What age do rugby league players peak statistically?

Now that the State of Origin distraction is over, we’re reaching a point in the NRL season where finals hopes are well and truly dashed for some teams, and they turn their attention full time to next season and player recruitment or retention.

As usual in the NRL media cycle, there’s been plenty of speculation about players moving on and teams trying to shore up their weaknesses. There was even a leaked report on how a 17th team would affect all aspects of the league including recruitment. Answer: it would mean they have to work harder. Friend of the site Liam at PythagoNRL has broken down this brilliantly here and is worth a read.

Circling back to recruitment, one of the bigger stories last week was the Tigers looking to “go all in” to sign 29 year old Dale Finucane, who would be 30 by the time the 2021 finals series arrives. This comes a year after bringing in 32 year old James Tamou and 25 year old Joe Ofahengaue to offset the loss of 25 year old Josh Aloiai.

In another move, across the ditch, the Warriors let 26 year old Ken Maumalo go, replacing him with almost four years (!) of soon to be 26 year old Dallin Watene-Zelezniak.

Given the ages of some of these players, this led me to look into some numbers to determine at what age do rugby league players actually peak statistically.

Is there a point in a players career where they are no longer developing statistically and from that point on they are what they are? Just how long should a team expect a 26 year old winger to be productive? Would or should this have any impact on recruitment?

There’s a tendency for clubs to assume players are finished products around 20-22, with players either leaving the game, heading to England or fighting for their career in second tier competitions. For some that is true, but for most positions it appears that their development peak may be later.

To undertake this analysis, we’re going to look at a group of running or attacking metrics by age to determine at what point do players hit their apex as a player. I’ve grouped games played by all players from 2014-2021 into one year age buckets from 18-34, based on the age when a game was played, rather than their starting age during the season.

For example, Latrell Mitchell turned 24 on June 16. His games before that date from 2021 would be in the 23 year old age bucket, whilst games after that date would be in the 24 year old age bucket.

I’ve also put in a minimum game limit of 30 appearances for each age bracket. In doing so I checked how these numbers would change if I dropped the threshold to 10 games or increased it to 80. Interestingly enough, it wasn’t until I dropped below 10 or moved over 90 that any of the numbers changed even slightly as you were capturing fewer players. That said we’re keeping it at thirty because that’s a robust enough sample size to keep things significant and ensure a wide spread of players captured.

And since most positions on a rugby league field have different roles, we’re going to use different sets of metrics for them. For middle forwards we’ll mostly look at runs and metres. For edge forwards the stats are runs, line breaks, and tackle busts. When looking at centres and wings, the stats are runs, line breaks and tries and finally for fullbacks and halves runs and playmaking (line break and try assists).

Keep in mind this is not necessarily meant to show a players physical peak or their high point from an impact perspective, just their statistical peak when their numbers stop improving. Many players will continue to be extremely effective players past these age groups, all that changes is how often they can do it or being able to pick the right spot to do so. Knowing is half the battle.

The point is that by pinpointing the age when players in certain positions peak, you can understand when the development of a player is likely to end, and they’ve moved onto a maturity phase. At a certain age, a player is what they will be and there’s no point expecting or hoping for further improvement. And that age doesn’t always line up with the point when players are moved on from clubs.

It is also worth keeping in mind that the later year players (30+) are more likely to be of higher quality and potentially mudding the stats in those age brackets. Players like Brett Morris, Cameron Smith and Paul Gallen played well into their 30s and the declines for ages over 30 are clouded by those players.

Let’s begin with middle forwards.

Middle forwards

Looking across the most of these statistics middle forwards are peaking at 25-27. At those ages, they’re playing the most minutes, taking the most runs and completing the most tackles. This makes sense as middles mature and move from interchange players to starting forwards. The drop offs from 27 onwards aren’t too severe either until you hit 30.

Middles do tend to start off more as damaging runners of the ball, with higher metres per run and tackle bust % (percentage of runs with a broken tackle), which starts to decline at 23. This is probably due to an increased workload and needing to manage effort with intensity, shown by their run rate (how frequently they complete a run) reaching its highest between 25-27.

Some of this is natural as they play increased minutes as they age. Yet some of the metrics shown like metres per run, Run % and Tackle Bust % aren’t pure volume statistics and aren’t affected by minutes played. Which again indicates why they would decline as minutes increase. It’s hard for players to sustain the same effort in 50 minutes that they were playing at 30 minutes.

An area that does improve later in their careers is the ability to offload. As middles hit 26, they offload the ball more than average until 30 when that drops off to below average.

One thing that doesn’t change much is defensively. I didn’t include the advanced statistic Tackle Rate for middles because it doesn’t change much – around 25-26% each year. So, whilst the number of tackles made per game ebbs and flows as minutes do, the rate at which players are making tackles is steady as players age.

Locks have slightly different profile to props and are much more consistent over time. Like front rowers, their runs per metre drops once they hit their mid-20s, as does their tackle bust %. However, they are running the ball at a lower rate from 27 years old and they’re less likely to be ball players until very late in their careers. Again, keep in mind that these post 30 buckets are affected by a small number of players so those 32-34 offload rates are skewed by likely one or two players.

Penrith recently extended James Fisher-Harris for another four years, and with him currently sitting at 25 it’s the perfect time to lock up one of the best middle forwards in the game as he hits his peak statistical years.

Returning to the Finucane and Tamou examples at Wests, whilst they’re reaching the tail end of their career, there is value in bringing in experienced heads to also assist in developing younger middles like Thomas Mikaele, Stefano Utoikamanu and Alex Seyfarth. Tamou had bucked the trend somewhat in his last year at Penrith, but players finding another level at 30 is the exception not the norm.

It looks though that there may be a market inefficiency in early 20s middles, who may have been discarded earlier by teams expecting them to progress quicker. Instead of bringing in late career middles, it might be better to cast a wider and deeper net looking for young players discarded too soon.

Not only is their output likely to increase as they age, but they would also provide similar production to older players at a cheaper salary. If you can find a bunch of young low cost, high work rate middles, that leaves a lot of salary cap space left to splash around on more impactful players.

Second row

Second rowers share some similarities with middle forwards in that their running involvement inreases in their early 20s and peaks at 26 years old. However their metres per run metric is slightly more consistent over their career and only decline marginally in their late 20s.

Backrowers also tend to be more dynamic earlier in their careers, with line breaks per game and percentage of runs with a tackle bust also hitting their high points in their early 20s. Later in careers, they’re running the ball less and take on more of a defensive role.


Running involvement for a dummy half peaks between 24-27, as does metres per run which maxes at 29 years old. As you would expect, hookers run the ball less frequently as they pass 27, not only on an overall basis but at a lower rate as well.

The danger of their running game comes a little earlier, from 22-26, when they peak in line breaks and line break assists per game.

Defensively, hookers are making a similar number of tackles no matter how old they are, at about 35 per game. Minutes played is similarly very consistent and actually increases post 30 (mostly due to Cameron Smith), and for this reason I’ve not included it in the chart above.


The halves are a position where age plays a big part in statistical peaks, although it is quite different to when forwards peak and also tends to develop key skills later.

Like other positions, running involvement maxes out at 26 and drops sharply from there. This is probably an indication of players maturing and knowing when to run the ball than any physical decline. Metres per run peaks at 24, whilst runs and run rate is relatively steady until 27 when it declines below average, and it is no coincidence that line breaks per game drops at that point as well. It is in these running statistics where the difference between a five eighth and a halfback occurs, with #6s lasting a year or two longer than 7s for their running peak.

The other big change for halves is that around age 24, they start to really hone their playmaking skills. Passes per game increases as they get more involved in attack, leading to big jumps in try and line break assists per game. It isn’t until they hit their early 30s that these numbers start to tail off. It’s another reason why the Warriors bringing Shaun Johnson home to finish his career makes sense, he’s been in sublime form as a playmaker even if his physical skills have started to decline.

Another reminder that when you’re writing off players in their early 20s, not all of them have finished developing or found their niche.

Outside backs

Centres have similar production as they age, peaking at 26. Their running statistics are very similar in late teens to mid 20s, with younger players having a slightly higher metres per run and older players running the ball slightly more.

The biggest drop occurs in attacking stats once they hit 27 when tries, line breaks and line break assists per game all fall off at a fast rate. As mentioned previously, this doesn’t mean that players \are leaving less of an impression on a game, they’re just doing so less frequently.

Wingers have a similar profile to centres but appear to peak a little later at around 27-28, as their run rate increases as they near 30 years old. Their try scoring rate doesn’t drop below average until 29 either, a year later than centres. But as we saw from Brett Morris this season, they can still be very effective into their 30s in the right situation.  

Returning to the Warriors example, they may have been right to let Maumalo walk and replacing him Watene-Zelezniak as they output similar numbers. However, Watene-Zelezniak will be almost 30 by the time his deal expires, and the above chart shows that his productivity is likely to decline over the coming seasons.


Fullbacks have an age profile that overlaps between outside backs and halves. Similar to outside backs, their running involvement and output is reasonably consistent from their early 20s until their late 20s, only starting to decline around 28 years of age.

Their attacking impact is more visible early in their careers, with tries and line breaks per game all climaxing before they hit 24. And just like halves, once this drop in pure attacking statistics doesn’t mean a player has hit their peak, it just highlights when they transition from an attacking weapon running the ball to more of a third playmaker role supporting both halves.

How Melbourne and Penrith differ on set restarts – NRL Round 15 stats and trends

After last week’s dalliance with the trivial subject of what brand of boots NRL players are wearing, it’s time for the Eye Test to get back to it’s bread and butter. That would be analysing set restarts and who is gaming the system to their advantage. The conclusion that the good teams don’t concede set restarts does hold true but with one very large exception.

The dominance of Melbourne and Penrith has been well documented this season, with both teams scoring and conceding points at historic rates and seeing some incredible momentum swings within games.

But one area where they differ is how often they’re conceding set restarts.

Last season I used a chart plotting set restart difference (awarded minus conceded) against margin after accidentally discovering that there was a negative relationship between set restarts and margin. That is, the more set restarts you conceded, the better your winning margin was, and the top four from 2020 all had a negative set restart difference.

Something changed this season though, as only one of the top four has a negative net set restart difference. The chart for plotting net set restarts against margin for 2021 is below, split into four quadrants – conceding and winning, not conceding and winning, conceding and losing, and finally not conceding and losing.

As mentioned, one team stands out from the rest of the league.

Back to set restarts, it is very clear to see that Melbourne have taken over the mantle of the biggest set restart offenders from the Panthers. The Storm have a -23 set restart difference, which clearly isn’t affecting their play as they’re scoring 300 points more than their opponents this season, and hit a new record in the process of trouncing the Tigers on Saturday evening. Penrith aren’t far behind either.

That -23 net set restart difference by Melbounre is by far the biggest disparity in the NRL, with the next worst being the Dragons at -18.

On the other end of the scale, the Panthers have a +18 set restart difference, quite a swing from 2020 when they were practically committing assault in the ruck and slowing down every play the ball. Their +18 only trails Parramatta (+22) for best set restart difference in the NRL this season. Looking at this you’d conclude that the theory that Penrith receive the most help from set restarts is true, and to an extent it is but not by receiving them. We’ll get to this more shortly.

You can also see from this chart that only six teams have a positive margin this season (those above the centre line), once again highlighting how lop sided the 2021 season has been. The bad teams have always been bad, but the rule changes have just widened the gap between the haves and the have nots. It’s not a binary thing, it’s a combination of events causing this seasons results.

Does this trend continue if you look at net penalties (awarded minus conceded)? The same chart showing penalties instead of set restarts plotted against margin is below.

Penrith still sit towards the top of the NRL in net penalties, second at +20 and only trailing Souths (+27). Meanwhile Melbourne is decidedly mid table here, with a penalty difference of -6, and one of just two teams with a positive margin and negative penalty difference, with the other being notorious penalty conceders the Sydney Roosters.

Up to now we’ve been looking at net set restarts and penalties. What about checking where Melbourne and Penrith rank for average infractions awarded and conceded?

The chart below shows the average set restarts awarded (blue) and conceded (orange) per game for each NRL club this season,with clubs sorted by their ladder position.

Again, the perception that the Panthers are awarded the most restarts doesn’t hold up, with Penrith sitting mid table for set restarts awarded. Where Penrith does benefit is that they are conceding the second fewest set restarts in the NRL this season at just 3.0 per game, only trailing the Sharks (2.8).

On the other hand, Melbourne is conceding the equal second most set restarts in the league, at 4.5 per game, tied with the Dragons and only trailing Canterbury (5.1).

When you combine that with the Storm receiving the second fewest set restarts this season at 3.0, it’s easy to see why their set restart difference is so large. It also makes their dominance this season even more impressive, considering the amount of possession and field position they’re yielding.

Again, does the same hold true for penalties? Below is the same chart as above, but substituting penalties awarded and conceded for set restarts.

For Penrith, they’re receiving the most penalties per game this season at 5.6, slightly ahead of South Sydney (5.57). That’s nearly two full penalties more per game than Melbourne are being awarded, and this may be where the perception that Penrith are benefitting from “leg ups” comes in.

Clearly being on the wrong end of set restart and penalty counts isn’t affecting the Storm, who are backing their defensive discipline to combat giving away extra possession. Penrith are riding the momentum wave of increased possession from set restarts and field position from extra penalties. Neither approach is better than the other, and it’s looking like the only outcome of this season is to see them collide at the end.

Cleary doing more with less

Talking about the excellent season Nathan Cleary is having isn’t any great revelation. Friend of the site Jason Oliver from the wonderful Rugby League Writers website has been demonstrating his greatness all season, most recently in his Round 15 Repeat Set post.

As a quick aside, if you’re not reading and subscribing to Rugby League Writers, you’re missing out on the best on field analysis of the NRL anywhere. Jason and Oscar deliver amazing content daily and offer a free newsletter as well as very affordable subscription option at just $5 a month.

Back on the topic of Cleary, unsurprisingly he leads the NRL in Net Points Responsible For (NPRF) this season, a statistic I use to assess playmakers involvement in points scoring (and conceding). The leader board for 2021 is below.

What’s also not surprising, given the rule changes and more ball in play with the introduction of the set restart rule last year, is that he’s also leading the NPRF table from 2014-2021 and holds third spot as well for his 2020 season, only split by Tom Trbjoevic’s incredible year.

The thing that stands out for me is that Cleary is averaging three fewer possessions than 2020, 71.5 down from 74.4 a year ago. It’s possibly a sign that Cleary is maturing as a playmaker, knowing when to inject himself, and also a nod to the continued improvement of Jarome Luai.

Despite the slight drop in possessions, he’s doubled his try scoring tally in just 12 games versus 18 games in 2020 and should easily pass his try assist and try contribution totals if he plays the rest of the season.

It’s also worth noting just how many players from 2021 are sitting at the top of the 2014-2021 leaderboard. Eight of the top 20 and three of the top five NPRF seasons have come in 2021, and four of the top 5 have come under the set restart era.

The top three NPRF seasons are from 2020-21, and the top 2 are almost two points per game higher than any other sason. Again it’s another data point about how the massive one sided scores this season are skewing statistics.

Set restart and referee update

Looking at set restarts and penalties awarded in Round 15, the crackdown seems to be well and truly over, other than the odd ridiculous sin binning. After peaking at 12 penalties a game in Round 12, we’re down to near pre crackdown levels with just 9.3 penalties called per game in Round 15.

The drop has been consistent across both halves for penalties, while nothing much has changed for set restarts since they’re rarely awarded in the second half anyway.           

As for referees, below is the breakdown of per game averages for set restarts and penalties awarded by official up to Round 15, sorted by the number of set restarts awarded.

Matt Noyen has only called two games this season, so his numbers don’t really mean much. Grant Atkins leads the way with nearly 9 set restarts called per game, with Chris Butler not far behind at 8.5. Last years #1, Adam Gee, has softened a bit this season and is only calling 8.2 per contest, the only other referee above 8. He’s still the king of first halves though, as you will see below.

As regular readers will know, one of the things I go on about regularly is the discrepancy between first and second half set restarts. Usually, I’d post averages by referee to show the gap, but this week I’m going to use raw totals to draw attention to the differences between halves, as shown below.

Here you can plainly see that there’s amazing consistency between halves in penalties awarded, but there’s huge variances in how many set restarts are awarded in second halves.

Another thing worth noting is that Grant Atkins has blown more second half penalties (86) than total penalties by Peter Gough (78) or Chris Butler (85).

The most inconsequential Eye Test statistics ever? NRL Round 14 2021 boot brand and colour analysis

If there’s one word to describe the period during State of Origin where NRL clubs are missing multiple players due to representative duties, it’s illegitimate inconsequential. Results don’t have much meaning, some fanbases have been telling me.

Given the lack of significance of on field action, this week is the perfect time for what may be the most inconsequential set of statistics on the Eye Test today. The Rugby League Eye Test was founded on inconsequential statistics. It’s right there under the site name, I wouldn’t lie to you, would I?

For Round 14 we’re looking at the distribution of colours of brands of boots worn by NRL players. I warned you it was inconsequential. After seeing the fantastic player exclusive designs for indigenous round recently, it had me thinking about what boot brands were the most popular among NRL players and who had the highest share.

For Round 14 we’re looking at the distribution of colours of brands of boots worn by NRL players. I warned you it was inconsequential. After seeing the fantastic player exclusive designs across multiple brands for indigenous round recently, it had me thinking about what boot brands were the most popular among NRL players and who had the highest share.

First up, the methodology. I went through every game from Round 14  and coded every player for the brand and general colourway of their boots. In some cases, I was able to get the 18th man for teams, whether they played or not, such as seeing Kyle Flanagan in the Bulldogs dressing room sporting a pair of black Adidas.

For colourways I’ve grouped them as predominately white, predominately black, and other. If I had unlimited time, I’d have put the actual majority colour down in the other field, but given they’re mostly tied to what colourway the manufacturer is releasing this season – Nike has some pink and orange models, while Asics has blue and neon green, for example – it wouldn’t be much more of an extension of brand of boot. Unfortunately for the outraged Boomers who read this site and wanted to know how many players you need to complain about because they wear pink boots, you might be disappointed.

Finally, a general disclaimer, there may (and probably are) mistakes in this data collection. The largest issue was trying to spot the brand of all black boots during day games as cameras adjusted between light and shaded parts of the pitch.

On to the brands. There were 277 players whose boots I could identify in Round 14, with six brands on field during Round 14 – Adidas, Asics, Puma, Nike, X-Blades and Concave.

Of those 277, just three players wore something other than the big four brands – Ben Hunt and Jake Granville in Concave, and Luciano Leilua in X-Blades.That means that those big four brands had 99% share of players last weekend. Here’s the share of NRL players by brand for Round 14.

Asics takes first place with nearly 43% of players wearing their boots, ahead of Nike (26%), Puma (19%) and Adidas (12%). No sign of New Balance or Under Armor, not that I expected them to have much of a showing.

As mentioned above, X-Blades had just the one player on field, Luciano Leilua. They do claim to have five players on board currently, with Luciano’s brother Joseph, Ben Hunt, Jake Trbojevic and Tyson Frizell also listed. Hunt as noted above wore Concave against the Bulldogs, Trbojevic wore Asics, which indicates that page might be out of date.

From a larger brand perspective, how does the NRL spread of brands compare globally? I’m already time poor, so manually capturing Super League data is out of the question. Luckily, French football site Footpack did a study in March-April of 2020 of 2,500+ football players to see which brands were the most popular. Obviously, there are differences between football and rugby league, but these brands play in the same place, and you would think there would be some similarities.

Across the five big European football leagues, Nike takes the top spot in all of them. Their share ranges from 56% in La Liga (Spain) down to 47% in Ligue 1 (France). Adidas was generally second with about 37-40% share, with Puma third in single figures. Every other brand sat under 1%, including Asics, Under Armor and New Balance.

90% of football players in those five leagues wore Nike or Adidas, and 98% wore Nike, Adidas, or Puma. The NRL comparatively isn’t as strong for the swoosh and three stripes, with only 38% share. If you add in the cat, those three brands add up to 57%. The big difference being the dominance of Asics here, with 43% share compared to less than 1% in Europe. That’s some sort of market inefficiency.

There are a few trends though when you delve into these numbers a bit closer. When breaking down players by their position, forwards still overwhelmingly prefer Asics (57%), while backs tend to wear Nike (41%). Only 25% of backs wear Asics, while just 13% of forwards wear Nike.

Puma has a much more even split, with 16% of forwards and 20% of backs wearing their boots. Adidas is also more likely to be worn by forwards, with 17% wearing the three stripes opposed to just 8% of backs.

Looking at colour distribution, it was a clear victory for predominately white boots at 51%, ahead of Other (36%) and traditional black boots being worn by just 13% of the NRL.

If you split out colourways by position as well, there’s not as much variation between forwards and backs. Both groups sat around 50% of players wearing white, while forwards were slightly more likely to wear majority black boots, and backs had a higher tendency to wear a non-traditional coloured boot.  

Does this data look any different by team? Here’s the brand share breakdown by NRL clubs for Round 14.

Asics takes the top spot for most clubs, with as high as 60% of players at the Warriors. Only five clubs have the majority of their players wearing a brand other than Asics. South Sydney and Melbourne players are more likely to be wearing Puma, while the Gold Coast, St George, and Cronulla wear Nike more than any other brand.

Adidas has the higher affinity with Bulldogs players, where they have a 29% share of players, the highest in the league. They’ve got a bit of work to do with the Gold Coast, Parramatta, and the Warriors, where no player was sporting the three stripes in Round 14.

Looking at colour distribution by team, most are wearing white with a few exceptions.

Penrith have just a quarter of their team wearing white boots, with South Sydney, Wests Tigers, Canterbury, and Canberra the only other clubs with less than 50% wearing white. The Rabbitoh and Titans are the only clubs who have the majority of players wearing something other than predominately white or black boots.

The final thing to look is if age affects boot brand or colour choice?

For brand, Asics looks to be less popular as players age, with Puma a strong preference for players in their late 20s. Nike has a stronger connection with younger players than Puma though.

For colour choice, white tends to be a more popular pick for those in their late 20s, while black interestingly declines as players age, which would go against people becoming more traditional and conservative as they age.

First half 2021 NRL advanced statistic leaders

The NRL season has hit the slow period of the season to accomodate for State of Origin, meaning it’s a good time to continue the Eye Test’s review of the first half of the season. Last week we looked at the decline of 80-minute players. This week we’re diving into the advanced statistic leaders for the season up to Round 12.

If you’re unfamiliar with the advanced stats used here at the Eye Test, I’d recommend checking out the glossary page on the site to get up to speed. If you want a longer run down, there are longer posts on Tackle %, Run % and Involvement Rate that talk about how they are calculated and how they should be used.

If you want an even quicker rundown than the Glossary page, here’s the cliff notes. Tackle % and Run % show the % of completed tackles or run made by a player whilst on field, adjusted for minutes played and possession. Involvement Rate combines both Tackle % and Run %, and all three are “work rate” metrics to gauge the contributions of players (typically middle forwards) who don’t play big minutes.

The average rate for each of these metrics can be seen in the chart below. Generally middle forwards (front row, hooker, lock and most interchagne players) make tackles at a rate in the 25% range, have a run rate in the 10-12% range and their Involvement Rate sits between 14-19%. As you move further out to the edges and outside back positions those numbers continue to drop as they are doing less of the grunt work.

One thing worth mentioning is that the Melbourne Storm use a similar metric for “work rate markers”, calculating runs, tackles, supports and kick pressure on a per minute basis. Great minds etc…

Net Points Responsible For (NPRF) is a metric to show the contributions of players beyond the typical try and try assist statistics. And finally, Error Rate is a measurement to identify players who are committing errors more frequently.

For the first three advanced stats, I’ve implemented a minimum of three games played and 120 minutes (playing at least three full halves) to filter out any players with a low sample size. For Net Points Responsible For, I’ve increased the threshold to give games played, as one or two blowouts can skew the data. For Error Rate, the limits are four errors committed and three games played, again to remove any small sample biases.

These statistics are for the first half of the season (Rounds 1-12), and don’t include the recently completed Round 13. And as usual I’m using Fox Sports NRL statistics.

Now the formalities are out of the way, let’s move on to look at the season leaders for the first half of

Tackle % leaders

North Queensland forward Reuben Cotter leads the way with a tackle % of 34.85% in his four games this season. Unfortunately, as he is out for a lengthy period due to Lisfranc surgery and probably won’t make the end of year minute qualification, which is usually between 150-250 minutes played, and is something I gradually increase as the season progresses.

Still his tackle % is almost 10% higher than the average middle forward, with props (25.1%), locks (25.1%), interchange (24.9%) and hookers (25.9%) all sitting in the 25% range, or completing a tackle on one in four possessions.

Coming in behind Cotter is the Sharks interchange forward Billy Magoulias with a tackle rate of 34.0%, followed by promising Tigers middle Alex Seyfarth at 32.74%.

Eye Test first ballot Hall of Famer Daniel Alvaro comes in at 7th with a rate of 30.87%, his seventh season with a tackle rate over 30%, meaning he completes a tackle in three out of every ten opponent plays whilst on the field. Alvaro owns three of the top four end of season tackle rates and can usually be spotted in the top 10-20 players on a weekly basis.

An interesting thing to note is that there are only seven players with a tackle rate of 30% or above. Last season by Round 12 there were 21. In 2019 there were 26. Is this another data point suggesting fatigue is an issue?  

Run % leaders

Penrith’s middle forward machine Spencer Leniu of Penrith takes first place with a run rate of 17.11%, indicating he is completing a run on nearly two out of every five plays for the Panthers. As noted above, the average front rower Run Rate is 12.6%, with Leniu nearly 50% above that number at over 17%.

Leniu was on fire early in the season getting through a huge amount of work in his limited minutes, ranking 1st, 1st, 6th, 80th and 1st in his first five rounds of the season. He’s also the only player with a run rate above 17% this season.

Second place is taken by Roosters prop Jared Warea-Hargreaves at 16.26%, with the Tigers Zane Musgrove rounding out the top three at 15.91%, ahead of former Tiger Josh Aloiai at 15.28%

This year we have seven players with a run rate above 15%, and unlike Tackle % that number is in the same range as 2020 (eight players with 15% or more) and 2019 (five).

Involvement Rate leaders

Given that Involvement Rate is a combination of Tackle % and Run %, it’s no surprise that Reuben Cotter sits at the top of the Involvement Rate chart, with a rate of 22.75%. This means he is either completing a run or tackle on more than one in five plays whilst on the field. The average middle forward has an involvement rate in the 14-19% range, with Cotter sitting at least 3% higher than front rowers and 4% higher than locks.

Second place goes to the Raiders Corey Horsburgh at 22.17%, with that man again Daniel Alvaro snaring third place at 21.96%.

This season there are 20 players with an Involvement Rate of 20% or higher, compared with 29 in 2020 and 48(!) in 2019. To me this is a clear indication that fatigue is having a huge impact on the game.

When looking at 80-minute numbers, the totals and averages may be the same or very similar, which could indicate there’s no change in the speed or pace of the game. However, if fewer players are completing a run or tackle at a high rate, that would indicate that they’re unable to keep up a high workload over the similar periods of time.

Remember that these advanced statistics are not only minute adjusted, but possession adjusted. This means that they are comparable regardless of any raw increase or decrease in minutes played or ball in play.

Net Points Responsible For leaders

Net Points Responsible For (NPRF) is a metric for a player’s overall contribution to a team’s scoring or defense. It is not a measure of overall performance, since it only uses points scored, contributed towards, or allowed. If you do want to put a number on a players overall production, Liam at PythgoNRL has a fantastic overall player performance metric.

Returning to NPRF, this the way it is calculated. Each score by a player is valued as it is on the scoreboard (try – 4, goal – 2, field goal – 1 or 2), plus 4 points for every try assist and try contribution. Four points are removed for every try cause a player concedes as a way of quantifying their defensive contribution. This total is then divided by the number of games played to get a plus or minus total points per game that a player is “responsible” for.

It is a great way of valuing a playmakers influence if they aren’t directly scoring tries or throwing passes that are deemed try assists. By adding in Try Contributions, something like the famed a “hockey assist” (the pass that least to the pass that sets up a score), you can more accurately judge the impact of a player whose involvement may not show up in regular statistics.

With that out of the way, here’s the top players for 2021 thus far.

No one should be surprised with Nathan Cleary topping the list, and at 11.82 NPRF/game, he’s directly adding almost 12 net points per game to the Panthers this season. To put that in perspective, the Bulldogs are averaging just over 11 points scored per game as a team. Given Cleary is touching the ball over 70 times per game, more than any other half in the competition, it’s not surprising his impact on a game is so immense.

Tom Trbojevic is adding +10.29 NPRF/game for the Sea Eagles this season, and his mere presence on the field has turned them from wooden spoon contenders to a legitimate finals football team.

Third and fourth places are taken by a pair of Storm fullbacks, Ryan Papenhuyzen (+8.67) and Nicho Hynes (+7.60). The Storm have four players inside the top 10, with Harry Grant (+6.4) and Jahrome Hughes (+6.18) also having strong seasons.

It’s also worth noting just how good Brett Morris was playing this season at 34, with the fifth best NPRF average this season. The Eye Test would like to wish him the best in retirement, he will be missed.

This chart also indicates how important defense is. Cody Walker is having a great season with the ball, scoring nine tries, handing out 11 try assists and contributing seven more, playing him second in the NRL for total try involvements. But with 11 try causes for the season (meaning he was deemed responsible for conceding a try) his NPRF per game is only +5.89. Without those try causes his points responsible for would be +9.81 per game, well inside the top five.

Error Rate leaders

The last advanced statistic we’re looking at this week is Error Rate, which is the number of possessions it takes for a player to commit an error. As a rough guide, anything less than 10 is terrible, and between 10-15 is bad, and 15-30 is not good.

Dragons winger Jordan Pereira takes top (or should that be bottom?) spot with an error every seven possessions in his three games. That equates to nearly three errors per 80 minutes, which you don’t need me to tell you is not good.

Second place is Warriors backrower Josh Curran, who is committing an error every 9.14 possessions, with Titans edge Beau Fermor at 9.33 making up the top three. Canberra veteran Jarrod Croker is the only other player with an Error Rate under 10 at 9.86.

Pereira is unlucky that I haven’t set the minimum games played threshold for Error Rate at two games played, as Wests Tigers winger Zac Cinto has committed seven errors in just 34 possessions this season for a rate of one error every 4.86 possessions. But as he has only played the two games, he’s currently excluded from this list.

On the positive side of things, Jayden Brailey has the best error rate in the NRL, with just one error in 1,419 possessions. Given hookers are usually just moving the ball from dummy half 150 times a game, it’s understandable they’re not committing too many mistakes.

A special shout out goes to NRL Physio’s favourite and grilled chicken connoisseur Alex Twal, who has played 653 minutes this season, touched the ball 171 times and is yet to commit an error. He’s the only player with at least 500 minutes played yet to commit one.  

The decline of 80 minute players – NRL Round 12 2021 stats and trends

The first half of the 2021 NRL season is over already. Time sure does fly, unless you’re a Raiders fan, and then it probably doesn’t move quick enough. With the passing of the midpoint, it’s time to look back at what has transpired and see what has changed thus far.

Fatigue has been the main talking point this season, with the NRL claiming there is no link between the rule changes and fatigue. I figured it was a good time to take a broader look at how fatigue and injuries has changed minutes played and players used in 2021.

Earlier this season the Eye Test already looked at how teams were using their interchange bench more often, and earlier in games. With fewer stoppages, greater fatigue impact and more line-up changes, it is no surprise that teams who excel in recruiting outside of their own back yard or prioritise developing their own juniors are excelling this season.

Let’s start by looking at a simple measure – the average minutes played by an NRL player.

Over the past eight seasons players about 61 minutes per game, with 2021 sporting the lowest average at 61.12 minutes, beating the previous low in 2014 of 61.32. This makes sense with the number of injuries and head impact assessments, as players who would usually be playing big minutes are spending more time off field.

Notice the average minutes played hasn’t changed much from 2020 to 2021? It’s down but only by a small margin. Keep this in mind for later.

If you look at the change in minutes played by position, you can see why the minutes played has dropped but hasn’t altered significantly.

Most starting forwards are playing fewer minutes, with front rowers, second rowers and locks all playing around 3% less than last season. This has been made up by interchange players, who are spending 5% more time on field, which makes sense given the number of injuries and reduction in stoppages.

The decline in mintues makes sense, given that there has been a high number of players injured and undergoing a HIA most likely due to the reduction in stoppages from poorly thought out rule changes.

Which raises the question, has that affected how many players each team has used this season? The number of unique players used each season is below. For comparison purposes, 18th men have been excluded.

Here we can see some of the affect of the “fewer stoppages” era of the NRL. 2014 was the only time in the previous eight seasons where over 400 unique players had stepped onto a field in the first twelve rounds. The jump from 2019 (397) to 2020 (420) was about a 6% increase. For those who like to claim that there isn’t enough talent to support a 17th or 18th team, we’re using an extra team of players each season due to attrition under the current environment.

Next up I’ve split this out by team for 2021, to understand which teams have coped the best with the increased fatigue, and who has struggled.

No surprises that Penrith and Parramatta have been the least affected by line up changes this season. Those at with the most changes have been brought about by injuries (Roosters, Warriors), suspensions (St George), reshuffles (North Queensland) or a combination of all three (Brisbane, Canterbury).

Moving back to the unique players used, those 423 unique players have played 3,255 games. Of those games, 1,533 players have played at least 80 minutes, or about 47%. How does that stack up historically?

Unsurprisingly, again it’s the lowest percentage of players playing 80 minutes since 2014, and nearly a 2% drop from 2020 when set restarts were implemented in Round 3. Coaches are finding they cannot leave players on the field as long as they used to, whether that be due to injury or fatigue leading to bad decisions. Keep in mind that some players are being forced to play 80 minutes because there were no available interchange players due to injury.

Even more alarmingly, backs (numbers 1-7) are playing the lowest percentage of full games since 2014 as well, at just under 89% of players hitting 80 minutes.

After only once hitting 8% prior to 2020, we’ve had two seasons of the percentage of backs not playing 80 minutes above 11%. Over one in ten backs are now not playing the full game. Regardless of the reason, whether it be concussion, injury or getting hooked, backs are playing fewer minutes than previous seasons.

The last thing I’m going to check is the distribution of minutes played to check for any trends. Below is a box and whisker plot, which can be used to show the spread and a five number summary of a set of data. The numbers are the minimum, first quartile (25%), median, third quartile (75%) and maximum. Here’s how the distribution of mintues over the last eight seasons looks.

The thing that stands out from looking at minutes is that the median (middle number in a set of data) is considerably lower in 2021 than 2020, or any other year in this set. Last year the median was 77, meaning that half of the all players up to Round 12 played at least 77 minutes. This season, it’s down to just 71 minutes, which is almost a 10% change in one season. Yet as I mentioned before the average was similar (61.39 in 2020 to 61.12 in 2021).

The third quartile hasn’t changed much, and at 41 minutes is in line with other seasons, confirming what we discovered earlier, that fewer players are playing the full 80 minutes and it’s dragging down the median. Those minutes are being distributed to interchange players who are spending more time on field and the average is masking the change. It’s also why you don’t use averages in your press releases to unsuccessfully quell player unrest.

It’s pretty hard to argue now that players aren’t being affected by fatigue this season, with fewer playing 80 minutes, interchange player are spending more time on the field and a larger percentage of backs aren’t playing the full game.

The NRL can use whatever unrelated metrics they like to show that things are working. but as a wise man once said, “the kind of people swayed by statistics are generally not swayed by bad statistics and have keen enough noses to smell dumb shit a mile off”.

Being concerned about player welfare and not being a fan of the current rules aren’t mutually exclusive. Unlike those with a predisposed agenda, the Eye Test is agnostic and only relies on facts.

Net post contact metres

Post contact metres are often used as a barometer for success for NRL teams, as run metres correlates highly with winning games. But how do you judge if a team won that battle? Gaining 600 metres post contact sounds good, unless you’re giving up 700m.

The answer is to look at their net post contact metres. Net post contact metres is derived by taking a team’s post contact metres and subtracting their opponents post contact metres.

The chart below shows all 16 NRL clubs and their net post contact metres over the twelve rounds played this season, with a line showing their average net post contact metres for 2021.

There are a few reasons I like this chart, which I used last season as well. The first is that I can see every team at once. The second is that it clearly shows who is performing well (Penrith, Melbourne), and who is struggling.

Canterbury haven’t had a game with positive net post contact metres all season and are giving up an average of 180 more to their opponents. Melbourne, on the other hand have only lost one post contact metre battle all season, which came in Round 7 against the Warriors in a game they won.

The third is that it’s great at showing the ebbs and flows of a team over a period of time. Parramatta’s success over the past two months can be seen before screeching to a halt in the last two rounds against Manly and South Sydney. The Raiders struggles can be seen as well, coming out on top only a handful of times.

It also shows that despite the Dragons strong start to the season wasn’t sustainable as they didn’t win a single post contact metre battle during that time. They have only come out on top once this season, which was Round 9 against the Bulldogs, who everyone comes out on top against.

To’o’s rise to Origin

The naming of Panthers winger to the New South Wales State of Origin team this week ruffled a few feathers, with detractors pointing to his lack of heigh compared to Daniel Tupou as a liability at that level of the game.

It may be the case, but if you compare the two players statistically this season there’s no doubt To’o is ahead of Tupou but not by a wide margin. Tries per game is the one metric where the Roosters winger comes out ahead.

The main difference is the incredible work rate and damage To’o does when running the ball. I’ve posted the next two charts a number of times but given his selection it’s worth revisiting them.

To’o is still blitzing the rest of the league in long runs (greater than 8 metres), with 16.6 per game. That’s almost as many total runs per game as Tupou has (18.2).

He’s also busting nearly twice as many tackles as Tupou as seen below.

With the fast past of State of Origin set to be even faster this season with the new rule changes, having someone with the work rate and explosive ability holding the ball that To’o has could be an important factor in spelling the Blues back when they’re gassed.