A Rugbycology report -Super Rugby AU week one - Green and Gold Rugby
Analysis

A Rugbycology report -Super Rugby AU week one

A Rugbycology report -Super Rugby AU week one

It’s good to be back, and this time we are in a position to apply some proper analytics to our game and see if rugby in Australia is indeed on the upward curve we all hope it to be.

I want to therefore introduce our readers to a few new analytical concepts we use at Rugbycology to better understand team performance. These will be used weekly in this new segment on Green and Gold Rugby, and I hope to convert you all into stats nerds over the next few weeks.

Spectacle ratings

In our context it’s never good enough to merely win or lose games, because union as a spectator sport is in constant battle with league and others for eyeballs. For union to progress, we need to offer a more entertaining product at elite level. But how do we measure if a game was good or average or poor?

While there are many other variables at play, we will over the course of this tournament focus on one particular metric: the amount of play events per minute.

Let’s quickly then compare our games this weekend to the Aotearoa competition across the ditch to better understand this metric:

1

It is clear that after 3 weeks, the kiwi games are ‘busier’ than the Australia games. For the same dollar amount, the Kiwis serve up a large chips, while the Ozzie’s currently only feed you a medium. Both competitions have a similar amount of set play’s, however in-between these ‘stoppages’, a lot more currently happens in NZ than here. 

This is not the end of the world, provided that our team’s stay uber-competitive and that games are evenly matched. However we want our game to be as dynamic and entertaining as possible, and in this context we will be tracking the spectacle rating of each game for the remainder of the season.

Efficiency rating

Some teams play a flamboyant and dynamic brand of rugby. Others prefer a more pragmatic approach. Both styles are cool provided that it yields a good outcome on the scoreboard.

To allows us to track whether teams are ‘efficient’, regardless of the style of rugby they play, we use a simple efficiency rating that works like this:

On attack, an efficient team is able to move up-field, keep possession and score points. So the most efficient attack play is one that starts in your own 22 and ends in 7 points being scored. If however an attack starts in the opponent 22 and yields 3 points, it’s a lower efficiency play. Attacks that does not gain territory and where a teams ends up losing possession, will have a zero efficiency rating.

In defense, we add a value to a teams ability to prevent line-breaks and to regain possession within a reduced amount of phases. An efficient defensive play is this one where you do not lose territory, prevent points being scored against you,and win back the ball in under 3 phases.

With this definition understood, let’s see how the 4 Ozzie teams performed on attack and defense this week:

2

What? The Waratahs were the most efficient team of the weekend, yet lost their game? These ratings are crap! 

Not entirely no. What this rating shows is that with possession, the Tahs were the most effective. The problem is that the Reds starved them of possession. The trick is to gain territory AND keep possession (conquer) in the same play. Otherwise it matters little how efficiently you attacked. The Tahs only managed to ‘conquer’ their opponent 22.8% of the time – the least of the 4 teams.

What the Blues will want to improve is their ball protection, because its now clear that these lads know what to do with the ball once they get it!

Over time, these efficiency ratings become more valuable and will help us understand which teams are making progress in attack and defence, despite the strength of their opponents.

Use of possession

Earlier in the piece I talked about how some teams choose an expansive approach and other a more pragmatic one. To measure which is which, we track 4 key metrics, and to illustrate this point I combine below our two winning teams, and then compare their average with the Crusaders season average. This gives us a nice benchmark to work with moving forward:

3

Handles per attack

Whether you kick, or carry directly, or pass the ball, teams who ‘let the ball do more work’, can be categorised as being more ‘dynamic’ or even flamboyant. The Crusaders handle the ball a bit more and over the past two games sported an attack efficiency rating of 32.9 compared to the Brumbies 31.1 and reds 32.2.

% Times kicked

It’s long been a fact that Kiwi sides kick more than all the rest. The Blues and Saders kick a good lot, but what sets the Crusaders apart is the variation of their kicks. Don’t worry mate, we will show you all the cool stats as the season unfolds.

3 pass sequences

Now this is the kind of metric that make us stats nerds forget the rusk in our tea. It is an extremely valuable metric because it shows the desire of a team’s attack and reflects on the skillset available to execute it in the face of a rushing defense. Secondly it helps us measure the line-speed of the defensive team. With a super-committed line-speed it’s often very hard for teams to string together 3 passes in a row.

In this regard the Kiwi teams (and the English national team) are way ahead of the curve. They are able to shift the point of contact and use more width, which all has an effect on how quickly the defense gets tired.

Tramline play’s

Imagine 15 players stretching arm to arm across a pitch. The gaps in between each player is much bigger than when the same 15 stretch over only two-thirds of the pitch. The same thing occurs when teams use the entire width of the pitch on attack. Not only are the fatties forced to run further to make tackles – the gaps in between defenders are also stretched. So look at that stat above again. This is the stark reality we are dealing with at the moment when comparing the Kiwi sides to ours.

Let’s see below how this impacts a team’s ability to break the line and create gaps:

4

The good news for us is that our team’s are not that far behind when it comes to punching holes into the defence. In my opinion, strike running is probably the one area where Australia currently leads the world.

What we need to get better at is creating overlaps, and we can only do so if we develop a greater appetite for playing the footy all the way to the tramline. The thing about overlaps is that players make on average twice as many meters on an overlap as they do with clean breaks. Forced breaks in contrast rarely gain more than the minimum 2 meters required to qualify as a ‘break’.

In an era where we are truly blessed with great strike runners such as Paisami, Koroibete, Kuridrani etc, it is crucial that we sharpen our ability to get the ball to them more regularly.

Last thoughts.

I hope to update you guys on these efficiency metrics over the coming few weeks. We will also discuss our kicking game in more detail. Lastly we are studying the origin of tries to learn from which field zones and off which platforms Australian teams are more likely to score tries.

There is a lot to look forward to, and I hereby ask that you feel free to ask any questions you may have around these analytics on this platform or by engaging me on Twitter.

 

  • Brisneyland Local

    Great work there Brendon. An interesting set of metrics that is for sure. But once we have a larger set of sample data I wsure that these metrics will start to play out.

    • Thats it mate. I am chucking together every ounce of data from all over the world this year, suing the same entry methods, so we can compare apples with apples for once. It will be interesting to see what happens. For example I was already amazed to see that the tahs actually played good rugby but could not get hold of the ball!

      • Brisneyland Local

        Isnt that funny, because I watched the game and thought their first half was utter tripe. Aka normal Tahs, thier second half was good but they just couldnt get the apples. Effectiveness at the breakdown is what is killing them. Not precision clean out that actually clean out. And a 7 that didnt get one turnover!
        BUt having a Science Degree majoring in mathematics, I am more than encouraging what you are doing.

        • Ed

          The Tahs should watch this is a good video on the Bokke winning 2019 side, particularly on defence. Hopefully it begins at 12:40:
          https://youtu.be/sbpEG9e8yVs?t=760

        • Brisneyland Local

          Thansk Ed.

        • John Tynan

          Amen big fella! That was my whatsapp comment to the vets team.

  • Brilliant work Brendon, really enjoyed reading it.

  • KwAussie Rugby Lover

    Nice work Brendon and good to see some analysis even if I don’t understand it all, which I don’t. I always wonder about this though as we seem to keep showing metrics that demonstrate how a losing team has played well and yet they’ve lost. Maybe the metrics aren’t there yet. Take your efficiency metrics for example. As you mention they make the Tahs look good but what isn’t shown is that 103% efficiency of bugger all ball means that their efficiency isn’t the issue, it’s the bugger all ball they get that is the issue and yet that metric isn’t shown. I truly understand that this is just a start point and that as the season goes on the metrics will become more meaningful but I still wonder if what we measure provides us real useful data. To me it seems a bit like the Wallabies continually saying how well they trained during the week leading up to another loss. Maybe the training that was done so well was the wrong training. In this case a metric that shows how well a losing team did against a winning team may actually be the wrong metric. And in saying all this I admit that I don’t know the answer as I’ve never been into analysis of this depth before.

    • idiot savant

      Got your grumpy pants on this morning Karl?

      • KwAussie Rugby Lover

        hahahaha no not really I guess I’m just tired of being told all the metrics are great in a losing team

        • Losing teams can play well, just not quite as well as the opposition…

          What the efficiency stats really show is that the Reds were relatively sloppy on attack (they had a lot of ball, didn’t use it incredibly well, but they still had enough they outscored the Tahs), the Tahs didn’t have a lot of ball, but scored (relatively) a lot of points with what they did have. Put those two up against each other, and you have a side that looks efficient in attack and defence and the Tahs look great, but lose, and a side that looks mediocre in attack and offence but wins.

          If you contrast that to the other side of the ditch, although it took a long time, the Saders had a lot of the ball and racked up 40, the Landers not so much and were held to 20. I can’t tell you what the numbers would be, because I don’t know Brendan’s formula, but I’m willing to bet it drops out a set of numbers that says the Saders were more efficient on both sides of the ball than the Landers (and probably than both the Reds and the Tahs).

    • Nutta

      Have a coffee Dude. Then go for a walk. Breathe deeply. Do some stretching.

      After all that, tune back in and you will still hear Capt Sk8R say ‘But we trained well’.

      The difference will be you will be better for the exercise.

      • John Tynan

        aka “Hitting them well in the nets”. Kept the Marsh boys well paid for years.

  • idiot savant

    Fascinating Brendon. I look forward to more insights. It reveals what my eyes saw. The Tahs were more threatening and more skilled in attack than the Reds and if wasn’t for the period of Reds scrum dominance which affected the Tah’s confidence they would have won that game. The stats also reveal that the Reds were not able to shift the point of attack against the rush very well. Again thats what it looked like and as others have observed, the new ruck rules make it harder to shift the point of attack as you dont have spare men around the ball to use multiple passes in close.

    The tramline stats are interesting. What does this mean for strategies to get the ball there moire often? Unless you keep working back to the short side, do you need to stand fairly deep to get the ball there without engaging the rush? Do kiwi sides stand deeper in attack than Aussie sides? And how are we going to achieve overlaps with these ruck rules where you need to commit at least 3 players in attack to prevent the pilfer thereby leaving you undermanned against the defence?

    • John Tynan

      You reckon the scrum was the difference? I thought it was accuracy/effectiveness at the breakdown that shaded the good guys over the bad guys. If the Reds didn’t have that, I think the attack stats probably would have won it for the Voldemorts.

      • Huw Tindall

        I agree you Idiot and you John! Scrum stuffed the Tahs penalty wise and led to the card, territory and points. Probably enough to swing it the Tahs way if parity in the scrum. The breakdown was another nail in the NSW coffin and it’s a carry on from earlier in the season. The Tahs support players are not at the ruck fast enough nor are they accurate enough on the cleanout. When you’re up against backrow like the Reds currently have you are going to be stuffed. If we took that and the scrum away then you’d think the Tahs get the win. Still, ruck and scrum are two MASSIVE areas of the game so fair play to Queensland on the win. Not like the Tahs were hard done by.

        • John Tynan

          The rest of the comment with the vets boys was that “I still don’t think it’s accurate enough for a Brumbies game”, so we’ll see soon.

        • Huw Tindall

          Sounds like the vets know what they are talking about! Really liking Super AU already. Every game has something intriguing about it already. Brums v Reds should be very insightful when it lands.

  • formerflanker

    I started to read with the assumption this was a gee-up.
    Happy to have been wrong. Keep up your creative approach.

  • Hugh Cavill

    I found this really interesting, thanks Brendon. In particular the ‘3 pass sequence’ stat is one I’ve never seen but strikes me as a really valuable way of looking at how teams play the game. Obviously the Crusaders are miles in front of us (as they would be every other team in the world), but hopefully we can see that number get up a bit locally as our teams gel a bit more. Looking forward to the next article.

  • John Tynan

    Love it. The insight to creating an overlap and the metres made is awesome. Also good proof that he fundamentals haven’t really changed in however many years. ie the Ella’s were just masters in creating the overlap.

    • Huw Tindall

      Run into space, not at people. Sounds simple but clearly the players prefer hit ups mungo style!

      • John Tynan

        That’s the old kiwi thing about “running at branches, not at trees”

  • Hi Brendan,

    Can I ask why you’re using a simple addition of the efficiency ratings rather than a weighted average? Attack efficiency x possession + defence efficiency x opponents possession for example. I’m sure it makes sense, it’s just not instantly obvious on a first read.

    Also, I’d be interested to see the stat you call “conquer”. I understand the concept, how reliably does it correctly predict the outcome in the matches where you’ve done the analysis using it? If you do multivariate analysis can you get a strong model from that and a reasonably small number of other stats?

  • Brumby Runner

    Apples and apples you say Brendon, but then in the Use of Possession discussion you use an average for the two Aus teams against specific stats for the Crusaders. Not a very useful comparison in my opinion.

    Do the Crusaders stats relate to SRA only or do they include a history from previous years that is already in your bank? What about a comparison of the two Aus winners against the two NZ winners in their Round 1? Are the other NZ sides as efficient in their use of possession as the Crusaders?

  • Max Graham

    Good article. But…. you’ve confused efficiency for effectiveness. Efficiency is the measure of resources or energy it takes to achieve something, while effectiveness is the measure of success at achieving an objective. /pedant

  • Max Graham

    I don’t think your metric relates to ‘Efficiency’ – seems more like effectiveness, but you define that to mean something else. For example, efficiency at the ruck would be measured by the number of players you use (not necessarily whether you win it or not) while effectiveness would mean the rate in which the ball was successfully retained in rucks.

Analysis
@rugbycology

I research schools and club sport to help coaches create better training sessions and smarter game plays based on science. I believe that data hides these coaching gems that are very rewarding if you are willing to mine deep enough! Yes it's nerdy, but it works!

More in Analysis