Because I’m on a roll with the R Shiny web app stuff, I figured I’d make another. This one allows you to look at how each player is trending throughout the season in their serve, receive, or attacking stats.
For serving, we look at the outcome of the serve (different reception qualities or a service error). We then look at how likely the opponent is to win in all of those situations (let’s say they pass it perfectly for example, opponents are hitting 0.158 on that first ball against Harker). Because we care about how likely Harker is to win, we just flip the sign and say that Harker is expected to win at -0.158 (we are more likely to lose on that first point when the opponent passes perfectly, which makes logical sense). We look at all of a player’s serves for each match and average out these outcomes to a single number. This is what’s presented on the chart.
For receiving, we simply looked at the other side of serving. When opponents serve and Harker passes the ball at different qualities, how likely is Harker to win or lose given that outcome of the pass. For us on a perfect pass, we are expected to hit 0.396. Therefore, the max score a passer can receive is a 0.396 meaning that he passes every single ball he saw, perfectly.
For attacking, we use a similar methodology. We look at all of the possible outcomes of attacking (kill, error, dig by opponent, cover by Harker) and we look at how each of these situations affects our ability to win the possession. Obviously a kill or error are valued at +1 and -1 respectively, but a cover by Harker is worth 0.324, while hitting into a good dig by the opponent is worth -0.106 to Harker. So we factor in all of these things are spit out an average of all the individual’s swings.