huizezhang-sherry
commented
5 years ago Summary (ToDo List)
- analysis the first serve behavior as well
- consider other dose indicators than counts: running distance, opponents ranking
- mixture model: incorporate the match fatigue and set fatigue
- player effect [comparison between groups]: male and female (decreasing rate)
- player effect: [subjective]: more aggressive player vs. more defensive player
- player effect: [20, 30] &[30, 40]
- player effect: an individual player's fatigue changing in time
- player effect: change in court type (clay, hardcourt, grass) - could go to individual's behavior
Q: Physiological question: will effort involved in first or second be more subjected to serve effects. Don't discredit one or the other now, but recognise they are different. Players could be actively accounting for this.??????
Significantly different serve speeds, second serve may indicate fatigue better: More conservative if fatigued, but will likely always be aggressive on first.
Densities: First serve very similar, Second serve is more varied
In the round, the speeds in the fourth round and sixth indicate fatigue
Volume dose model: Not obvious increase in fatigue until 5th Consider other options than count.
Tennis: First and second serve are very different. Second: Influenced by risk of losing on a double fault, usually more spin, which is slower First: Try to win on an ace if they can, or use powerful and well placed to gain an advantage.
Physiological question: will effort involved in first or second be more subjected to serve effects. Don't discredit one or the other now, but recognise they are different. Players could be actively accounting for this.
Within match, and within events. Rounds may make players more vulnerable.
Consider a hierarchical approach: event: round: match: player: shot A general problem with tennis, players affect models dramatically with characteristics specific to players and the opponent.
Definitely consider player affects, fatigue effects across a match and event differ by player.
Account for opponent effect: use a ranking or ratio of rankings
Serve is likely independent from opponent effects, as there is not interaction.
Player effects over the years: Injuries, per-retirement, age, eg 20 vs 30 year olds, could age be an advantage eg. more experienced, learned how to manage physical demands
Consider being more flexible, likely not a linear relationship
Helpful to have a data set of player characteristics, Differences in genders: height, upper body strength,
Serve as an advantage 65% win in mens, 57% win in womens Advantage in age and development curves of genders, younger women may be more likely to win grand slam, and older men more likely to win Consider interactions or separate models