The great rump of rugby views are determined by impressions. There may be a few statistics sprinkled on top, like hundreds and thousands on a Krispy Kreme donut, but the essential sweetness is still shaped by a view in real time.
When he was head coach of England, Eddie Jones once told Rugby World: “As a national coach, I’m going to games to watch players. I look to see how they react to situations, how they relate to other players, what they do off the ball.
“Most of the vision is about on the ball and I want to see what players are doing off the ball. Even if you look at the best players in the world, 90% of what they do is off the ball.”
If you pick a team based on performance off the ball, or on stats the real-time eye doesn’t usually appreciate, some of the selections will change. You may even finish up with a matchday 23 you do not recognise! Who would make the 2025 Rugby Championship team of the tournament under those rules?
Front-row
Loosehead prop: Angus Bell (Aus) starting, Ox Nche (SA) bench
Angus Bell started every one of his five TRC games off the pine but his ball-carrying load and impact was truly eye-watering.
Key Stat: Bell had the most ball-carries of any prop [averaging 14.2 per game], second among all ball-carriers behind NZ number eight Wallace Sititi, for a whopping average of 67m per game, with 4.2 tackle busts per 80 minutes of power-packed performance.
Ox Nche’s impact was hiding in plain sight, embedded within the collective scrum stats. Together with Bell, they make the perfect prop pairing.
Key stat: Nche cannot match Bell for performance in the open, but the Springboks were +10 in scrum penalties for the tournament, far ahead of everyone else. All Nche’s opponents finished in negative equity at scrum-time – especially Francisco Coria Marchetti in round six [2.7 pens conceded per game] and Tyrel Lomax in rounds three and four [2.5 per 80].
Hooker: Malcolm Marx (SA) starting, Julian Montoya (Arg) bench
Malcolm Marx was the outstanding hooker on his return to the starting berth in the national side.
Key stat: Marx was hugely physical, scoring four tries from close-range, making the most dominant tackles for a hooker [5] and pilfering three breakdown steals [third in the comp], while being part of a dominant Bokke scrum effort.
Julian Montoya was his customary consistent self, holding a skittish Argentine scrummaging effort together in the middle of the front row.
Key stat: The Pumas lineout was the most reliable in the competition running at 86%, despite having the shortest complement of receivers – a tribute to Montoya’s throwing arm. The hooker added 134 attacking cleanouts [second in the comp] and 30 breakdown defences [sixth].
Tighthead prop: Thomas du Toit (SA) starting, Wilco Louw (SA) bench
The selection of a resurgent Thomas ‘Tank’ du Toit and Bulls man Wilco Louw is a no-brainer on the other side of the scrum.
Key stats: The combined 280kg mass of Du Toit and Louw provided the platform for the Springboks’ +10 penalty count at scrum-time, with two further turnovers to be added when the Boks pushed first the All Blacks, then the Pumas clean off their own feeds. Between them the twin towers added 155 cleanouts, with both contributing an average of 27 attacking cleanouts apiece per game [fourth overall].
Second row
Second row: Fabian Holland (NZ) and Ruan Nortje (SA)
With most of the second-row pairings in flux – between them the four nations used 16 different locks over the six rounds – the obvious pairing was young All Black giant Fabian Holland and new Bokke lineout caller Ruan Nortje.
Key stats: With no lineout in the comp reaching the golden target of 90% retention, clean lineout ball was at a premium. Holland was the most reliable source of possession, contributing 34 takes [14 more than anyone else] at an average of nine per game. Between them Holland and Nortje added five lineout steals, with Holland the leading New Zealand tight forward tackler with 66 at an average of 17.5 per game [third overall] and Nortje contributing another 24 cleanouts per game.
Back row
Back row: Pieter-Steph du Toit (SA), Ardie Savea (NZ) and Jasper Wiese (SA) starting; Fraser McReight (Aus) and Pablo Matera (Arg) bench
There were many more outstanding candidates in the back row than in the middle row, so we will pick five back-rowers in our 23-man squad with the intention of promoting the inimitable Pieter-Steph du Toit up a row in the final half-hour.
What more needs to be said about PSDT and Ardie? Neither shows any signs of slowing down with age. They played all but 80 minutes of the tournament between them.
Key stats: Savea had the most carries by any forward whether playing seven or eight [58] with Du Toit only eight carries behind, and they combined for a total of 636m and 11 successful offloads. Savea was top of the charts for breakdown pilfers with eight. Du Toit added 11 lineout takes and three more steals against the throw. Wiese was a difference-maker at number eight for South Africa midway through the competition. Wiese had the most tackle busts of any forward, averaging 5.7 per game on 14 carries for an average of 72m per game. He only conceded one penalty or turnover in the three games he played.
In the last half-hour we will shift Du Toit and bring on Queenslander McReight and Pumas talisman Matera for speed and lasting impact.
Key stats: Contributed a combined 11 breakdown pilfers, with McReight a tackle zone machine boasting 86 tackles [second in the comp] and 37 combined cleanout arrivals on both sides of the ball. Sprinkle on Matera’s 10 lineout takes and 44 carries and it is a potent final quarter mix!
Half-backs
Scrum-half: Cam Roigaard (NZ) starting, Tate McDermott (Aus) bench
Scrum-half was not a strong position for TRC 2025, but the two most promising candidates were Cam Roigard [starting] and Tate McDermott [bench].
Key stats: Roigard and McDermott were the two best running scrum-halves in the tournament, with the Hurricane averaging 52m per game and the Queenslander 39m per 80 minutes. Both averaged over four tackle busts per game and Roigard was the only player to average more than two successful offloads. The kicking game offered nice contrast with McDermott kicking the most [13.5 kicks per 80], and Roigarad the fewest [5.4].
Fly-halves: Sacha Feinberg-Mngomezulu (SA) starting, Santi Carreras (Arg) bench
The two best goal-kickers from 10 by far were Sacha Feinberg-Mngomezulu and Santi Carreras, who potted 43 of their 48 goals for a stupendous 90% success rate.
Key stats: Feinberg-Mngomezulu and Carreras averaged 33 points per game, before you include their 16 average kicks out of hand, 35 total breaks or busts, and extras such as eight shared break or try assists and the Argentine’s two breakdown pilfers. So, the South African starts at 10 and Carreras covers fly-half and the back three from the bench.
Centres
Centres: Santiago Chocobares (Arg) and Jordie Barrett (NZ) starting, Joseph-Aukuso Suaalii (Aus) bench
Another hotly contested area, and finding the right synergy at 12 and 13 is the key to the success of any side. I have opted for Pumas 12 Santiago Chocobares and Jordie Barrett at 13 to start, with Joseph-Aukuso Suaalii offering x-factor and cover off the bench.
Key stats: Chocobares and Barrett were very solid in D with 85% plus tackle completion rates and sharing seven dominant hits [five of which belonged to the Kiwi]. They also combined for exactly 100 carries and 26 breaks or busts. Barrett’s kicking game is an undeniably valuable bonus [22 kicks for 612 metres]. Suaalii’s presence on the bench boosts the break/bust tally to 41 and the dominant hits to a satisfyingly round 10.
Back three
Back three: Will Jordan (NZ), Juan Cruz Mallia (Arg) and Damian Willemse (SA)
With all of the back three spots already covered by the presence of Carreras, Barrett and Suaalii, there is only space for two wings and one full-back, and they are Will Jordan and Juan Cruz Mallia out wide, with Springbok Damian Willemse at 15.
Key stats: Willemse was the only member of the back three to average over 100m gained per 80 minutes, compared to Jordan’s 76m and Mallia’s 84m. Jordan offers the burning speed and breaking ability [six breaks and 18 tackle busts] while Willemse and Mallia provide the all-rounder, extra playmaking ability with 28 kicks and 15 offloads between them.
Summary
The final balance reflects the true balance of power in 2025: nine Springboks, five players apiece from Argentina and New Zealand, and four from Australia.
A tight five weighted towards the world champions, with a mix of Argentine and Kiwi strike runners to fire the bullets and South Africa’s brand-new playmaking axis to supply the ammunition. Martians beware – planet earth, or at least the southern half of it, is ready and raring to go!
I think there was a lot of politics at Munster, whereas he was given complete control of the rugby side at Bath and he has made it such a positive environment at the club.
He’s a super intense hard working guy, there are no shortcuts with JvG, he’s a “do the work and the results will follow” type of guy and if you lose a game he says “we need to work harder and not let it happen again”.
Fairly easy to build progressive success when you create that type of environment. It’s been brilliant to see after the tough times before.
Yes it will certianly have put Rassie’s antenna on high alert! If they can get him to around 135 kilos he’ll be Test-worthy for sure.
Van der Messcht is an incredible ball carrier and the fitter they get him the better and deeper he will go into games. He’s the best space generator I have seen in ages, it takes 3 people to bring him down, so you are playing 12 vs 14 in attack.
Earl only played the last half hour, and Tom W spent most of his time understanding the real meaning of power from Van der Mescht!
See next piece!
I’ve just watched Saints vs Sarries, with WILLIS, EARL & POLLOCK all on the pitch together and I thought this was a coming of age performance from Pollock - he was exceptional. . . . I don’t get excited in just one performance but his acceleration into contact has improved this summer and he could well be the dark horse to get the starting spot at 8.
Once again a very well written article Nick. Thanks for this stats based team. As I suspected, there is a few surprises there. Any team usually chosen has some bias in it, that's why a stats based team are more accurate in deciding the best players in each position, although it isn't necessarily the best team overall should they compete. Depending on the coach and what tactics he use, the team might be different. Still….that's a mighty strong team that I doubt any team but the Boks can beat. On personal choice, there is a few players who wouldn't make my team due to bias, but via stats, there is no doubt that this team would destroy everything in front of them. It would be interesting to see what Rassie would do with a team like this. It is easy to say this player is better than that player etc. but that's not what they are getting judged on. Those who played the most minutes should be given preference like you did. It's easy racking up high stats in 2 or 3 games with few minutes played, but much harder to keep high stats and a high quality game play throughout the RC. I really can't argue on this team chosen. Very well done and a much closer description to the team of the tournament of the player's who should have made it.
There is always scope for personal choice, and ofc diff coaches will always look for players who suit their systems the best, but it is an interesting exercise [not least for me] to take that out of the equation for a moment.
Compared to the final table SA came out more clearly ahead but that matches my gut instinct in any case!
Agree, although the defence looked to be the best it has been in ARG this summer, so not having all our Lions helped. 🤣
The other aspect which you don’t mention is defence, and there Earl plays a crucial role in the system that Willis cannot fulfil.
Spot on as always NB and I agree the stats will strongly favour Earl.
I think my gut feel comes from the things that enable the stats but my gut feel from watching the games suggests Willis does more of the close quarter, heavy traffic carries around the edges to tie in the back row defenders (creating space), whereas as Earl does a lot of the 12 channel carries, or pre-contact pass on the outside shoulder or a carrier but makes longer deeper carries when he breaks the line (ie his speed makes it harder to catch, so he makes greater metres per line break.
My other gut feel is that Earl is less impactful in the first 50 mins of games where the defence is tight) and a lot more of his metres are made in that +50 min timeframe, as more gaps start to emerge (which is natural in games). It’s not a perfect theory but once Earl finds space he is exceptional at maximising those opportunities (with his pace) so they both offer something completely different to each other.
The other thing that I accept is changing is that ENG appear to be keeping ball in hand through their improving back line attack, where they are starting to attack the wider outside channels (since 55mins of the IRE game when Fin Smith came on), which does then favour Earl over Willis if our plan is to maintain that wide attack (which i think Lee Blackett will further improve).
I think there was once a time where the International game was very much about having your best 15 players on the pitch but it’s really turned into a game of balancing out the skills of the 23 over the full 80 mins, which has a lot more nuance and is more game strategy influenced, rather than individual players per se. It certainly makes selection a lot more nuanced and is much more complex than it used to be for sure. 🤣
And a shot-putter of some repute!
The big diff between us is that the stats do not support the assertion that Willis is a far better ball-carrier than Earl. Last 6N for example, he had more carries for more metres than the likes of Alldritt, Doris and Faletau, the more traditional number 8’s.
He may not look the part but he plays it, and then you add in all the bonuses at BD in and wider D. He is a major reason why England’s ‘three sevens’ B/R works. They will not cycle backwards now.
Great question NB. My thoughts are not really against the individual player per se, it’s more the collective output of the team.
I think Tom Curry (at his best) is exceptional and is potentially the best England 7 ever. I think SB is favouring the Curry’s together as the 1st choice pair, as they do at Sale, with Tom nudging over to 6.
I think Sam Underhill has a more destructive 1st line D tackle but Curry has a better all round game and the speed of his decision making and jackal is incredible. SU was used behind Ben Curry last year and I think SB favours the more mobile, expansive game, which the Curry’s do very well and their combined breakdown work is exceptional. It just felt like SB favoured the 2 x Curry’s over SU last year and whilst there is no doubt they are the best 3 English 7’s, I think ENG are at a point where they are having to develop Pepper & Pollock and give them game time, which is where Sam will potentially miss out (whilst being the better player) - that’s my fear but injuries have avoided that difficult decision so far (which is also part of the modern game).
I prefer the Mark McCall approach, of playing Willis for the attritional first 60 mins and then letting Earl loose for the final 20 (where he usually excels). I thought ENG lacked ball carriers last Autumn against SA. NZ & AUS, especially as this isn’t a strong point of Itoje’s game but Genge, LCD (Lions) & Chessum stepped up last 6 N’s (alongside Willis) which gave us a better forward platform in the tight exchanges and more variation of options to keep the opposition defence honest, as we were too predictable last Autumn, with a 2 man welcoming party waiting from the SB’s that smashed us.
It’s also why I thought SB may favour Itoje/Martin, with Chessum at 6 (which injury has prevented again), which then gives him the carriers and takes the pressure off Earl, so it’s not about Earl per se, just how effective our carries are and the difference to that balance that Willis made last year (he became our best tight exchange carrier).
FRA are exceptional, they are happy to go 2 or 3 phases in the middle of the pitch to disrupt the defence, wait for the miss match and then send through a big powerful carrier on a smaller defender, make the forward breakthrough/line break and then play off quick ball to attack out wide. There is absolute beauty in how they set up those attacks and how comfortable they are with the ball.
Hope that explains my thinking and why it’s not a case of EARL vs WILLIS because they are at opposite ends of the spectrum on style, just ironic that Mark McCall has his philosophy and SB has another (with the same players).
Sam Underhill only lost his place because of injury, at his best he is every bit as good as Tom Curry. So he goes back in.
Interested to hear why you have such an issue with Ben Earl at 8? Tom Willis would likely have stayed if he had been given any assurance he would start for England. Who wouldn’t?
He obv did not get what he wanted, so he left for France. Which is understandable, because he was in a contest with a current Lion who has the best record in the 6N of any 8 on the carry over the last two years, and who should have started all three Tests in Aussie.
I wasn’t overly impressed by Willis when he first came back to the Premiership with Sarries but I have to say Sarries have done a great job improving his fitness, S&C and he’s really improved these last 12 months and has become quite an impressive player.
The Top 14 are taking our abrasive ball carrying type of players, it can’t be too long before Underhill, Ted Hill, CC-S & Dan Du Preez get offered highly tempting contracts to take them to Top 14. It does feel like they would all fit in with the style of play they prefer and will be hard to compete financially whilst they are not designated marquee players (they will probably be offered 2x more) to go.
If they want PEPPER & POLLOCK in the squad for RWC 2027, they need to give them game time. Given this is a first time combo back row, I went for Pepper & Underhill starting (as they did play together quite a lot last year and worked well), with Pollock being used at 8 as we try and find the best combinations but there is very little between all the different options
I don’t think they will want both Pepper & Pollock starting together, as that’s a lot of inexperience against quality back row opposition, so this is more about trial and error.
Not sure if you saw SB’s interview but he basically said it was his decision to only focus on players that will be available in 2 years time, which is why WILLIS was dropped immediately, so the key outcome from this Autumn is to decide who are his favoured starting and bench options at 8 between Earl/Pollock/CC-S and to decide if Pollock is better on the sides at this level rather than 8.
Starting back row feels like a decision between Pepper or Pollock at 6 (alongside Underhill) to start and to define who are the favoured order of 8’s so we have clarity for when the Curry’s return for the 6N’s.
He may also put CC-S on the bench as lock cover and give him the option of playing 8 from the bench if both Itoje & Chessum look good to go the distance (which they normally do).
It’s going to be fascinating to see how this all evolves during the Autumn series.
Interesting column Nick. Still, stats are blind to history and the immediate context. Good servant, poor master, as the Eddie’s comment inadvertently implies.
The point of the exercise was really to find out how much overlap there was between ppl’s perceptions of the best team and what the stats revealed. Did it work?😁
with the Hurricane averaging 52m per game and the Queenslander 39m per 80 minutes
just with that comment Nick, I was wondering, do you adjust their figures so that it is averaged over an 80 minute period - i.e with McDermott mostly playing 30 at the end do you calculate a metre per minute and then multiply it out? Or did McDermott average 39 metres in each game he contributed to. Pity he was injured as he was playing very well.
Yes that is exactly what happens Ally. Averaged over an 80’ minute game, and you can do it by metres carried or metres gained [i.e. across the ad-line].
Keen to see what a year in the URC will do to Bell’s scrummaging. He is wonderful around the field but playing against those oppositions and in those conditions, might take to another level. I just hope he can stay fit and healthy. Twelve months ago we were wondering if he would ever play again after issues with his feet and a few operations to fix them.
Ulster are on the up and Bell will contribute mightily. For the the WBs he is the natural man to start at LHP if you have AAA [a work rate guy] at THP - esp as Tupou will no be out of the equation at least for a while.
I felt his career might go the same way as Corbs [Alex Corbisiero] who eventually had to reture despite many foot surgeries.😒
NB - from a stats perspective, who was NZ’s 2nd best lock behind Holland? From a purely stats perspective, who were their best pair?
Not much to choose between Scooter and Vaai P. Vaai more effective at lineout, Scooter slightly more active around the tackle.