NRL 2020 – how the game changed statistically this season

The NRL 2020 regular season has drawn to a close and much like the Brisbane Broncos it is time to get to the bottom of it. Statistically that is.

There were some significant rule changes this season that have impacted the way teams are playing. The biggest ones were the move to one referee and the introduction of six again calls for ruck infringements, which I’ve gone over extensively before.

The other less talked about one was the change to provide attacking players the same protection under high kicks that defensive players have received. This change has certainly played a part in how teams are attacking at the end of sets, with a sizable increase in kicks aimed between 1-10 metres out from the try line instead of trying to place the ball in the in goal area. It has also resulted in some teams, most notably the Melbourne Storm, just running the ball on the last tackle to hand it over a few metres out, rather than have a mistimed kick result in a seven tackle set from the twenty metre line.

Just how much did those rule changes affect the way the game looked statistically? To check how much things have changed, we’re going to look at the percentage increase on a per game basis from the average of all 25 rounds in 2019 to 2020 from Round 3 to 20, for groups of publicly available statistics from Fox Sports.

If you read my breakdown of the Warriors statistical improvement under Todd Payten from earlier in the season, you’ll know why I’m just using the percentage change and not the raw number change. If not, then the reason is that it’s hard to show per game shifts in statistics with massively different ranges. You can show the change in run metres from 2800-2900 per game, but you would never see the change in average metres per run from 8.92 to 8.85. To deal with that I am purely looking at the percentage change, which for most statistics is in the single figures to low double figure range, which allows for a greater distinction of change.

Now we’ve defined what we’re looking at, what changed under V’Landysball in 2020? And by how much? Turns out quite a bit.

Time in play

The biggest change was time in play. With around 22% fewer penalties being called, the ball wasn’t sitting idle as long and time in play jumped by 6% (golden point games excluded from each season). You can see the round by round breakdown and three round rolling average (orange line) below.

The average time in play increaed by 6% from 54.15 minutes to 57.73 minutes. A lot can happen in three minutes in the NRL, although from the statistics below most of it seems to be middle forwards running the ball. The interesting thing was the decline in time in play late in the season. Rounds 3-12 had an increase in time in play of 9.6%. Yet Rounds 12-20 only had a 1.6% increase. Something to investigate…

Possession

Moving on, these rule changes had a flow on affect to practically every other single statistic in the game. More time in play means more possession. More possession means more runs. More runs mean more run metres. Which leads to more kicks at the end of sets. More tackles need to be made. And so forth.

It does mean that anyone averaging a “career high” this season that is less than a 4-6% increase is probably not having a career high if you adjust 2020 stats to be in line with 2019. That doesn’t mean they haven’t had a career season by effort, just that their numbers are slightly inflated and not completely comparable to previous season without adjustment.

% change in possession statistics, 2019 vs 2020

Looking at the above chart, the orange data points show the 2020 percentage change, and you can see that average sets per game are up nearly 8%, as are average play the balls (+7.47%) but average tackles per set is basically flat (down 0.4%). This is no surprise with the increase in early kicks seen this season.

Completion rate also increased slightly, up 2.4% to 78.3% per game from 76.5% last season, pointing to a slightly more conservative approach despite the increase in tries and points scored.

Penalties and set restarts

As you’d expect with the significant rule change this season, here’s where you see the most changes and it’s had a flow on effect to other parts of the game. There was also the removal of one referee.

Penalties declined by 22% with the introduction of set restarts. This led to a large number of them being called in the first half and a one third drop to the second half as seen below.

Consistency of set restart calls in the second half has been an issue for the second half of the season. We had a run of 6-7 rounds where games had zero set restarts in the second half, and then we had Round 19 where three of the four highest second halves for set restarts occurred. It’s something that needs to be tightened up for 2021.

Scoring and passing

% change in scoring and passing statistics, 2019 vs 2020

Thanks to some high scoring final rounds of the regular season, scoring increased almost in line with the increase in time in play or possession, increasing 5.78% per game to 41.74 points per game, up from 39.5 per game last season.

Tries were up 9.78%, goal attempts were up 1.8% while goal makes were down 2.1%. Part of this is probably linked to the decline in penalties with set restarts being introduced, as penalty goal attempts are down 32% this season on the back of penalties awarded declining by 22%. As penalty attempts were usually taken in positions where the goal kicker was likely to succeed, it makes sense that the overall percentage would decline.

Line breaks increased at a lower rate than tries, at just 4.72%, which makes sense after you read the kicking analysis.

It is interesting that the increase in possession and time in play hasn’t led to an increase in general passes, which were only up 1%. We’ll see why though in the analysis of runs and running metres. Offloads were down by 5%, again supporting a theory that coaches were playing a more conservative game. That is also shown with errors being basically flat on last year, from 21.7 to 21.5 per game.

Running and metres

As you’d expect with more time in play there’s been an increase in runs, up 4.6% and total run metres are up 3.8%. The interesting thing with run metres is how we’ve arrived at that increase. Post contact metres are up 3%, while pre contact metres are up 5.4%, leading to a decline in average metres per run from 8.92 to 8.85. It’s not a lot but when you consider there’s been over 50,000 runs this season it does add up.

The type of runs has seen a big change as well, with dummy half runs down just over 9% on last year, dropping from 17.6 to 16.0 per game. This has mostly moved to one pass runs, a standard hit up, which has increased 7.3% to 155 per game from 145 per game in 2019. These numbers would explain the higher increase in pre contact metres and a slightly lower metres per run, and the decline in passing and offloads as mentioned above.

Again, this points to a more conservative approach, which again leads into the next section. Set restarts are also having an effect here, as when the tackle count restarts, they continue to push the ball through the middle.

Kicking stats

% change in kicking statistics, 2019 vs 2020

One of the biggest changes this season was in kicking type. Overall kicks have increased by a similar amount to runs and run metres, up 3.9% to 18.2 per game. When you drill down into that, you can see that long kicks are up 11.5% and attacking kicks have increased by 5.7%, while weighted kicks are down by a huge 25%, which you can see by the orange data point on the left. I’ll quote my theory from the Round 14 Notes and Trends post as to why that is the case.

“Here’s my crackpot theory – teams have gotten more efficient and accurate at aiming their attacking kicks just outside the goal area to avoid a seven tackle set. The rule change which came into effect in Round 1 that gives airborne attacking players the same level of protection as airborne defensive players is surely a driver for this, as Daniel Tupou was showing before succumbing to injury.

This explains the drop in weighted kicks but the large increase in attacking kicks. Fewer kicks reaching the in-goal area leads to fewer dropouts which can take up to 45 seconds each. By aiming them a bit shorter than the try line, at worst a team will give up possession less than 10 metres out or a scrum at the same point. This is a much better result than a seven-tackle set from the 20 metre line.”

The Melbourne Storm were one of the teams driving this drop in weighted kicks, as I noted earlier in the season. They were quite happy to run out the ball on the final tackle and ensure their defense was set well inside their opponents 20 metre zone.

There’s been a decline in forced dropouts as well, down from 3.33 to 3.28 (-1.5%), and fewer kicks going dead (-6.9%) which also supports this trend. The reduction in dropouts taken has also led to part of the increase time in play, it’s not purely the cause of set restarts.

I mentioned before that line breaks hadn’t increased at the same rate as tries, and one of my theories is that there are more tries being scored from attacking kicks, which aren’t awarded any line breaks.

Defensive stats

% change in defensive statistics, 2019 vs2020

Not a lot has changed in the few defensive statistics publicly available, with tackles up just 3.2%, and missed tackles climbing by 2.5%. Tackle efficiency, which is a cautious stat to be using in the first place, barely changed, sitting at 92.41% last year and 92.46% this year, which is why you can’t see the orange data point.

When you bring all these small changes together, it shows a change in the way the game has been played. We’ve seen more ball in play thanks to the change to set restarts for ruck infringements. This has led to an increase in hit ups through the middle of the field, and a further neutering of dummy half running. The law of unintended consequences led to conservative one out running with little ball distribution becoming a larger part of the game.

This has been offset by the dramatic change in kicking profiles, with teams favouring attacking kicks within the 10 metre zone. This has come at the expense of short weighted kicks that are aimed to sit up in the in-goal area and draw repeated sets of six. Coaches have become even more risk adverse, happier to hand over the ball a few metres out instead of potentially giving up a seven-tackle set from the twenty metre line.

Given the gravity of the changes made this season, hopefully we’ll see a more nuanced approach to rule amendments in 2021.

Final set restart update

Hopefully, this is the last time I have to write something about set restarts for at least a month (*notes Grand Final date*). After Round 20, we had one of the lowest numbers of total infringements called since Round 4, with a penalty or set restart being called approximately every 22 play the balls.

On the positive side, there was a bit more consistently among whistleblowers this round, even with the wacky rule changes that were being used. I would have bet my house that Andrew Gee would have given at least 15 once they let him call them for offside.

And that restraint has ensured that Gee didn’t finish the season with an average of 10 or more set restarts called per game. He did manage to call four more per game than Chris Sutton though. There’s always next season Andrew.

Final Error Rate update

I’ve been posting Error Rate updates throughout the season, and with the regular season finished it’s a good time to reflect on 2020 and see who had the worst hands in the NRL this year.

I had planned to put more than a two-game minimum to qualify for this list, but with Nene McDonald making six in just two games, I stopped at a minimum of three errors required. McDonald’s rate of an error every five times he touched the ball and three every 80 minutes is horrific.

Only slightly less horrific is North Queensland’s Shane Wright at one every 7.8 possessions and the Tigers Asu Kepaoa at 8.3 possessions per error.

Not that it seems to be causing the Roosters many problems, but Josh Morris the highest profile name on this leaders list, with an error every 9.88 possessions. Of 28 NRL players who have made at least 20 errors, only one of them has fewer possessions. That would be another Tigers back, Tommy Talau, who has 20 errors in 206 possessions for a rate of 10.3.

Final NPRF update

And finally, as this is the last (regular) post for the season we’ll finish on a high with the full season look at Net Points Responsible For (NPRF).

Nathan Cleary hangs on to first place at +9.72 net points per game responsible for. Luke Keary and Shaun Johnson round out the top three, but the big story is Cody Walker charging into fourth spot after his amazing game against the Roosters on Friday.

Jarome Luai has also had a fantastic month to close the season and takes fifth spot at +6.0 NPRF per game, equal with Cameron Smith and Jahrome Hughes. You can see the impact AJ Brimson has had for the Gold Coast as well, averaging 4.0 NPRF per game.

Here’s the bottom 20 for the season with a minimum of five games played.

Brisbane’s Jesse Arthars holds the worst NPRF per game this season, giving up 5.33 points per outing. Manly’s Albert Hopoate (-4.80) and the Bulldogs Christian Crichton (-4.50) make up the remainder of the bottom three.

Eels fans won’t be surprised to see Blake Ferguson sitting on this list either, especially after his defensive lapses against the Tigers, at a stone cold -3.16 per game.

Advertisement

NRLW Advanced Stats – 2018 and 2019 seasons

If you’re a regular reader of the Eye Test^tm, you’ll be (somewhat) familiar with the advanced statistics I use to analyse players throughout and across each NRL season. And now, they’re also available for previous NRLW seasons.

What do these statistics show? The issue is that generally middle forwards don’t play big minutes or put up big numbers and go unnoticed besides the odd comment about how much of an impact they make. To do so I created three advanced statistics for rugby league – Tackle %, Run % and Involvement Rate.

1. Tackle % estimates the percentage of opponent plays whilst on field where a player completed a tackle.
2. Run % estimates the percentage of team plays where a player completed a run during their time on field.
3. Involvement Rate combines them and estimates the percentage of total plays a player completed a run or tackle whilst on the field.

If you want to read more about them, I’ve linked the explanations of them from the site.

Now I’ve explained them, let’s see how NRLW positions compare for these statistics to their NRL equivalents and look back at some top performers within each statistic from the past two NRL seasons. I would like to note that we’re dealing with some very small sample sizes, and even with a lenient minute restriction of 40 minutes I would still take these as indicative rather than representative of performances. Also, “current team” may also mean “previous team” for someone not playing NRLW currently, which is one of the issues looking at multiple seasons of data at once. Anyway, lets move on to the analysis.

Tackle %

There’s not a huge difference between the NRL and NRLW for Tackle %. Hookers and Locks are making tackles at the same rate, around 25-25%, indicating they complete a tackle on one in four defensive plays. Props and interchange players still sit over 20%, but around 4% lower than their NRL counterparts. This could indicate that a lot of the basic hit up work could be centered less on the middle of the field.

Positions on either edge of the field are similar as well, other than Second Row being down over 2% and Five Eight being up almost 2%, which given that they defend in the same spot could cancel each other out. Fullback is slightly higher as well at 4% for NRLW compared to 2.8% for NRL.

Who are the top NRLW players for Tackle % then? Below are the top 15 players by Tackle % for seasons 2018 & 2019 who played at least 40 total minutes.

Kate Haren formely of the Dragons leads the way with a Tackle % of 43% from her two games, indicating that she made a tackle on two out of every five defensive plays whilst on the field. Aliti Namoce who last plaeyd for Roosters placed second at 31.45% and Talesha Quinn who also last played for he Dragons came in third at 31.32%.

Rebecca Young from the Roosters was the only other player to have a tackle rate above 30%. As noted above, the average Tackle % for middle forwards is about 20-25% so each of these players are tackling well above average for their position.

One of the interesting differences here compared to the NRL is seeing Second Rowers at the top of this list, which is usually just middle forwards for the NRL. In addition to Namoce, Lorina Papali’I (29.38%) and Holli Wheeler (25.93%) also make the top 15 whilst playing in the second row. Again, the average for second rowers is 16.2%, which puts Namoce’s tackle rate almost twice as high the average second rower.

Run %

Looking at the average Run % across positions for the NRLW against the NRL, there’s not as much variance as there was for Tackle %. Differences fall between 0.5% to 1% for most positions, and the only significant change is at lock, where NRL players make a run on 10% of plays, whilst NRLW locks make a run on 7%.

So, who has the highest run rate among all NRLW players who played at least 40 minutes across 2018 & 2019?

Ngatotokotoru Arakua takes first place with a Run % of 20.92%, meaning she completes a run on at least two out of every five plays the Dragons used the ball. Second place is Chloe Caldwell from the Roosters at 20.45%, the only other player above 20%. Another former Rooster in Elianna Walton picks up third place with a run rate of 17.99%.

With the average Run % for middle forwards sitting in the 10-12% range, anyone over 15% is putting in an elite amount of work.

It’s also worth noting that new Dragons signing Isabelle Kelly at centre was extremely close to making this top 15, sitting only 0.15% outside with a run rate of 12.04% which is exceptional for a centre.

Involvement Rate

Given that Involvement Rate is a combination of Tackle % and Run %, it would make sense that any changes we saw in the previous two statistics would be reflected here as well.

Involvement Rates for middle forwards are slightly down as was exhibited with Tackle % rates. NRLW Prop forwards suffer the biggest drop at 2% compared to their NRL counterparts, otherwise things are relatively consistent.

The top 15 players in the NRL for Involvement Rate from the 2018 and 2019 seasons who played at least 40 minutes are shown below.

Kate Haren takes top spot here, as she did with tackle %, with an Involvement Rate of 26.37%, meaning she completed a run or a tackle on one in every four plays whilst she was on the field. Ngatotokotoru Arakua came in second with an Involvement Rate of 21.97% while Chloe Caldwell rounded out the top three with an Involvement Rate of 21.42%.

Brisbane halfback Tarryn Aiken is also worth mentioning, sitting in 19^th place with an involvement Rate of 17.25% in a list that is dominated by middle forwards. This is mainly due to her running game, as she sits just outside the leaders in Run % with a run rate of 10.23%.

NRL Round 20 advanced stats – Involvement Rate

Involvement Rate is an advanced statistic for rugby league that I created to identify players who have a high workload but don’t play a lot of minutes. If you’re new to the site and want to understand how it works, I would recommend reading this post on Involvement Rate.

With that out of the way, here’s the all minutes leaders for Round 20

It’s the single digit crew again after these three took over the Run % chart. The Raiders Jarret Subloo sits first (49.23%) ahead of Jake Friend (33.97%) and Daniel Alvaro (28.66%).

The highest double figure minutes player was Aaron Woods at 27.77%.

Next, we’ll look at those players who spent 40 minutes or more on field

Alex Twal takes first place in Involvement Rate after taking first in Tackle % as well. This time his Involvement Rate was 26.48% meaning he made a tackle or run in over a quarter of all plays whilst on field for the Tigers against the Eels.

James Tamou from Penrith placed next with a rate of 24.61% and Christian Welch was next with an Involvement Rate of 22.61%. Nathan Brown was the only 80 minute player on the leader board this week with an Involvement Rate of 19.11% in his full game.

There’s that Darren Schonig again, popping up at the bottom of the leader board with 18.61% in 44 minutes. Always a good sign when a high workload player can maintain that in extended minutes.

Finally, we have the leaders for Involvement Rate this season.

No changes here with Jaimin Jolliffe snaring first place for the season with an Involvement Rate of 21.93%. A huge effort for the rookie and things continue to look up for the Titans in 2021. Despite not playing, he managed to hold off Jai Whitbread who again snuck into the top three after not qualifying with enough minutes beforehand, with a rate of 21.79%.

Third place went to the Warriors Jazz Tevaga at 21.39%, however he was only 0.02% ahead of the Dragons Blake Lawrie and 0.03% ahead of Melbourne’s Christian Welch.

NRL Round 20 advanced stats – Run %

For those new to the site, I’d recommend reading this post on Run % which details how it is calculated and how to use it.

Here’s the leading players in the NRL after Round 20 without a minute restriction

Canberra rookie Jarrett Subloo sneaks into first with two runs in one minute for a Run % of 93.02% and a lesson in small sample sizes. Eye Test^TM Hall of Famer Daniel Alvaro placed second with 23.81% in his active eight minutes on field as the Eels ran down the Tigers. Jake Friend completed the single digit minute trifecta with a Run % of 22.22% in his three minutes. Again, how good are small sample sizes?

Moving on, let’s look at the 40 minute plus players for this round.

Another raider nabs first here with Dunamis Lui capping off his strong 2020 campaign with a Run % of 18.43% in the Raiders win over Cronulla. Second place went to the Warriors Lachlan Burr at 17.74% with James Tamou placing third at 17.67%.

Two Roosters had the only 80 minute games in the leader board this week, both backs with James Tedesco and Daniel Tupou sporting the same Run % (15.00%).

To finish up we’ll take a look at the 2020 season leaders for Run %.

Unlike with Tackle %, there was no last minute change in the order. Cronulla’s Andrew Fitifa takes the crown here with a Run % of 17.34% for the season. Second place was Melbourne’s Nelson Asofa-Solomona at 16.50% and Parramatta’s Kane Evans in third place with a run rate of 16.35%.

No other player managed a run rate above 15% for the season, showing just how far ahead these three were from the rest of the league.

NRL Round 20 advanced stats – Tackle %

Let’s skip the intro – if you’re new to the site, I’d recommend reading this post on Tackle %, which explains how it works and why I think it’s an important statistic for identifying high motor middle and interchange forwards.

Here’s how the Tackle % chart looked for Round 20 without a minute restriction

Single digits winners are back with Jack Friend making 3 tackles in 3 minutes for a tackle rate of 51.61% before failing a HIA. True first place went to NRL Physio’s favourite Tiger Alex Twal with a Tackle % of 41.08% in their game against the Eels.

Next up was Corey Jensen from the Cowboys at 41.03% and third place was Darby Medlyn from Canberra at 40.67%.

Next, we’ll look at those players who spent at least half a game on the field this round

Twal repeats here with his tackle rate of 41.08% coming in 51 minutes. Second and third place went to a pair of Panthers, with Api Koroisau next at 33.48% and James Tamou not far behind at 32.92%.

Cameron Murray (31.67%) from the Rabbitohs and Reed Mahoney (31.51%)from the Eels placed fifth and sixth behind Christian Welch (31.99%, which is amazing considering those two played the full 80 minutes.

Finally, here is the 2020 season leader board for Tackle % with one round to go.

Here’s one for the ages. Another Titan has snuck in and grasped the Tackle % title for 2020 away from Nathan Peats at the last minute. Jai Whitbread has Stephen Bradbury’d the Tackle % crown by passing the 250 minute restriction with 23 in their game against Knights. This pushed him into first place with a Tackle % of 35.93%.

That puts him ahead of team mates Nathan Peats (34.08%) and Mitch Rein (32.65%), who looked all but certain to claim the top two spots. Whitbread’s surge pushes the retiring Tim Glasby (31.75%) to fourth. Congratulations to Whitbread for taking top spot in 2020.

NRL Round 19 notes and trends

Set restarts remain consistently inconsistent

Regular followers and readers will know I’m a bit of a stickler for consistency with set restarts. I’d pointed out recently that there had been a number of games over the past two months that had zero ruck infringements in the second half. The theory that referees were setting the standards early holds some merit, although I still refuse to believe fatigued players were that well behaved after misbehaving early on.

The good news is that Round 18 was the first time since Round 13 that we didn’t have a second half where zero set restarts were called, and the first time since Round 9 where every game had at least two set restarts called in the second half. Is this progress?

The bad news is that things have swung completely in the opposite direction. Round 19 featured three of the four highest second half set restart counts this season, as you can see below.

That’s a considerable shift in second half interpretations of ruck infringements, going from multiple games with zero to three of the four highest second halves this season. And one of them was by Ben Cummins, who has the lowest average for set restarts called per game. This round in particular had some wildly different numbers of transgressions, as you can see below with each referee’s breakdown in Round 19.

We have three games where five set restarts were called in total, and four that were in double figures. Penalties were much more consistent, with six falling in the 7-10 range with two other outliers.

Let’s not even get into the fact that one game this weekend is going to be giving set restarts for what were penalties for offside. The kicker is that the referee for this game is Adam Gee. Reality is often far better at comedy than any team of writers. But I digress.

The interesting thing is that penalties have been called far more consistently, as you can see from the chart below showing the number of penalties called per half since Round 3. The reference lines for each pane show the average per half.

The difference between halves is around 8% for penalties and 31% for set restarts. That’s a pretty big disparity between periods, compounded as seen above by the wild swings between games.

The historic drop in between penalties between halveshas been about 21%. The decline in total penalties in 2020 (regular penalties plus set restarts) is around 18%. This would indicate that there has been slightly more consistency this season. Although if you look at the drops this year for penalties (-8%) and set restarts (-31%), you could assume that the drop in previous years was entirely from ruck infringements. That’s a long bow to draw though.

Whatever the reason, the inconsistency of application for set restarts is extremely obvious to most fans. Overall, I think the change has been a success but needs some fine tuning for 2021. Just not the type of fine tuning we’re seeing in dead rubbers for Round 20 however.

Penalty goal attempt analysis

Another update from earlier in the season, this time on penalty goal attempts as a proportion of total goal attempts. One of the things I like about this chart is that you can see how a team has performed in attack over time, as the lightly shaded area represents conversion attempts (and therefore tries).

It also shows where teams have made changes to their strategy in taking penalty goal attempts – the sharp increase in penalty goal attempts for the Rabbitohs as soon as Wayne Bennett joined was noted last time.

Below is the update for Rounds 1-19 from 2014-2020, with conversion attempts in the light shade and penalty goal attempts in the darker shade.

It’s extremely easy to see the Panthers rise this season from their panel, and it looks almost as if 2019 was an anomaly. Generally the more successful teams have a higher number of penalty goal attempts, either from trying to extend a lead or being able to generate penalties in favourable areas.

The steady decline of the Broncos since Bennett left is also exceptionally apparent, and whilst the Tigers haven’t improved much on the ladder their attack has picked up with the coaching change to Michael McGuire.

Advanced statistics for middle forwards season leaders

One of the things I’ve been interested in tracking for a while is the performance of middle forwards. I know I’m not the only one who enjoys watching the likes of Christian Welch, Daniel Alvaro or Toby Rudolf putting in the hard work.

The issue is that generally middle forwards don’t play big minutes or put up big numbers and go unnoticed besides the odd comment about how much of an impact they make. To do so I created a three advanced statistics for rugby league – Tackle %, Run % and Involvement Rate.

Tackle % estimates the percentage of opponent plays whilst on field where a player completed a tackle.

Run % estimates the percentage of team plays where a player completed a run during their time on field.

Involvement Rate combines Run % and Tackle % to estimate the percentage of total plays a player completed a run or tackle whilst on the field.

If you want to read more about them, I’ve linked the explanations of them from the site. I also post weekly updates for each statistic after each round. The aim isn’t to show the quality of a middle forwards performance, but to quantify it and identify players with large motor who put in consistently high levels of effort.

To give you an idea of the average rates for each statistic, the below chart shows them broken down by position.

Now we know how the average player performs, the table below shows the 2020 season leader board for Tackle % with one round to go. From the chart above, you can see that anyone with a tackle rate above 25-26% has an above average workload. I’ve also put a minimum of 250 minutes played which ensures a decent sample size to work with as individual games can have significant volatility.

The top three for Tackle % has been relatively consistent over the past month, with a pair of Titans, Nathan Peats (34.08%) and Mitch Rein (33.22%), as well as the Tigers Elijah Taylor (32.28%) sitting in the same spots as last week and unlikely to change with one game remaining. A tackle rate of 30% would indicate that a player is making a tackle on 3 out of every 10 defensive plays whilst they are on the field.

One of the reasons you usually only see hookers on the Tackle % chart is that their role in modern rugby league is that of defense and distribution. They’re tackling through the middle of the field and passing from dummy half, rarely running with the ball. That can also be seen from the above chart by position, showing hookers with a Run % of just 3 %, only ahead of halfbacks at 2.9%.

Two more Titans sit inside the top six for Tackle %, with Jaimin Jolliffe (31.61%) and Jarrod Wallace (31.35%) taking up fifth and sixth spots. The presence of so many Titans at the top of this list is due to the incredible amounts of defense they were required to do earlier in the season as they were routinely getting pumped.

The other notable name in this list is the Panthers Moses Leota, who I’d noted as one of the most improved players in the NRL this season in another post this week.

Next, we’ll take a look at the 2020 season leaders for Run %. Again, from the chart above, any middle forward with a run rate of over 12-13% is putting in greater than average effort.

Andrew Fifita is the season leader and looks like going unchallenged for the remainder of the 2020 as he sits with a 2020 Run % of 18.12%. The lingering knee issues have forced him to change his game and he’s used for an impact in short bursts on the field with the ball, rather than conserving his energy and spacing out his efforts. This run percentage indicates that Fifita is makign a run on nearly 1 out of every five plays for the Sharks whilst he is on the field, or basically a hit up every set of six they have.

Melbourne’s Nelson Asofa-Solomona has been a strong runner of the ball this season and occupies second place at 16.50% but is too far behind to close the gap with one game remaining. The Eels Kane Evans maintains third spot with a run rate for the season of 16.24%.

There’s a huge gap of nearly 1.5% then to fourth place, which is Jason Taumalolo (14.87%) of the Cowboys who returned on the weekend in limited minutes.

Finally, we have the leaders for Involvement Rate this season. The chart above shows that an Involvement Rate above 20% would be considered elite.

Titans rookie prop Jaimin Jolliffe sits first at 21.93% and that won’t change with the front rower out with an injury. New Zealand’s Jazz Tevaga not far behind at 21.69% but has been named at lock this week and may drop slightly if he plays big minutes.

Usually Involvement Rate declines as time increases as workloads have to be managed, as shown below.

The Titans Jarrod Wallace sits in third spot at 21.39% but has Blake Lawrie from the Dragons breathing down neck, just 0.03% behind at 21.36%. There are another five players sitting just behind Lawrie between 21.24% and 21.13%, and any one of them could move into the podium with a high workload game in Round 20.

Involvement Rates over 20% indicate that these players are macking a tackle or completing a run on one in every five plays whilst on field, which is a huge effort when you consider some of the minutes they are playing.

NRL Round 19 advanced stats – Involvement Rate

Involvement Rate is an advanced statistic for rugby league that I created to identify players who have a high workload but don’t play a lot of minutes. If you’re new to the site and want to understand how it works, I would recommend reading this post on Involvement Rate.

With that out of the way, here’s the all minutes leaders for Round 19

After leading the Tackle % chart, it’s no surprise to see North Queensland’s Emry Pere first for Involvement Rate as well, at 30.33% for the round. This indicates he made a run or completed a tackle on 30.33% of all plays during the Cowboys loss to Penrith.

Second place was Warriors debutant Tom Ale who had an Involvement Rate of nearly 28% in his 12 minutes, whilst Andrew Fifita

Next, we’ll look at those players who spent 40 minutes or more on field in Round 19.

As with Tackle % the leader takes the cake again here. New Zealand’s Lachlan Burr is in first with an Involvement Rate of 26.41%. The Sharks Toby Rudolf nabbed second spot with a rate of 23.32% while the Roosters Lindsay Collins rounds out the top three at 23.17%.

Patrick Carrigan is the only 80 minute player in the leader board this week with an Involvement Rate of 20.12% as the Broncos lost again to the Eels.

Finally, we have the leaders for Involvement Rate for 2020 with one game remaining.

With Jaimin Jolliffe not playing for the Gold Coast this round he still sits first at 21.93%, with New Zealand’s Jazz Tevaga not far behind at 21.69%. The Titans Jarrod Wallace preserves his third spot at 21.39% but has Blake Lawrie from the Dragons breathing down neck, just 0.03% behind at 21.36%.

There’s another five players sitting just behind Lawrie between 21.24% and 21.13%, and any one of them could move into the podium with a high workload game in Round 20.

NRL Round 19 advanced stats – Tackle %

Let’s skip the intro – if you’re new to the site, I’d recommend reading this post on Tackle %, which explains how it works and why I think it’s an important statistic for identifying high motor middle and interchange forwards.

Here’s how the Tackle % chart looked for Round 19 without a minute restriction

No single digit minute shenanigans this week. The Cowboys Emry Pere takes top spot with a Tackle % of 46.58%, indicating he completed a tackle on nearly half of the possession his team defended whilst Pere was on the field.

Second and third place goes to a pair of Warriors, and sadly one wasn’t Daniel Alvaro. Lachlan Burr took second place with a tackle rate of 43.94% and debutant Tom Ale at 41.67%. Ale completed 9 tackles in his 12 minutes.

Sam McIntyre (40.55%) from the Wests Tigers was the only other player above 40% this round.

Next, we’ll look at those players who spent at least half a game on the field this round

Burr takes top spot for those who played at least 40 minutes, with his tackle rate of 43.93%. One of the NRL’s most improved players Moses Leota from Penrith grabbed second spot with a Tackle % of 35.63%. Jacob Saifiti takes third place as the Knights prop posted a tackle rate of 33.93% in his 41 minutes on field during their win over the Dragons.

There was a number of 80 minute hookers making this list this week – Penrith’s Mitch Kenny (31.61%), North Queensland’s Reuben Cotter (30.98%) and South’s Damien Cook (28.50%).

Finally, here is the 2020 season leader board for Tackle % with one round to go.

No movement in the top three, with Nathan Peats (34.08%), Mitch Rein (33.22%) and Elijah Taylor (32.28%) sitting in the same spots as last week and unlikely to change with one game remaining.

Want to also wish fourth placed Tim Glasby (31.75%) the best in his post NRL career.

NRL Round 19 advanced stats – Run %

For those new to the site, I’d recommend reading this post on Run % which details how it is calculated and how to use it.

Here’s the leading players in the NRL after Round 19 without a minute restriction

Zane Musgrove led the way this round with a Tackle % of 25.81%, meaning he completed a run on over a quarter of all plays the Tigers had whilst he was on the field. The remainder of the top three comprised of Souths’ Mark Nicholls (21.64%) and season leader Cronulla’s Andrew Fifita (20.88%).

Abbas Miski takes the spot for top back this round, playing 18 minutes and making six runs for a Run % of 18.65% after coming on to the field as an interchange player. Charnze Nicoll-Klokstad of the Raiders was the only other back in the top 20 with a run rate of 13.19% in his 80 minutes.

Moving on, let’s look at the 40 minute plus players for this round.

Musgrove’s teammate Josh Aloiai placed first among high minute players with a run rate of 19.05% and will be sorely missed by the Tigers next season as he recovers from injury. The Gold Coast’s Jarrod Wallace placed second with a Run % of 18.23% whilst Daniel Saifiti showed no lingering affects of injury with a Run % of 17.28% in 43 minutes.

The Sea Eagles’ Curtis Sironen was the only 80-minute player inside the top 20 this round, with a run rate of 14.69%, coming from 21 runs.

To finish up we’ll take a look at the 2020 season leaders for Run % with one round to go.

As mentioned above, Fifita is the season leader and looks like going unchallenged for the remainder of the season as he sits with a 2020 Run % of 18.12%. Melbourne’s Nelson Asofa-Solomona still occupies second place at 16.50% and is too far behind to close the gap with one game remaining. The Eels Kane Evans maintains third spot with a run rate for the season of 16.24%.

There’s a huge gap of nearly 1.5% then to fourth place, which is Jason Taumalolo (14.87%) of the Cowboys who returned on the weekend in limited minutes.

NRL Round 18 notes and trends

Just how much does regular season performance dictate finals success?

Last week Rugby League Analytics posted a fantastic visualisation of points for versus points against for every Super League team since 1996, which showed that the majority of teams that either topped the table or won the grand final sit in the bottom right quadrant. This quadrant contains high scoring teams with great defense, which makes perfect sense. Below is the tweet, please give them a follow if you’re into Super League or analytics for the sport.

Following yesterday, this shows the per game performance of specific teams in their season.
Red indicates the team topping the table at the end of the regluar season. Black = Grand Final win (+premiership trophy in 1996/7). Yellow = League + GF#SuperLeague #RugbyLeague #intRL pic.twitter.com/al7C5oS6uT

— Rugby League Analytics (@AnalyticsRugby) September 10, 2020

This led me to wonder if the same trend was evident in the NRL era (1998-2020). Did each Grand Final contain only the best attacking teams who could also defend, or were there some outliers who made to the grand final (or even won it) despite having obvious weaknesses in defense or attack?

First lets define what we’re looking at. The below scatter plot shows every team in the NRL era, and their average points scored plotted against their average points conceded. All data Is taken from The Rugby League Project, the best rugby league resource on the internet.

I’ve changed the legend up a little bit from Rugby League Anlaytics colour coding, adding a few extra segments. It’s a little bit busier but it allows a greater level of analysis. How they finished the season is still colour coded – pink for winning the grand final and minor premiership, purple for winning the minor premiership but losing the grand final, orange for grand final winners only, red for grand loser, blue for winning the minor premiership only, yellow for a top 8 finish and green for no finals played.

For simplicity sake I’m going to include the discarded grand final wins by Melbourne, which as an Eels fan is pretty painful. Additionally, the “Top 8” moniker is a proxy for finals, I am aware there was a 10 team final series in 1998.

Now we know what data we’re looking at, lets have a look at the overall picture before breaking it down into segments.

Just like the above chart from Rugby League Analytics, this one has four quadrants. Moving clockwise from the top left have: bad defense/bad attack, bad defense/good attack, good attack/good defense, and good defense and bad attack. Most of the teams sit in the bad defense/bad attack and good defense/good attack quadrants with a few outliers sitting in the other quadrants.

It’s quite clear from that the successful teams look to be sitting in the bottom right quadrant which is the good defense/good attack area. But from an initial glance there are some grand finalists and top 8 teams sitting outside that quadrant. Let’s take a deeper look and see who the were and if it might give some indication of what could happen this season.

Firstly let’s look at those those teams who achieved the minor premiership and grand final win double, the pink data points below.

No team that has achieved this double had a bad attack or bad defense, except for the mythical 2003 Penrith team that defeated the Roosters with that Scott Sattler tackle.

Next up those minor premiers who didn’t win the grand final, the purple data points in the below chart.

There’s been seven instances of the minor premier making the grand final and losing, and not one of them had anythong other than a combination of good defense/good attack.

What about teams who just won the minor premiership but didn’t make the grand final (the blue data points)?

Again relatively straight forward here – no team that was minor premier but didn’t make the grand final in the last 22 years has had bad defense or bad attack. Again, this is logical, since taking the minor premiership requires a season of consistently good results. I’ve also assumed Penrith wins the minor premiership here with a three point lead with two rounds remaining.

Lets look at grand final winners next, shown in orange.

Of the 22 grand final winners over the past 22 seasons just five of them have come from outside the bottom right good attack/good defense quadrant. The exceptions fall into two groups – historically elite teams like the 2006 Broncos and the 2009 Storm (*cough*), or high scoring teams with a transcend star half. There’s something else those latter two who could score easily have in common and we’ll get to that in shortly.

Now we’ll have a look at grand finalists only, coloured red.

It’s a similar tale, with the majority of losing grand final teams also sitting in the good defense/good attack quadrant. The lone side that lost a grand final despite bad defense was North Queensland in 2005, which also included another team with “bad defense”, the Wests Tigers. Otherwise every grand final team at least had good defense, indicating that at least having a strong defense gives you a chance of making the grand final.

The other thing to note is that those grand finalists within the bad defense/good attack quadrant all occurred before 2006, which is an eternity ago. In the current NRL era, if you don’t have an elite defense you’re not making the grand finals. Pre 2006 NRL seems to be the wild west where you could simply try to outscore a team before the banality of modern NRL where coaches refuse to take any risks and would rather lose a low scoring game than try to win a high scoring one. How good are endless block plays.

Moving on, next we’ll look at teams who played finals football, shown as yellow in the chart below.

This one has a wider spread, although again the majority of teams playing in September had a quality defense, or at least were able to put points on the board. There were a few outliers, with the “worst” team to make the finals being the Canberra Raiders of 2002 who made it in with 10 wins, 13 losses and one draw, and a points differential of -170, sitting well inside the bad defense/bad attack segment.

Let’s flip things now and look at teams who didn’t make the finals, represented by the green data points.

This provides a slightly clearer picture, as teams who have both good attack and defense usually make the postseason. The two “best” teams who didn’t make the finals were the 2002 Bulldogs who were stripped of 37 points, and the 1999 Canberra Raiders. Coincidentally the 1999 Raiders were almost an exact mirror of the 2002 Raiders – 13 wins, 10 losses, 1 draw and a +173 points differential.

After looking at all of this, now we know that you’re unlikely to make the top eight without a good defense, and you’re even more unlikely to make the grand final unless you have a good attack and good defense, where do teams sit for the 2020 season? Here’s the last 22 years with 1998-2019 in blue, with the 2020 season in orange.

The three raging favourites for the 2020 title – Melbourne, Sydney and Penrith –understandably sit in the important bottom right quadrant. The other top four team, Parramatta have sit in the good defense/bad attack quadrant, which was epitomised in the Eels inability to cross the line against Penrith on Friday despite being able to withstand a torrent of pressure in the first half. Both Canberra and Newcastle also sit in this quadrant, signifying that they are also unlikely to feature on the final weekend of the season.

The Knights are nearly inside the quadrant but given their wildly inconsistent results you wouldn’t give them much of a chance even if they were. Of those three Canberra would be the most likely th crash the party and make it to the grand final.

The team I haven’t mentioned yet is one that could be a smokey a grand final win this season – South Sydney. Despite sitting currently in sixth place, they are the only other team besides the top three that sits in the favourable good defense/good attack quadrant. The loss of Latrell Mitchell definitely hurts, but the Rabbitohs have the foundation of a side that could challenge for the title. They aren’t as deep into that quadrant as the other three, but they had started to gain momentum at the right end of the season.

It’s a bit of stating the obvious that there’s only a handful of teams who look like serious contenders this season, but as I’ve stated before analytics is the art of being less wrong . By looking at teams in this lense, we’ve been able to practically eliminate Newcastle, Parramatta and Canberra as legitimate contenders, although the Eye Test^tm would have eliminated the Eels a few rounds back and Newcastle would have never been in the conversation.

It’s also shown that you need an elite defense to be even in the hunt for a title, and the most likely grand final winners are coming from Melbourne, Sydney, and Penrith. The final key point is that Souths are a potential dark horse for the title but will face a battle without Mitchell as their defence and attack are good but not on the same level as the other three contenders.

Are the top four running the ball more consistently?

Over the past few weeks, I’ve been delving into the total metres and pre/post contact metres splits of teams in the NRL with a focus on the Wests Tigers, since they’re having issues promoting the ball lately. To get you up to speed here’s the update after Round 18 for pre/post contact run metres for starting forwards and interchange players.

Wests still have the fewest average pre contact and average total run metres by forwards and interchange player and have the third least average post contact metres. On the other end Penrith have the most post contact and are second in total and pre contact metres. This led me to wonder, are the top sides this season consistently gaining metres?

The below chart shows the total run metres per round for 2020 by all sixteen NRL clubs, with a reference line across the middle showing their average run metres, and a 95% confidence interval (95% certain of the true mean sitting in this range) either side.

At first glance you can see that there is more consistency in run metres from some of the top teams. Melbourne, Sydney, Penrith, Souths and Parramatta are either above the line or very close to their average most weeks. Souths notably started the season poorly but have had a strong run of late, while Parramatta were running the ball more effectively earlier in the season but have dropped off considerably of late, which is concerning considering their injuries are to positions that don’t usually run the ball often (Dylan Brown and Reed Mahoney).

Teams who are struggling this season tend to have wilder fluctuations in their total running metres. Brisbane tend to swing wildly above and below their season average, as do North Queensland and the Warriors. The Gold Coast had some very low points early in the season, but their turn around in form has seen them only dip below their season average once since Round 12. And you also have the Tigers, who finally had a game close to their season average after back to back games where they couldn’t move the ball.

One thing to consider here is that we’re looking at total metres gained during a game. What if we looked at just pre or post contact metres? Would that show anything different? Historically pre contact metres correlate better with scoring points, whilst pre-contact metres correlate more with a higher margin.

Above is the same chart but just for pre contact metres. Turns out there’s not a lot of difference, just some of the peaks and valleys are smoothed out a little bit. That makes sense as around two thirds of total run metres come from pre contact metres, with the other third coming from post contact metres.

So, what do you see when you look at post contact metres only?

It’s quite different! Even a team like Brisbane that was struggling has some massive spikes for pre contact metres. The issue, like other bad teams, is that they lose big when they’re not able to generate metres after contact. The erratic season Newcastle is having can be seen here, alternating from high to low post contact metres on a weekly basis over the past two months.

Penrith has seen a huge uptick in post contact metres in the last month when they’ve been lapping teams. Souths have only had one game below their season average in post contact metres since Round 8, which came against Melbourne. Parramatta had been excellent at pushing through contact earlier in the season but that has waned of late.

There is one notable exception to this rule and that seems to be the Roosters, whose run consistency profile shares more with the bottom eight than the top eight. One theory behind this could be that they’re incredibly good at creating space for their outside backs and putting players into huge gaps in the defense doesn’t actually generate post contact metres because they’re not getting touched in the first place.

#BIGMANSZN kicking stats

Round 18 was a bonanza for middle forwards moving out of their wheelhouse and trying their chances with a kick. First up was Tom Burgess on Thursday night against the Tigers forcing a line drop out with this deft touch that you can see in the below video:

Second was Martin Taupau kicking and regathering (eventually) for himself against the Bulldogs on Friday Night.

Based on this, I wanted to look and see which forwards had been kicking the ball this season. Most other positions generally have a licence to kick the ball when needed, with middle forwards generally on a tight leash. I’ve also removed second rowers as well, since Wade Graham has kicked over 20 times this season. Below is the table of kicks by middle forwards this season.

This season Adam Elliot from Canterbury leads the way among middle forwards with four kicks and one forced drop out. Des Hasler seems to have the longest leash on his middles, with both Taupau and Jake Trbojevic kicking three times each this season, but without any forced dropouts. The only other forced dropout this season by a middle forward was by the Panthers James Tamou.

So we really did see something of a unicorn on the weekend. Let’s hope we see more of it #bigmanszn.

Net Points Responsible For Round 18 update

Net Points Responsible For (NPRF) is a statistic I put together to track the contributions by playmakers for their team that isn’t always shown in raw numbers for try assists. By including the try contribution statistic that Fox Sports uses it brings in players who don’t necessarily throw the deciding pass for a try. I’d first brought it up earlier in the season in this post, highlighting how well Jahrome Hughes was playing.

Players like Nathan Cleary, Brett Morris, James Tedesco, and Luke Keary show up on the the leaderboards for try scoring or try assists, but that doesn’t always provide a true measure of their worth. And given that defense is just as important in rugby league, I’ve deducted four points for every try cause allocated to a player to use as weighting of their defense.

Here is the calculation for NPRF – tries *4 + try assists *4 + try contributions *4 + field goals but adds in a negative four points for every try caused.

Now that we’ve explained it, let’s look at the top 20 players for NPRF after Round 18:

Nathan Cleary still leads the way but has dropped under 10 points added per game for Penrith. His halves partner, Jarome Luai has been exceptional of late and is adding another 5.6 points per game himself and is one of just eight players have a NPRF of more than 5 points per game.

Let’s not forget that Shaun Johnson has been exception despite allegedly “not doing enough” a month or two ago, and Hughes is still having a very good season.

Here’s the bottom 20:

As you’d expect, the majority of this bottom portion of the NPRF table is littered with Broncos, Cowboys and Titans. Jesse Arthars was the worst this season, giving away 10 try causes in just five games for a NPRF of -6.4 per game. Ouch.

The Eels struggles of late down their right edge is seen here too with Blake Ferguson sitting inside the bottom 10, giving away 3.06 points per game. He’s one of the few players in this list to be playing for a top four side, the other notable one being Ryan Hall at -2.4 per game.

If you scan both lists, the one thing that jumps out (especially in the top 20) is the lack of forwards. Other than Cameron Smith, every player in the top 20 is a back or half.

Regular readers will know one of my key tenets is that you (usually) can’t compare players across positions in rugby league as players perform very different roles. With that in mind, lets look at the top 20 players for NPRF filtered purely for forwards, and by forwards, I mean their regular or default position and not named position, as utility interchanges would dominate this list otherwise.

The one thing I like about looking at Net Points Responsible For this way is that it isn’t just hookers leading the way, there’s a smattering of edge and middle forwards as well. A few key takeaways from this list:

– Cameron Smith is in a class of his own as a hooker even at 37, but it’s clear he’s more halfback than hooker with the ball contributing nearly 6 net points per game.

– Tom Startling is having an incredible season for the Raiders, and it’s a crime he wasn’t starting sooner. The return of John Bateman also has played a huge part in their strong recent form, with both players part of the six forwards who have a NPRF average higher than 2.

– Warriors rookie Eliesa Katoa was an incredibly damaging runner for the Warriors, at least when he was fully fit and not carrying injuries into games, adding 2.33 points per game.

And finally, we’ll take a look at the bottom 20 for NPRF segmented for forwards only this season.

Here we have the Eye Test^tm confirmed, Bryce Cartwright is quantifiably the worst defensive forward in the NRL, with he only player in the league averaging over 2 per game conceded. A number of Broncos also make up the top five, along with another reported turnstile in Coen Hess.

It’s also worth nothing is that volumes of tackles don’t equate to quality defense, as Cameron McInnes’ appearance on this list would show. And I wonder how long Josh Jackson will continue to live of his good defensive reputation that was acquired in 2015?