Lies and statistics

Thanks to scientific data, team coaches see a match differently compared to the ordinary spectator, writes MARK KEOHANE in Business Day Sport Monthly.

My mate thought the player was awful. I thought he was pretty good. My mate thought his team attacked well. I thought his team was poor with ball in hand. My mate had a seat on the halfway line, 10 rows up. My seat was behind the posts 30 rows up.

‘We clearly were watching a different game,’ he said as we discussed the merits of who played well and which team deserved to win.

‘We clearly were,’ I said. ‘You were on the halfway line and I was behind the posts. Now go watch it on a television replay and see how different it looks.’

It was only through my involvement with the Springboks a decade ago, in the capacity of communications manager, that I was exposed to the detail of a rugby game and just how different opinion, depending on the seat location and the video analysis, can be.

It was also during this time that I was introduced to the detail of a match and player analysis. Watching rugby has never been the same.

Rugby professionalism was still an infant back then and in-depth statistical analysis was also very new, especially in determining a player’s off-the-ball work rate. Nevertheless, it changed how I viewed the players and the game.

Coaches, when talking about successful teams and players with that something out of the ordinary, will always single out how a player and a team scramble on defence because scrambling defence is not something that can be coached. It is not a skill, but an attitude. It is also a measurable and because of technological advances in analysis it can also be monitored.

A player who scrambles on defence and a team known for scrambling on defence have an association with passion, desire and a belief in each other. They are usually successful teams. These players and these teams are also well conditioned and physically capable of matching the intent to work harder off the ball with actually working harder.

In the amateur era no one would be the wiser if a player on the opposite wing did not anticipate the opposition play, turn and make the metres in a crosscover tackle attempt. All eyes were on the ball and the ball-carrier.

In the profession rugby is today, the eyes that matter most are those that aren’t on the ball.

Prozone, the most detailed of all rugby analysis programmes, has changed perceptions and players have nowhere to hide. Prozone provides coaches with a player’s off-the-ball contribution. It also shows when a player is operating at maximum and when a player is coasting.

The television camera and the spectator eye follow the ball. The data that interests each coach is what happens off the ball.

The public may judge a player with what he does when he has the ball but the quality of the player is more in how he reacts to situations, how he summarises the play in an instant and how his off-the-ball contribution manipulates what happens next. All Blacks flyhalf Dan Carter’s brilliance, as one example, is emphasised even more when you realise how he reads the game when he doesn’t have the ball.

Rugby is a game played as much without the ball as it is about a team having possession. It is also a game won often by the team whose players understand aggressive defence and the philosophy of defence being the greatest form of attack. Great defensive units are historically potent counter-attackers.

Chatting with good coaches is an inspiration and an education. The game they analyse is most definitely not the game the average Joe watches. What the public see in a player and what coaches see in a player is very different. It is because the devil is in the detail, and it is a detail to which the rugby public doesn’t have access.

The rugby supporter, through television match statistics and (most recently) the impressive Vodacom Stat app, are given the basic team statistics of possession, territory, penalties, lineouts, scrums and turnovers. The stat app has more detail in a team and individual capacity but it still is very different to the data from which coaches make their analysis.

Coaches talk of effectiveness in the first three phases of play and more and more of them don’t want their teams to have the ball if they haven’t made a territorial, field position or points gain from having had the ball in those first three phases.

Global statistics, which include all professional rugby played in the northern and southern hemispheres, show that 70% of tries are scored within the first three phases of the team in possession, with a large percentage within the 70 coming from counter-attacking. Multiple-phase play, in which teams can take it through 20-plus phases, inflate statistical returns and create the illusion of expansiveness but often amount to no field position gain and very little scoreboard reward.

The best teams are more effective when they don’t have the ball and the dominant teams rely on an ability to score from an aggressive defence that forces mistakes from the team whose players are carrying the ball.

The statistics the public use for pub discussions are not the ones influencing selection and game plan strategy

There is no escape for the professional rugby player. The data available to the coach allows him to monitor when a player walks, jogs, accelerates and sprints during a game. It also allows a coach to expose those players whose laziness is no longer excused because of a perceived lack of insight to how play unfolds.

The data shows just how long it takes a player to react to a situation and one of the most studied statistics is the time it takes a player to get back on his feet and make another contribution to the game.

Some players miss more tackles on their left shoulder than their right and some players are never within the first three arriving at a ruck situation. Some players control games and others are controlled by the flow of the game.

In recent weeks I chatted with several of the game’s most influential and successful coaches. I wanted to know how much value they put in individual statistics and how much in team statistics. I also wanted to know if they felt statistics could lie and if there was a paralysis of the game because of analysis through statistical saturation.

The majority didn’t read much into the individual statistics with ball in hand. So much was dictated by what players were doing when they didn’t have the ball.

The interpretation was also of greater relevance than the raw statistical return and no player statistics were viewed in isolation. The team statistics told the coaches more about the players than individual data revealed about the team.

The statistics the public use for pub discussions are not the ones influencing selection and game plan strategy. Statistics, correctly identified, applied and interpreted, make for a very different perspective of a player and of a team.

The coaches I interviewed said they had always trusted their gut instinct and nothing replaced an initial feel, but all of them said statistical returns, over a sustained period, had often shown up their bias towards a player’s strengths or a prejudice towards a perceived weakness.

The coaches spoke of times when their belief in a player influenced how they viewed the performance, yet data over a period of time showed the performance to be sub-standard. Conversely, they spoke of a player whose performance, when analysed with the appropriate data, consistently surprised their initial match day impression.

‘Don’t confuse numbers as being a statistical representation,’ was the message of the coaches. ‘The numbers may show a player has made 20 tackles but the analysis may show that 10 of the 20 were ineffective and some of them were tackles that took him out of the game and didn’t need to be made.

‘Nothing about the individual can be viewed in isolation. It has to be seen in the context of the play, the time of the play and the match situation.’

The level of detail about a player’s performance is staggering.

A player told me a story of missing a tackle in the last 10 minutes of a Test that led to a try that determined the result. He couldn’t believe he had missed the tackle and said as much in conversation with the opposing coach after the match. The coach told him they weren’t surprised because their analysis had shown the player, when fatigued in a match and because of several career operations, never made a tackle on his right shoulder in the last quarter. It was only then that they would run an attacking play directed at this player’s right shoulder. The player had been unaware of how he had always defensively favoured one shoulder over the other in the last quarter.

That’s just one example of many told to me by players and coaches of why the game played today differs to the free-spirited affairs of the amateur era.

Statistics and an informed application in the reading of the data have changed the dynamic of the game. A player’s effectiveness can be measured in more than an opinion. His vulnerability can be exposed statistically over a period of time and a coach’s brilliance in understanding the nuances of a game can be measured in the reading of his team’s data.

Equally, the naivety or incompetence of a coach. Data available will expose the limitations and laziness of a player. Conversely the attributes and decision-making.

These are but two examples.

Coaches aren’t obsessed with how many tackles a player makes or how many times a player kicks the ball or the metres gained. It’s when he does it, how often he does it and how accurate the act is in giving the team an advantage that builds pressure and ultimately leads to points.

The good coaches talk more about data than they do statistics and the very good ones know how to read this data.