  • 57. Good Perspectives
    • Theoretical performance analysis
      • Progress in theoretical modeling
      • Capability to develop formal models with practical relevance
    • Practical performance analysis
      • Computer-based coaching gives real-time support
      • Wide-spread in top-level sports
      • Professional branch: game analysists
    • Technological developments
      • Continuous stream of challenges
      • Unforeseeable increase of impact
    Final Remarks
  • 58. Vielen Dank!