Deconstructing the ball runner’s options

Our hirsute Saffa analyst has been analysing and interpreting the performance of different positions and attempting to discover new statistical tools to review players for the last few years.

This exercise is an attempt to provide him with possibilities to evaluate those players receiving a ball in open play.

I will not include here receiving a ball in a lineout, only open play where a player receives a pass from a teammate or picks up the ball from the ground. I will also try to show that receiving a ball from a kick from the opposition should have different measures than one received from a pass.

I will not provide statistics of the different options or provide an evaluation of the relative importance of each option. That is an exercise for someone else or a debate for the panel. I will also try to limit the amount of statistical discussion.

When a player receives a ball in open play from a teammate he has a few options assuming he doesn’t drop the pass. He can pass it, kick it, score a try, run out of bounds or be tackled or held up in a maul. The sum total of all these options is 100 per cent. You can then evaluate different positions by the distribution of options by a player during the game.

You can also look at them by forwards versus backs and any other option you prefer. Back rowers against front rowers, for example. This work assumes that the role ends once one of these options occurs.

But it gets complicated. The pass can be successful or not successful. An unsuccessful pass means the receiver could not or did not catch the ball or was intercepted. You could have a ‘successful’ hospital pass, but I digress.

James O’Connor. (Photo by Anthony Au-Yeung/Getty Images)

The kick has so many options that it becomes impossible to describe them without getting subjective. How do you define a successful kick?

Let’s look at the options when the player is tackled. Here we have quite a few options again. The player can lose the ball at the point of tackle. He can perform a successful or unsuccessful off-load. When the player gets to ground his team can maintain possession of the ball or lose it to the opposition. Committing a penalty is part of losing possession for this exercise.

Before the outcomes of passing, kicking, scoring or being tackled, the player can evade tackles or break through tackles. But if a player is caught with the ball, there are only two outcomes: the player is tackled or a maul ensues even if I described different outcomes from the tackle.

Some of these outcomes are reported in statistics of games. For example, you know how many times a scrumhalf passes the ball and how many times he kicks. It is similar for flyhalves. You have pass-kick-run ratios. In the case of flyhalves, statistics (if I am correct) calculate run metres, including receiving the ball from kicks. Modern flyhalves position themselves as a second fullback in defence, so they have a high chance of long runs from the kick compared to the metres run from a pass.

I am proposing that we separate run metres from kicks to run metres from a pass, and the same for kick passes. Distance run after a kick pass is closer to run distance after a pass than a kick from the opposition. And then you have box kicks where distance ran after receiving tends to be minimal.

So statistics will be very different for some players. For example, scrum halves should be tackled very infrequently if they pass and kick but not run with the ball unless they receive it in open play not from a ruck or scrum. Their run-pass-kick-tackled ratio should be very different to a prop.

The front five will tend to be tackled a lot but achieve relatively low run metres. These players should be judged by how many times they lose the ball in ‘contact’ compared to maintaining possession.

We have also seen reported the number of times a player’s runs beyond the advantage line, but this is not relatively common as reported per player. Based on position, some players have more chance of going beyond the advantage line, such as back rowers. These are the outliers that will become interesting to study!

Using these tools you can evaluate backs in play. Does Rieko Ioane get tackled more than he passes the ball? How many times do backs lose possession when tackled compared to forwards? If Blues wings do not run many metres with the ball when Ioane plays centre compared to when someone else does, you have coaching tools to use.

I haven’t even thought of how to use all these metrics. But if you have a player with a relatively large number of lost possessions per game, you can work on these skills or target that player if in the opposition. If your forwards have few metres per run as a group or too many lost possessions, you can do something.

What is going to happen is that many of these outcomes at a senior level will occur quite infrequently, so you will have to do group analysis to understand what can be done. How many knock-ons in a match? How many penalties in open play around the tackled ball carrier?

What is important to understand is that the ball carrier can only be tackled once per carry. He may receive many tackle attempts, but the last one that brings him to ground will by definition be successful. If a player never passes or kicks (or chooses the other options), then the tackle rate for him will be equal to the sum of outcomes, or 100 per cent. Never more than one tackle per carry.

You could eventually see a report such as 30-35-30-5 for kick-pass-tackled-other for each player in the game plus distance run. By then, looking at trends over many matches, you can see tendencies of players and teams that may help understand how they play or how each player performs. But in a game where the number of tackles can never be more than the number of runs it will always be a proportion of runs.

I know this is a rather superficial opening of options. But I hope it is a platform to help find or develop statistical tools to measure player or team performance. Now it’s your turn.