Halfway to the 2019 Rugby World Cup

The analysis of international rugby is laden with danger. Fixtures are infrequent (on average, Tier 1 nations have played fewer than 12 games per season across 2016 and 2017), yet the traditional depiction of test matches – imbued with a deep emotional resonance – lends itself to the inference of significant meaning from single data points. A loss by New Zealand in a test match is attributed with significance that a loss by Saracens in a Premiership game is not, because of the sparsely populated environment in which it stands. For another data point on a club or franchise, it may only be necessary to wait another seven days; months can pass between internationals. The degree of light which a single game can shed on the underlying abilities of the respective teams is the same in each scenario, but the difference in the first case is that an additional data point to confirm or deny a suspected trend is usually not far away; analysis can be measured and sceptical, knowing that in a short time a little more information will become available, rather than conclusive and emphatic.

Two years into a World Cup cycle there is sufficient information available to review the performance of each team, with Tier 1 nations having played between 20 and 29 test matches each. What follows is an attempt to establish not only which teams have been successful, but also the tactical approach which they have employed in order to be successful.

NB: the following graphics display ESPN data for the ten Six Nations and Rugby Championship and teams in all test matches during 2016 and 2017. ESPN data is not available for Italy vs. South Africa (November 2016), and so this fixture has been excluded from the analysis.

When average points difference and try difference per 80 minutes in this two-year period are plotted for each team, a number of groupings emerge:

PD_TD - final

By this measure, New Zealand stand far ahead of the pack in their performances to date in this cycle. Their average points difference in 2016 was an exceptional +24.3 per 80 mins, and given their tougher schedule, spate of injuries and loss of depth to overseas clubs it was unsurprising that they took a step backward in 2017 (+18.0 per 80 mins).

England also suffered from injuries during 2017, but their average margin stayed consistent from 2016 to 2017 at +15.0 per 80 mins. An excellent series of results in five tests against Australia has helped the second-ranked nation perform markedly better against Southern Hemisphere opposition than Northern Hemisphere opposition over the first two years of the cycle, with average margins of +17.0 and +12.9 respectively. However, a potentially worrying indicator for Eddie Jones is the fact that the average number of tries they concede doubles against sides from the Southern Hemisphere (2.4 per game, vs. 1.2 against NH teams).

New Zealand’s North-South splits follow a similar pattern: they have beaten Southern Hemisphere sides by an average of 27.0 points per game, while against Northern Hemisphere that margin falls to +14.4. In contrast to England, their defensive performance is consistent across both hemispheres (NH: 1.7 tries conceded per game; SH: 1.5 per game); this drop in margin is due almost entirely to a lower number of tries per game (6.2 vs. 4.2). However, this average of 4.2 is still 0.6 per game higher than any other team managed across all fixtures.

Scotland and Ireland are the only other teams with a positive average points difference across the period. Scotland were consistent across both years, recording average margins of +5.4 and +5.0 in 2016 and 2017 respectively. Ireland, in contrast, were considerably stronger in 2017 by this measure; however, their average of + 16.5 (2016: +5.7) was boosted by three fixtures against Tier 2 opposition on their June tour with an average winning margin of 28.9 points per game.

Both South Africa and Australia – part of a group of four teams with an average margin close to zero, as noted above – suffered dispiriting defeats to Home Nation opposition in the most recent set of November fixtures. The prevailing narrative in the wake of the 2015 World Cup – the strength of the South, in contrast to the lack of quality in the North – has been reversed, and in the case of these two ailing powers of the world game this reversal is backed up with clear evidence. After two World Cup cycles (2011 and 2015 – i.e. 2008-10 and 2012-14) where both sides posted positive average margins on end-of-year tour fixtures (Australia: +8.8, +1.8; South Africa: +6.8, +6.1), both are in the negative in the 2019 cycle to date; Australia’s average points difference per 80 mins is -4.6, and South Africa’s is -5.6.

The search for an explanation for this trend often overcomplicates matters: put simply, it is a natural consequence of an environment in which the concentration of economic resources in Europe has allowed unions and clubs in the home nations to invest considerably – and successfully – in both the development of young home-grown talent and the selective recruitment of established talent from Super Rugby, a competition offering relatively little financial reward to its players.

The talent exodus from South Africa – which begins to affect the domestic system at levels below Super Rugby – has been well documented. Australia have also had to suffer the loss of 20 capped Wallabies born in 1987 or later to European contracts across the three offseasons between 2015 and 2017. All of those players would have been aged 32 or younger at the 2019 World Cup, and their loss constitutes the decimation of almost an entire generation (in sporting terms) of talent. Of those 20 players, only three continued to represent Australia in test matches (Will Genia, Quade Cooper and Kurtley Beale), and those three were also the only players of the group who have subsequently returned home to play Super Rugby. For comparison, in the three preceding offseasons (2012-2014), 12 internationals in the same relative age range (born in 1983 or younger) signed contracts with European teams, with only 2 players later returning to Super Rugby.

Southern Hemisphere talent plying its trade in European competition has been an ever-present feature of the professional era, but the continuation and acceleration of this trend – in concert with increased investment in talent development in the Home Nations in particular – appears to have reached the critical mass of players necessary to disrupt the status quo in the international game, and any analysis of international rugby is incomplete without an acknowledgement of these economic factors.

The accepted wisdom has been that the efficiency of New Zealand’s talent development system is immune to these economic factors, but it is inarguable that they too have suffered the diminutive effects of the global transfer market: there are currently 147 players who have come through the NZ domestic system (i.e. players born in NZ – or born overseas and schooled in NZ – who have played NPC or Super Rugby) playing in the Top 14, Pro14 and Premiership, roughly equivalent to another four squads’ worth of  Super Rugby-calibre players unavailable for international selection. 32 of those players are former All Blacks, and a further 15 have been capped internationally for Tier 1 Northern Hemisphere nations.

In addition, Super Rugby continues to develop along a different stylistic path to the European game, which is a closer facsimile of test match rugby than the Southern Hemisphere’s domestic competition. There is some statistical evidence that supports the common depiction of Super Rugby as more transition- than breakdown-based: per ESPN data, there are approximately 10% more carries for every ruck or maul in Super Rugby than in test rugby – likely indicative of higher offloading rates and more carries resulting from broken-play situations such as turnovers and kick returns.

This stylistic difference manifests itself in the physical requirements of the lock position in different competitions: players listed as second rows in the 2018 squads of New Zealand’s Super Rugby franchises on average weighed 3.6 kg less than their Aviva Premiership counterparts, with an average weight of 113.5kg falling considerably short of the 118-120kg range considered optimal by Steve Hansen and his coaching team for the position in test rugby. It was also evident in the ways in which the absence of key tight-five forwards Joe Moody, Owen Franks and Brodie Retallick provided the All Blacks with problems at the attacking breakdown in the second half of 2017, as their replacements struggled to clean out as efficiently and effectively. This was especially noticeable in the case of Kane Hames and Scott Barrett, with Nepo Laulala – 184cm, 116kg – proving himself capable at tighthead after injury to the 118kg Franks. (The fact that Ardie Savea has been unable to reach his desired weight and body composition has also likely been a factor in his inability to translate his game fully to international level, with the attacking breakdown one particular area where the contrast between his ability and Sam Cane’s is marked.)

The All Blacks have not yet regressed against Northern opposition in the same way as their SANZAAR colleagues; their average winning margin so far in the 2019 cycle is actually higher than their 2015 cycle average (+14.9 vs. +11.3). While this average is inflated by a 58 point victory over Italy in 2016, their overall pattern of results is relatively consistent with previous cycles. Single-digit wins against Northern Hemisphere opposition – such as the five-point victory over Scotland this past November – were far from uncommon pre-2015: 6 of 22 November victories between 2008 and 2014 came by 8 points or fewer, versus 2 of 6 across 2016-2017.

Appraisal by this simple measure thus gives a clear indication of the top tier that has begun to emerge in this cycle: the ascendant home nations (minus Wales) behind New Zealand. For comparison, these are also the four top teams ranked by winning percentage, although England (96%) and New Zealand (86%) swap between second and first place.

There are scheduled to be a number of fascinating clashes between these sides in 2018: Scotland vs. England at Murrayfield, England vs. Ireland and New Zealand at Twickenham and Ireland vs. New Zealand at the Aviva Stadium are undoubtedly the four most interesting fixtures in this year’s international calendar, and they will be made even more compelling by the fact that each of these four teams achieves their success in different ways, using contrasting tactical approaches.

The ultimate tactical goal in rugby is progression of the ball up the field, and – the obverse – the prevention of ball progression by the opposition. Whether this is done with ball in hand or via the kicking game, it is difficult for a team to sustainably score enough points if they are unable to both move the ball into the attacking half of the field and generate point-scoring opportunities effectively when in those positions. In terms of ball progression with ball in hand, there is considerable distance between New Zealand and the rest; they have also been among the best at preventing opponents’ ball progression:

Ball progression

In a similar fashion to their points difference, New Zealand’s exceptional attacking performance by this metric over the period was powered by an other-worldly 2016; they took a step back as a result of the factors outlined above in 2017, but still averaged the most metres carried (517m per 80 mins) of any team over the year. There is a small but noticeable difference when comparing their performance against North and South by this measure, averaging 553m per game vs. NH opposition compared to 600m against Southern nations. Their slight drop-off in carry effectiveness vs. NH teams (4.6m/carry, vs. 4.8m) is accentuated by the fact that they make fewer carries relative to the opposition in these games (125 carries per 80 mins vs. 113 by SH opposition; 120 vs. 128 by NH opposition).

There is a subtle but important distinction between ball-in-hand ball progression and territory that must be emphasised: this is well exemplified by the fixture between Wales and New Zealand in Cardiff in November 2017. Wales spent a considerable amount of time inside New Zealand’s half in the opening forty minutes of the game: their first half territory % per ESPN was 74%. However, this progression was achieved primarily through New Zealand’s concession of penalties – which Wales kicked to touch for an attacking lineout – and open-play kicks which were able to be reclaimed. While Wales were stationed in opposition territory, the New Zealand defence was able to effectively repel the Welsh phase attack; they averaged only 2.3m per carry across the entire 80 minutes. By contrast, Scott Williams’ first half try was an example of excellent ball-in-hand ball progression, but this score came in one phase from an attacking lineout approximately 50m from the try line, and so would likely have little, if any, impact on their territory % for the period.

It is very rare that New Zealand are outgained with ball in hand by their opposition. This has only happened in 4 of 28 games in the period: twice against France on their November tours (in 2016, they were outgained by 217m in large part due to France offloading on 17% of their carries, the highest rate allowed by the All Blacks in 2016-17), and twice in the third and final Bledisloe test of the year against Australia.

Ireland’s triumph in Chicago in November 2016 has been discussed at great length and picked apart for clues as to how teams could have success against New Zealand. However, one conclusion that was reached – that to win against the All Blacks, you have to ‘take the game to them’ in open–play – lacked any great basis in fact. Joe Schmidt’s side certainly didn’t rely upon fearless attack with ball in hand to outscore their opponents 5 tries to 4: they carries the ball 100 times for only 194m to New Zealand’s total of 514m (on 120 carries). Rather, they relied on ball progression from kicks both in open play and from penalties to generate scoring opportunities from set pieces deep in opposition territory. They were particularly cute at engineering penalty opportunities through clever work at the attacking breakdown as in the four examples below, where New Zealand tacklers are pinned in the ruck by Irish players:

This slideshow requires JavaScript.

This sort of ruck work – in addition to effective clearing beyond the ball, which both takes opposition players onto the ground and out of the game and lengthens the ruck to slow their fold – was not just evident in this fixture; it also appeared a couple of times in the win against Australia:

Following penalties awarded against New Zealand in Chicago, Ireland progressed the ball 215m upfield via kicks to touch; in open play, they progressed the ball 662m with kicks (kicking statistics from this fixture were manually tracked). In comparison to this 877m total, New Zealand kicked for only 440m – with this deficit more than offsetting their advantage with ball in hand.

Ireland were able to strike the balance between kicking enough to make significant territorial gains without – after the first try of the match – giving New Zealand’s back three too many opportunities to counter-attack. Indeed, teams have generally avoided kicking frequently to the All Blacks so far in this cycle. Their opposition average 6.1 carries for every kick they make across 2016-17:

Kicking game

New Zealand’s own kicking frequency is around average, and it is notable that three teams usually thought of as playing ‘attacking’ rugby kick the ball with the lowest frequency.

By contrast, England’s position on this graphic is one of the main reasons that the November 2018 clash between the two top-ranked teams in the world will be so tactically interesting: Eddie Jones’ side kick the ball more frequently when they have possession of the ball than any other team. While their ball-in-hand ball progression (and prevention of opponent’s progression) was middle-of-the-pack based on their period averages as displayed in the graphic above, these marks improved noticeably in 2017 (471m vs. 397m conceded). Moreover, their average ball-in-hand ball progression over the two years of the cycle to date is driven more by possession effects than by a low level of efficiency; they make clean breaks with the third-highest frequency of all teams, and average the second-most metres per carry.

Not only are New Zealand the most effective team in terms of ball-in-hand ball progression, they are by far the most efficient: their average of 578m carried per game (62% of total metres made in their fixtures) comes on approximately 50% of their fixtures’ total carries. This efficiency is powered by an average carry distance of 4.7m (2016: 5.2m; 2017: 4.3m), beating defenders more regularly than any other nation and generating a percentage of clean breaks that would be a leading mark in Super Rugby (where the top three teams in 2017 averaged between 11-12%), let alone test rugby:


Despite a drop in m/carry in their less impressive 2017 season, it is notable that their clean break % (clean breaks/total carries) remained consistent at 13% in both years. There is also a reasonable level of consistency in their NH-SH splits: their clean break % is 14% against Northern opposition, and 12% against Southern opposition.

Scotland have seen a slight uptick in their clean break % (6.6%) and average metres per carry (3.2m) so far under Gregor Townsend, but those numbers would still place them firmly at the lower end of both measures.

The fact that Joe Schmidt’s Ireland generate clean breaks with a very low frequency is a continuation of a trend seen in the previous RWC cycle. Despite averaging the third-most passes per carry across the period (1.32, behind Australia’s 1.40 and New Zealand’s 1.38), this ball movement has not punched holes in opposition defences to the same degree as those two sides. (Their approach against Australia in November 2016 is a good example of this: an average of 1.35 passes per carry, but with combinations of 3 passes or more on only 12% of phases.)  Instead, their attack is built on the retention of possession, and the avoidance of turnovers and penalties: they average 54% of all carries made in their games (the highest of the top 4 teams by 3%), and concede fewer turnovers (12.1 per 80 mins) than any other team. Ireland are not visible on the graphic below as their averages for turnovers conceded and opposition turnovers conceded are exactly equal to England’s:


As Ireland have had a much higher share of total carries than England (54% vs. 47%), their rate of turnover concession is in all likelihood more impressive. This is a clear limitation on using simple per-80-minute averages; ideally, a rate of turnovers relative to a team’s total number of carries, rucks or phases could be calculated, but this is difficult and potentially misleading without access to more detailed data on types of possessions and sources of turnovers.

It is notable that the two teams with more passes per carry than Ireland both concede significantly high numbers of turnovers per game on an average number of carries per game markedly lower than Ireland’s. This adds more context to Joe Schmidt’s attacking approach, as they move the ball in such a way that limits the downside risk of frequent passes; however, this is offset by doing it in a fashion that also limits some potential upside (i.e. the generation of clean breaks).

In addition to the low number of turnovers per game which Ireland concede, they are also extremely effectively at limiting their concession of penalties:


The same caveat on the use of simple 80-minute averages applies to penalty statistics, and Ireland’s possession-heavy style was likely a key factor in their low average figure: per World Rugby’s official Statistical Report for RWC 2015, 70% of all penalties at the tournament were conceded by the defending side.

On the defensive side of the ball – where they have been coached by Andy Farrell since June 2016 – Ireland have also been very strong, conceding on average the third-lowest number of points per game and the lowest rate of clean breaks conceded:


Much has been made of the success which teams with Farrell-led defences have had against New Zealand over the course of his coaching career, but – as noted above – Ireland’s historic victory in 2016 was not a result of preventing the All Blacks’ ball progression; across two games against New Zealand, their rate of clean breaks conceded and average metres conceded per carry were considerably above their period average (8.5% and 4.4m).

Against Farrell’s Lions these figures fall to 6.0% and 3.3m respectively, and New Zealand’s offload % falls to 6% versus an average of 11% for the two-year period – data which further indicates how effective this defence was as the touring side sneaked a series draw. However, it is important to note that New Zealand’s tactical choices in the second test after going down to fourteen men likely skew these figures downward. The All Blacks averaged only 3.0 carries per kick (vs. 5.5 across 2016-17, and 5.9 across the first and third tests) and 1.10 passes per carry (vs. 1.38 across 2016-17, and 1.30 across the first and third tests), clear indications of a more conservative gameplan with more kicking in open play and an increased focus on narrow carries off 9. This approach was undoubtedly appropriate for the game situation and had proved successful until the Lions’ late surge, but significantly reduced the frequency with which New Zealand made clean breaks (1 in 90 carries – i.e. 1.1%).

Nevertheless, the ability of Farrell’s defences to bring New Zealand’s attacking effectiveness down to a level in line with test match averages is very impressive. The pre-tour familiarity of a large proportion of the eventual test match squad with his coaching methods was likely an important factor in the Lions outperforming what was expected of a side without much pre-tour preparation time, and with Joe Schmidt’s Ireland he has been able to quickly and clearly implement his preferred strategy.

As the graphic above shows, England’s ability to limit clean breaks and metres per carry under Paul Gustard are much closer to average than Ireland’s under their former defence coach. However, they still rank second in points conceded per game (17.3), and the sample is large enough to conclude with some conviction that they have found a way to be effective on the whole while not excelling in this specific regard.

Regardless, it is interesting to consider where England would currently stand if the RFU had not decided to dispense with Farrell and head coach Stuart Lancaster after the failed 2015 World Cup campaign; the former’s success in the wake of this has been highlighted above, while the influence of the latter in his new role as Senior Coach at Leinster has been widely heralded as extremely positive for the province. It is possible that allowing that coaching team the opportunity to learn from its mistakes may have yielded a more coherent and developed product at the halfway point of their second World Cup cycle together than that produced by Eddie Jones – who pared back England’s game drastically in 2016 – two years into the job. The question of how much of the team’s progression over the past two years can be attributed to Jones’ coaching, rather than to the continued development of an excellent generation of players first blooded under Lancaster’s watch into their primes, is perhaps one which ought to be treated with a certain level of scepticism.

In summary:

  • England, New Zealand, Ireland and Scotland – in that order – are the top four nations by overall winning percentage in the two-year period. They are also the top four teams by both points difference and try difference per game, although New Zealand are considerably ahead of England by both measures. At this point it is worth noting that the All Blacks have played an even split of home and away fixtures, while England have played 13 games at Twickenham versus 10 away from home.
  • South Africa and Australia – both World Cup semi-finalists in 2015 – have both fallen back to the pack, and form part of a group of four teams (along with France and Wales) with an average points difference around zero, while fellow semi-finalists Argentina now languish towards the bottom of Tier 1 (along with Italy). Given the patterns of player movement experienced by Australian rugby in particular since the 2015 tournament, it is hard to look past economic migration as a primary cause of this trend. New Zealand have also been deprived of a significant amount of talent as a result of this, but have so far not been affected as negatively at international level; their continued excellence is even more impressive in light of this.
  • New Zealand’s attacking performance indicators are exceptional in terms of both efficiency and effectiveness; however, Ireland and the Lions have been able to limit this performance somewhat (while still being outscored 12 tries to 9 across 5 games).
  • The Lions’ second test victory – where they scored two tries in the final quarter against fourteen men – is somewhat of an anomaly in comparison to other fixtures in which the Home Nations have exerted pressure on the All Blacks. As noted above, New Zealand average a significantly lower share of possession against Northern Hemisphere sides; Ireland (in November 2016), Wales and Scotland (both in November 2017) were able to starve them of ball by reclaiming kicks and forcing the concession of penalties, and with an approach based on holding on to possession.
  • This sort of approach has been the hallmark of Ireland’s play in the period: an attritional gameplan with ball in hand focused on limiting penalties and turnovers. To this they have combined a defence which limits their opponents’ attacking efficiency.
  • Scotland’s defence has also been effective at limiting clean breaks and metres per carry, but their attacking efficiency has been at the lower end of the spectrum; this has seen a slight uptick under Gregor Townsend in 2017, and it will be interesting to see if this trend continues during 2018.
  • England’s performance over the period remained consistent from year to year, while showing some development in their tactical approach: they kicked the ball slightly less in 2017 (although still more frequently than any other team), and their ball-in-hand ball progression numbers improved. However, the fact that their average number of tries conceded doubles against Southern Hemisphere opposition – and that some of their defensive performance indicators in areas of attack where New Zealand thrive are not as robust as their competitors in the top tier – may give them some cause for concern going into their long-awaited clash with the All Blacks at Twickenham in November.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s