There have been media reports about how picking George Smith this weekend constitutes a ‘tough choice’ for the Wallabies.
Peering past such characteristically bland insights into the bloody obvious, I thought a brief statistical analysis (ie a form of the objective basis that the paper should actually be basing their conclusions on for the reader’s benefit) might be useful.
The Findings
Using the Fox Sports’ stats from Super Rugby, two things become very obvious. First, Smith, Gill, and Hooper are all very different types of 7. They do different things, they have different strengths and weaknesses.
Second, based on the stats, Smith is clearly the best pick.
The Analysis
Looking more in detail, neither Gill nor Hooper is the full package (yet). Gill is basically a poacher/a tackle jackal/and not much else. He is what Smith was when Smith was 20 years old.
Other than that, he adds very little – he doesn’t run the ball much and he doesn’t make a lot of tackles. (Anecdotally, one might add that he’s just not good enough in the in close contact for a full 80 minutes of test rugby at the moment.)
As for Hooper, he does a lot, but not everything he does is effective. On average he gives up two penalties every single match (to 1 for Smith, and 0.5 for Gill) and he also misses more than two tackles a match.
On balance, however, he remains a more well-rounded pick than Gill. He also adds an x-factor with his attack: he is in the top 15 players in all of Super Rugby for line breaks. He has more linebreaks, in fact, than Israel Folau.
Smith is clearly in a class above both of these players. He contributes in every category (except linebreaks). He was never known as an effective ball runner, but that is obviously something he has added to his game in his time away from Australia.
He runs the ball more often even than Hooper, he busts tackles, and he offloads like it’s his job (1.3 per match versus 0.2 for Gill and 0.43 for Hooper).
It is worth noting that Izzy’s second try in Brisbane came from an offload, and our offloading on Saturday was particularly poor (one from Palu to no-one when he had four players to pick from springs to mind).
He also effects slightly more turnovers and pilfers even than Gill (1.3 per match versus 1.06; the figure is only 0.64 for Hooper). In other words, even at what Gill does best, Smith does it better.
Who to pick?
Obviously, stats only tell part of the story. But they can also help debunk human preconceptions and I think these figues do that effectively.
There are no grounds, statistically, for picking Gill over Smith. The logical choice would therefore be to bring Smith into the squad for Gill. Again, it is worth noting that 7s in the modern era classically get better as they get older and diversify their skills.
Everyone blames Carlos Spencer for the All Blacks losing the World Cup semi-final to the Wallabies in 2003. The real culprit was none other than Richie McCaw, who gave away no fewer than four penalties in the match, three of them kickable.
Since then, of course, he has added more running and brawn to his game and I think we can expect Gill to do the same.
Regarding Smith’s fitness, it’s worth noting that he himself said before last week’s match that he was ready to go. I think it can be taken that Smith is a professional and knows his body.
He is equal tenth on the all time test caps list and finished playing in 2009, so he could have added more had he so wanted. Few have been able to match that sort of longevity in a high-contact position. He knows when he is fit to play.
There’s also Smith’s experience and history of great performances in big matches. Neither of the other two come close to him on that count.
Finally, it’s often a decent rule of thumb to do what the opposition don’t want you to do.
Again, this is a no-brainer. Especially with Warburton out, the last thing the Lions want is to run out at Sydney, with the emotions high, no momentum in the series, and see George Bloody Smith at 7.
For all these reasons, Smith needs to start, with Hooper on the bench, probably to come on for Palu at the 50 minute mark.
Reginald Munday
Guest
An injury in the backrow and the three quarter line and Australia are toast
Reginald Munday
Guest
Interesting that you mentioned Reds v Tahs. Nobody else did...
Jereme Lane
Roar Guru
mate we might have to call this a reds fan and a tahs fan agreeing to disagree. some food for thought though. hooper was way better over the ball last year at the brumbies than he has been this year but i suspect thats more to do with the position cheika has him defending. for example he has defended in the centres from the lineouts and is making way more tackles than gill but gill is getting more steals (about 11 from 15 games from memory) so maybe we can say that gill is better at arriving as the first man still on his feet. one thing i firmly believe though as someone said below, the rules have changed and there aren't as many turnovers at the breakdown as there was 3+ years ago as shown by gill's comp-leading stats (although there are more in tests without a doubt and interestingly second on that list is Rene Ranger with 8). the responsibilities are shared more evenly among the forwards now and you need to be more than just a fetcher. think about david pocock when he first hit the scene (and George Smith too for that matter) he was a fetcher and brilliant at it but then he filled out a bit more and sorted his leg drive and body height and built himself into a strong ballrunner and thats when people started really talking about him. the one thing that hooper has that the others dont is blistering pace. people say hooper is small (i dont, ive seen him and he's not small but i think he has a 'boyish' physique which makes people think he looks small. he's listed on wikipedia as 97 to gill's 94 but i dont think that sounds right) but you can substitute size for explosive pace and when hooper fills out a bit and works out how to use his pace in the other areas of his game he will be even better. gill on the other hand is a fetcher and when that style of ref and game comes about he is awesome but those games aren't every game. just looking at Gill's stats for game 1 and maybe there's a bit of proof for what i'm saying here. with that sort of refereeing, ie not much of a contest, gill's stats were awful. 33mins, 1 tackle, 1 offload, 2 runs (8m) and 1 ruck hit
David Barbeler
Guest
Hey mate. Nice article. Not all us journos are lazy with the statistical analysis though. Here's an example I did with AAP in the lead up to the squad announcement. Similar to yours actually. Bit shorter, but word limit constraints are a killer... Cheers! :-) http://www.theroar.com.au/2013/05/18/smith-hooper-battle-for-wallabies-jumper/
Minz
Guest
Funny, I thought Palu was terrific in the first game, not so much in the second. Mowen's been everywhere in both. But I haven't seen the stats :)
ohtani's jacket
Guest
But do the Wallabies really want to get into an arm wrestle at the breakdown if the Lions' tactics are to win breakdown penalties and kick penalty goals? What exactly are the Wallabies trying to stop? The Lions attack in Melbourne consisted of mainly a kick chase game and mauling. We never got to see what they planned to do from set piece since their lineout and scrum broke down. Personally, I think it would be worth fanning out ala the All Blacks and encouraging the Lions to run at them. At least test them out. Not giving away kickable penalties would go a long way to winning the game, I believe.
Josh M
Guest
It wasn't a tough choice. It was a no brainer.
RPollock
Guest
Geo Smith & Richie Mc are the standout No 7 of the last decade or so. Imagine if McGaw was available for selection and missed out for the All Blacks, there would be a riot in N Z. Thank God sense has prevailed and hopefully Geo gets the chocolates in what will almost certainly be his final game for Aust. He deserves it.
Golden
Guest
OJ, I agree with you on the stats, until they become consistently reliable they are nothing more than a general guide. I think averaging stats across a super rugby season does start to give a decent picture of performance and in this article's case the stats seem to support the general opinion about the 3 backrowers in question. There is a focus on the breakdown because if the Wallabies can slow the Lions' ball by contesting at the breakdown it slows the game down, giving the Wallabies a better chance of lining up defensively, shutting down opportunities for the Lions. Furthermore, contesting the ball at the breakdown will suck Lions forwards in to secure their ball, increasing their work rate and reducing their impact on the next phase. So this is why Deans has picked a combative backrow. In regards to George Smith's additional contributions, there will be no other Wallabies player on the field this Saturday who has the experience at test level that he offers. He is a natural leader and the younger players will gain confidence from playing alongside him. Finally, I think you must have missed all the press - George Smith doesn't have a dodgy knee - he "had" a dodgy knee and declared himself fit last week. I think you can expect to see Smith in full flight on Saturday because he rarely has a bad game. He just might be the difference the Wallabies need to beat the Lions.
Nick Turnbull
Guest
Mate you need the ball before you can run with it. How many meters did he run at 12 not 7? Gill has had little game time, given opportunity with more game time who knows? I'd back him over Hooper on the deck any day.
Cross
Guest
-- Comment from The Roar's iPhone app.
Cross
Guest
ND it isn't too outlandish, while at the brumbies academy Laurie Fisher told TPN to make the move from back row to hooker. That being said as Pocock has already cracked the big time I doubt it would ever happen... -- Comment from The Roar's iPhone app.
ohtani's jacket
Guest
It's odd how rugby went from being a game with barely any statistical analysis to a sport with an over-imbalance of unreliable statistical data. It's all right to have rugby stats websites and bloggers creating their own numbers, but there's such little consistency between the numbers that it only paints a general picture. The reason I bring this up is that arguments are getting awfully stats laden in recent times and I don't really get that since a lot of the analysis is amateur, done from a single camera angle, not official data from the company used by the broadcaster and questionable in many cases, particularly tackle stats. With that, I will offer opinions. The Wallabies don't need Smith to beat the Lions. I don't get where this singular focus on the breakdown is coming from. What the Wallabies need is better ball handling. If Smith brings his passing skills to the match then that's an area where Hooper and Gill can't really compare, but to me that's more the job of the No.8. It seems to me that Palu wasn't tackling or offloading enough in the Melbourne test. The Wallabies getting over the advantage line inflated individual stats but accomplished nothing because they never punched a hole in the line. They need to start offloading behind the line. If Smith can help there great, but he's got a dodgy knee and was washed up in 2009 in terms of top flight, test match rugby. To me it's like recalling Jerry Collins. I don't get it, but maybe it'll be some kind of sporting fairy tale like with Origin comebacks. If Smith is MoM, how does that affect the next game? Might be a bit of a come down.
Robbo
Guest
Great article- Nice to see someone inject some proper stats to the debate. Clearly written either by a budding sports journo or a student with way too much time on his hands... Much as I prefer Hooper over Gill for the full 80, I think the article itself proves that in this series you'd bring Gill off the bench- he's 4 times less likely to concede a penalty. Leigh Half'waytotheposts'penny will surely lick his lips at the prospect of dumb breakdown penalties late in the game. Dingo Deans evidently agrees with you Mr Miller, but if my suspicions are correct you're an arts major- best leave the number crunching to the pros ;)
Jereme Lane
Roar Guru
No definitely don't think hooper has won by any stretch I've just found gill to be equally ineffectual. One line out steal, the so called Justin Harrison moment from gill doesn't put him above hooper. Hoopers meters ran have been very impressive so far while gill has almost no running game which you need in this day and age as a back rower. Also I'm a massive fan of the pick and drive and it infuriates me that deans refuses to use it.
bennalong
Guest
Pilfering is less important than it was because it reaps penalties so frequently, which goes to the posts re Pocock (above) I'm surprised that my comments re Joubert's reffing of the breakdown produced little comment Like everyone I was impressed by Warburton at the breakdown, but I noted Joubert was willing to call it a legal steal when no time was given to place the ball, and Wallaby players arrived only ALMOST instantly making it a ruck IMO I believe in the team going foreward having the advantage. He does too........usually Having said this I am pleased George is in the team for Saturday. He's fit, he's been training with the team and he IS a dead set legend It will be a great decision I'm sure But I don't know why people think pilfering will be so successful this time around under Poite. Anyone cast a light here? So unless pilfering is the rage I'd have Hoops on the bench to add a running back, rather than pilferer Gill GO the WALLABIES ! ! !
Nick Turnbull
Guest
Thats obviously why James Horwill praised Liam Gill for his work at the breakdown and why Robbie Deans is flawed as a tactician and selector because P & D works when done well. The Lions just got bigger. Hooper got worked by Warburton who is big but not as powerful as O'Brien yet the logic is Hooper will be able to stand up to it. Its flawed. Gill is proven better in tight and better pilferer. If you think Hooper has won the breakdown this series its a position I cant support and think it only logical to play our best pilferers; Smith & Gill.
A Mac
Guest
Insightful. Well researched. Well written. Big advantage now with Smith at 7 and Warburton out. Clearly Robbie agrees too. Nice one Jamie!
Jereme Lane
Roar Guru
Interesting point nick when you say that stats aren't everything. Hooper's missed tackles are usually one off the ruck where he gets down low (as he has too against the big boys) and drives them back inside and then someone else brings them to ground. I Would be interested to know if anyone can remember an important missed tackle from hooper this series or in the super rugby because I can't. (Not including defending at inside centre) As for Liam gill I have found him to be just as ineffectual as hooper at the breakdown. In fact in game 1 I thought gill was berrick Barnes twice (because of the headgear) because he was getting smashed in the rucks before I realized Barnes was off and it was actually gill. Hoopers running game is way above what gill offers in the pick and drive stakes especially considering that deans rarely allows pick and drive unlike mckenzie who insists on it
RalphVanDee
Guest
Jamie, your insights speak of a man who plays a lot of rugby and enjoys tight yellow speedos.