Скачать или смотреть Predictive Modelling - Sports Analytics Methods

https://www.agilesportsanalytics.com

Take a free sports analytics assessment:
https://www.agilesportsanalytics.com/...

Predicting Modelling

Victor Holman, The Sports Analytics Expert, presents his Sports Analytics 3 Minute Drill - Predictive Modelling. Predictive modelling uses statistics to predict outcomes.

Most often the event one wants to predict is in the future, but predictive modelling can be applied to any type of unknown event, regardless of when it occurred.

For example, predictive models are often used to detect patterns, after the game has taken place.

In many cases the model is chosen on the basis of detection theory to try to guess the probability of an outcome given a set amount of input data

Going Inside the Inner Game: Predicting the Emotions of Professional Tennis Players from Match Broadcasts

Tennis is often considered to be a mental game as it is an individual sport and the majority of the time is spent preparing for the next play.

Consequently, coaching is often focused on the players' mentality.

The purpose of this study was to develop a method to predict emotions in sports.

It is believed that emotions can affect player performance, both positively and negatively.

In order to determine the effect emotions have on a player's performance it must be possible to measure emotion during competitions.

Facial recognition programs aide in this process.

This study focuses on seven emotions: anger, annoyance, anxiety, dejection, elation, focus and fired up.

Images were gathered from a variety of data sets.
Faces of players were isolated and assigned an emotion.

In total there were 7,952 images.

Workers were found through Amazon Mechanical Turk who were asked to rate the intensity of each emotion from 0 to 10 with 10 being the most intense.

Each image was rated by 5 people and the median (middle number) of the ratings was used as the final intensity.

This information was then used to train the model to detect emotions and how to label its intensity.

This model was then used to analyze the facial emotions of Novak
Djokovic, Roger Federer, Andy Murray and Rafael Nadal at the 2017

Australian Open including two matches per player.
From this analysis profiles were created for each player.

These profiles indicated that each player experienced unique emotions throughout a competition.

Anxiety was the most common emotion among all the players and the most predominant emotion of Rafael Nadal.

Roger Federer was the player whose emotion was typically either neutral or focused. Andy Murray and Novak Djokovic both experienced high level of anxiety.

Djokovic experienced fired up, dejection and anger more often than the other players indicating that he demonstrated the most emotions during the competition among the four players.

Elation and annoyance were rarely seen, with Andy Murray expressing these emotions the most.

The fact that he experiences these two opposing emotions with a greater frequency corresponds with the reputation he has for being one the most volatile players on tour.

This data was used to test two commonly held beliefs.

The first is that players' emotions are a response to how they perform and the second is that a player's emotional reactions influence how they perform throughout the rest of the match.

These beliefs were tested by combining each facial image with game context including the score, the winner of the point, the server of the point and the importance of the point.

Three of the players showed strong emotional responses to their performance, the exception being Roger Federer.

To test the belief that a player's emotion affects their play was tested by looking at the play following each facial image.

Emotions affected the next play significantly for Nadal and Djokovic with much weaker links for Federer and Murray.

Coaches can use this information to help their players understand the relationships between their own emotions and how they play.

If players were able to focus on 'letting go' of the emotions that hurt their performance while 'hanging on' to those emotions that improved their performance it is likely that they would improve their overall performance and chances of winning.

And that’s predictive modelling applied in sports analytics… in 3 minutes