Watching the game on Saturday I was alarmed at how poorly the Wallabies scrum was performing. And watching a replay on Sunday it didn’t seem any better. Now I’ve taken a closer look, it was even worse than I initially thought.
Firstly, let me repeat my usual warning on the folly of relying on statistics alone to justify any argument you choose to make on any topic in rugby. Depending on which sources you use for statistics, you’ll probably get a different result from each and even then, just because the statistics are published on a website or flashed up on your television screen it doesn’t necessarily mean they are correct or tell you the whole picture.
Let’s look at the scrum statistics from Saturday’s game from some different sources:
Rugby Stats Website (Sports Data) | Fox Sports Website | ESPN Website | My Count | |
Scrums Set With Wallabies Feed | – | – | – | 4 |
Scrums Re-set | – | – | – | 1 |
Scrums Decided | 1 | 1 | 3 | 3 |
Scrums Won | 1 | 1 | 2 | 1 |
Scrums Won – Pen / FK | – | – | – | 1 |
Scrums Lost | – | – | 1 | – |
Scrums Lost – Pen / FK | – | – | – | 1 |
% Won | 100% | 100% | 67% | 67% |
Scrums Set With Wales Feed | – | – | – | 11 |
Scrums Re-set | – | – | – | 3 |
Scrums Decided | 3 | 3 | 9 | 8 |
Scrums Won | 3 | 3 | 9 | 2 |
Scrums Won – Pen / FK | – | – | – | 6 |
Scrums Lost | – | – | – | – |
Scrums Lost – Pen / FK | – | – | – | – |
% Won | 100% | 100% | 100% | 100% |
It should be no surprise that the Sports Data and FOXSports numbers are the same, as Sports Data lists FOXSports as one of its clients. ESPN’s data comes from Stuart Farmer, who used to provide the statistics to Planet Rugby. I don’t have the time available to record full statistics for each game any more so this is just my count from watching the game.
Overall my numbers are in line with ESPN’s, except for one extra scrum they recorded for Wales. I may simply have missed one, but I can’t find the ninth scrum they recorded. I presume that the difference to the Sports Data numbers is because ESPN only recorded scrums that were not decided with a penalty or free kick. If that is the case, I don’t follow the rationale behind that method and it doesn’t really give an indication of what happened.
If you relied on statistics alone and used the data from either of the first two websites, you’d think all was well with the Wallabies scrum. After all, they won 100% on their own feed, the same as Wales! If you use the data from ESPN, you’d be a little concerned with a winning percentage of 66.67%, but given that was over just three scrums, it doesn’t look too bad. However, a look at any of those first three columns of statistics doesn’t give the real picture.
Even without viewing any footage from the game, more detailed statistics would get you asking further questions regarding the number of scrums reset and the number of scrums determined by a penalty or free kick. That doesn’t apply so much to scrums on the Wallabies’ feed because I don’t see those numbers as being too far away from what you’d expect from only three scrums awarded. However, the more detailed numbers when Wales were feeding the ball show 3 of 11 scrums set were reset (27.27%) and 6 of the 8 scrums decided (75.00%) resulted in a penalty or free kick to Wales. Certainly the percentage of penalties and free kicks awarded to Wales on their feed indicates that the referee believed the Wallabies had significant problems with their scrum, and I can’t disagree with him.
So let’s dig further. Of the six Welsh scrums decided by penalty or free kick, two were free kicks for the Wallabies engaging early. In addition, the only scrum the Wallabies lost on their own feed was a third free kick against them for the same offence. Digging further again, two of those three offences occurred when the starting front row was on the field for the Wallabies, and the last in the first scrum packed with Alexander and Moore on as replacements. The first early call came in the first scrum of the game on the Wallabies’ feed.
There are two primary reasons for an early engagement – early in the game when timing is not quite right, and when you are losing the hit and you try to get in early to overcome this. I’ll put the offence in the first scrum of the game and the first scrum with replacement front rowers on the field down to the latter, even though the first scrum with the replacement front rowers was five metres out from the Wallabies’ goal line, which was a scrum in a real pressure situation. The second free kick for an early engagement came at a scrum after the Welsh had monstered the Wallabies scrum, which suggests that the Wallabies were looking for something extra to help them try and compete. The fact that the Welsh didn’t incur a free kick all match for an early engagement suggests that they were more confident of their scrum and could be more patient when it came to the engage. If you know you can afford to be slightly behind with the engage but still dominate the scrum you’re less likely to engage early and risk giving a free kick away.
The remaining four penalties the Wallabies conceded at scrums fed by the Welsh were as a result of Benn Robinson going to ground on one occasion and having his head and shoulders below his hips on another occasion, Sekope Kepu failing to hold the weight on one occasion, and Ben Alexander collapsing the scrum on one occasion. The last penalty awarded against Robinson could just as easily have gone against Alexander.
While the Wallabies scrum hasn’t been great in the first two Tests, in this match it was under enormous pressure. Some examples where the scrum was in trouble are shown in the following video.
I’m confused as to whether Patricio Noriega is still the Wallabies scrum coach or whether the forwards coach, Andrew Blades, has now taken over that role, but whoever’s in charge is going to have their work cut out for them prior to the commencement of The Rugby Championship as Argentina, New Zealand and South Africa all have strong scrums.