RUGBYCOLOGY: The Wallabies player rival teams target at the restart, and why he's a weakness

By Brendon Shields / Expert

All supporters know the feeling: your team just scored a great try, only to lose possession from the restart and then concede points themselves.

With the July series in the southern hemisphere in full swing, we can now work with fresh data to study how teams approach restarts and who is having the most success.

What is a restart?

Restarts are essentially kickoffs and can occur from the halfway line, the 22-metre line and the goal line. These are considered ‘structured’ platforms because teams can train these set plays and formulate strategies. Unlike a turnover, for example, coaches and analysts can roughly predict where their opponents will land their kicks and plan accordingly.

Good and bad outcomes

To better grasp which teams have good or bad restart outcomes, we need to first study the range of outcomes that teams experience each time they get an attack opportunity, be it from a restart or a scrum or a turnover or whatever.

The worst possible outcome is to lose possession to a turnover or a penalty without having gained any territory. An average outcome would be to retain possession, even if no territory was gained. Similarly, gaining territory but losing possession with, for example, a forward pass is considered an average outcome because you get to contest the next scrum.

Great outcomes are when you gain territory while also retaining the ball. An even better outcome is when you score points. So with these broad classifications of outcomes now defined, we can study how well teams performed this weekend when attacking from restarts.

Last game Poor Average Great
All Blacks 14.2 71.4 14.2
Ireland 0 75 25
Wallabies 40 60 0
England 14.2 71.4 14.2
Springboks 33.3 55.6 11.1
Wales 14.3 85.7 0
Argentina 12.5 87.5 0
Scotland 50 50 0

Based on the above data from this past Saturday (2 July 2020), it’s the Wallabies who take the wooden spoon, while Ireland have the best outcomes.

Data analysis, however, always takes a long-term view. We try to spot trends and often try to measure the strengths and weaknesses that teams may not even realise they have. With this in mind, let’s have a look at our sample teams’ performances from restarts over the last six or so games.

Season Poor Average Great
All Blacks 27.1 61 11.8
Ireland 37.1 45.7 17.1
Wallabies 41.7 52.8 5.6
England 21.1 65.8 13.1
Springboks 28 62 10
Wales 37.5 62.5 0
Argentina 23.5 76.5 0
Scotland 33.3 66.7 0

With restarts I would argue that it’s better to have fewer ‘poor’ outcomes than to have a high rate of ‘good’ outcomes. This is because the vast majority of restarts land in one’s own half, and losing possession to a turnover or penalty can have dire consequences. It’s something you just have to get right.

So while Wales achieve no great outcomes, they also have a lower percentage of poor outcomes from restarts, which means the Wallabies are currently the team with the worst attack from restarts out of the teams sampled.

Towards predictability: how to ensure your opponent gains a poor outcome

Many years ago I was fortunate to attend a seminar with Springboks coach Nick Mallett. Nick’s technical consultant at the time was a young Jake White, who told a good story pertaining to how video analysis changed the game around the start of the new millennium.

When studying game tapes White noticed how the All Blacks always kicked their restarts directly to Rassie Erasmus, who at the time was the Boks No. 6 flanker. Erasmus had a very good skill set and was safe under the high ball, so White could not understand why the Kiwis would always kick on him.

After the Tri Nations in 1998 White asked his New Zealand counterpart why they kicked on Rassie when this player was the safest bet to secure the ball. According to the Kiwis, Rassie’s behaviour when fielding kicks was entirely predictable. The player would run a metre or so to the left and then step off the left foot and take contact. Rassie would not pass or kick the ball, ever.

The Kiwis knew that Rassie will not fumble the ball, but his behaviour meant that they could plan entirely how to defend the Boks restart.

For the same reason, we see teams kick on Marika Koroibete. The winger is by far Australia’s most explosive player, yet his behaviour when fielding restarts is so predictable that he is the preferred fielder for most teams kicking onto the Wallabies.

UK rugby writer Mick Cleary joins Brett McKay and Harry Jones to discuss the first Test and look ahead to game two

Marika will catch the ball, accelerate and try to bash through contact. He often succeeds, but his support is too slow to the ruck and the Wallabies end up turning over the ball, often in their own 22. Being a backline player of slightly smaller stature means that Marika leaves a larger ‘gate’ for the opposing defence to play through. This is why forwards fielding kicks are often safer bets, even if they are not as safe under the high ball as backs are.

The reality is that the Wallabies restart is struggling, and the reason for this is known to most opponents. It will be interesting to see whether the Wallabies can adapt during this season and make the restart a strength instead of a weakness.

The Crowd Says:

AUTHOR

2022-07-07T20:39:27+00:00

Brendon Shields

Expert


You are absolutely right. Marika fielding, and the predictability is brings - may be something the Wallabies also prefer. However you cannot refute the data over 2021 where the Wallabies restart exit is clearly shown to have worse outcomes than the rest. I dont think Marika is to blame here at all. I do however argue that Marika has an opportunity to mix things up and offer an element of surprise.

AUTHOR

2022-07-07T20:35:59+00:00

Brendon Shields

Expert


No single aspect of a game (such as the restart platform) will mean the difference between losing and winning for any team. However when we look at a higher win % over time - or even dominance - we have to isolate those parts of teams that prevent them from performing optimally. The Wallabies have ironed out so many weaknesses in their game - but this one lingers. Similarly Ireland may have the best restart, but they have other issues.

2022-07-07T20:22:05+00:00

jcr

Roar Rookie


So does he ask the opposition to kick to him?

2022-07-07T10:40:13+00:00

Ankle-tapped Waterboy

Roar Rookie


In truth I don't default to such uniformity, due to precision bias and false precision, and I know enough now to keep an eye out for those things in your data sets. That is not to cast doubt on all your data, two significant figures being justified at those times when the data sets are sizeable enough to permit two significant figures. (Which is why you get asked about the size). Metrology. Gotta love it.

AUTHOR

2022-07-07T09:03:16+00:00

Brendon Shields

Expert


I just always prefer 2 digits because there are stats where it makes a difference. For example passes per ruck. Often two teams will register 1 each. But add one digit and its 0.8 each. Add another, and its 0.76 and 0.79: Eureka! For uniformity I then always just work with two digits.

2022-07-07T05:59:17+00:00

Sneaks

Guest


Useful analysis, but it seems like the story doesn't match what actually happened on Saturday? By my count, Australia fielded 6 kickoffs on the weekend. Marika caught 4 and they were all what might be considered 'average' (by your categories) and were followed by reasonable clearing kicks (1 was a bit short). The two we failed to field were shorter kick-offs that didn't go to Marika - they were failed to be secured by the forwards in those areas. If I was Rennie and Co, I wouldn't look at Saturday and think that my biggest problem was Marika fielding kickoffs when every time he did, it ended up as a positive or neutral outcome. It's also all well and good to say that Marika fielding kicks is a predictable outcome and therefore bad for Australia, but aren't we discounting what Australia wants from this scenario. Maybe we're ok with this being a predictable scenario and having our big forwards on hand to a) Secure the first ruck with three forwards and then b) have 3 forwards on hand ready to set up another ruck to open up the angle for White to clear if needed.

2022-07-07T04:10:28+00:00

numpty

Roar Rookie


You could put him plenty of other spots. You do not put a player you consider weak at restarts in that position. So the question is, what does rennie see that Brendon and his stats doesn't?

2022-07-07T03:45:50+00:00

Crusher_13

Roar Rookie


It’s a safe spot to put him. He doesn’t have the attributes to be anywhere else. Not tall enough to be lifted, probably not tall enough as a lifter. No natural kicking game to be mid field or any deeper. So you put him against a sideline, no real possibility to kick or pass, as long as he takes the kick you hope he makes a few meters and doesn’t run away from his support.

2022-07-07T03:39:12+00:00

numpty

Roar Rookie


This was sort of my point elsewhere in the thread. If Aus were trying to hide him, they wouldn't put him there, one of if not the most common spot to kick to. So what does rennie see that Brendon doesn't? Brendons stats are obviously right, but wallabies must think marika isnt the issue.

2022-07-07T03:03:13+00:00

Crusher_13

Roar Rookie


It’s the type of kick. An accurate high kick still benefits the receiver. They can get set with a lifter. A less accurate low kick benefits the chasers as the receivers have no chance to get set with lifters. DC was great at the low / flat restart that isolated a jumper and allowed the chasers to compete in front of the receiver.

2022-07-07T02:39:00+00:00

Crusher_13

Roar Rookie


The problem is you can’t hide him. Teams will kick to him no matter where he goes. It’s a weakness in his game that he doesn’t like to pass or kick, and that he likes run into bigger humans. I’d guess he stays where he is so if he does turn it over oppositions have limited options against a sideline. Who do the wallabies position near him to help in contact? Hooper would be my pick, good footwork, able to support the fastest and not needed as a lifter, catcher or kicker.

2022-07-07T01:37:39+00:00

Jonty Shonty

Roar Pro


Love this!! And you’re right – it is 100% my biggest frustration. Score points, then fail to exit efficiently and concede easy points. That’s a failure when trying to build scoreboard pressure which at test level is so important. So interesting the insight re Rassie Erasmus and now Koro. Although I did guffaw at Koro being “a backline player of slightly smaller stature”. Dude hits like a freight train loaded up with ore.

2022-07-07T01:11:41+00:00

Adam (Though An Imposter)

Roar Rookie


The Kerevi kicks certainly looked like a deliberate strategy. I'm not entirely convinced it worked, but maybe needs to be practised.

2022-07-07T01:07:46+00:00

numpty

Roar Rookie


They have. He who should not be named (cheika) got Marika to kick out of hand early in a number of games to put teams in two minds (with good effect). Clearly Rennie is onto it considering kerevi kicked off first phase prime attacking ball on the weekend. Not something generally expected.

2022-07-06T22:53:20+00:00

Sinclair Whitbourne

Roar Rookie


Gentlemen (or are we 'poor, twisted fools'?), sad to say I think we show our ages. For the joy of a childhood with the Goons on the radio on Saturdays I am eternally grateful.

2022-07-06T22:16:02+00:00

Spew_81

Roar Rookie


Definitely a combination of the kicker and receiver. I'm unsure if it was Kieran Reid who always received the restart. The kick has to be pinpoint to start with, otherwise you'd be risking giving possession to the opposition right outside the 22. So they must've had great faith with Carter's accuracy.

2022-07-06T14:48:43+00:00

Francisco Roldan

Roar Rookie


Very interesting your metrics Brendon...! I loved the detail commented on RSA resets. I think that in contemporary rugby, the amount of details to take into account to execute a collective movement is quite overwhelming. And the anecdote about Rassie Erasmus is proof of this: a no-brainer to ensure the start of an 'orderly' execution. But in that same obviousness lies our fascination. Greetings :happy:

2022-07-06T14:18:25+00:00

SDRedsFan

Roar Rookie


Maybe it's just that if we were better at these 1% things we may have won by more?

2022-07-06T14:16:49+00:00

SDRedsFan

Roar Rookie


Maybe because Dan Carter was playing with Keiran Read as he seemed to be very good at winning kick restarts.

2022-07-06T14:09:16+00:00

Ankle-tapped Waterboy

Roar Rookie


Almost there. I'll continue trying, with thanks. You have a choice about the percentages. Why do you go to one digit when it appears that the digit adds no value to the analysis and therefore is just clutter. Real life example: 14% could be: 14% 14.2% 14.23% 14.23146% The last two are silly and I made them up to show the point but my question is that you are providing accuracy of 14.2% and not 14% and so you are obviously seeing some meaning in the data that is escaping some readers e.g. me, otherwise you would have simply reported "14". Because meaningless data is just clutter, and I am curious to understand what underpins your reasoning and interpretation where including the .2 adds value as compared to not including it and just saying "14". Particularly as the original data set is so small.

More Comments on The Roar

Read more at The Roar