The return of NRL expected points and Souths poor second half field position

First up this week is a bit of good news. I’ve managed to be able to produce some Eye Test Expected Points (ETxP) data for a few games this season, specifically the Round 2 and 3 losses by South Sydney. Hopefully I’ll be able to continue producing them throughout the season but no promises.

Secondly, one of the reasons we’re focusing on Souths, is that their two losses this season can be traced back to a lack of quality field position in the second half and their inability to contain their opponents in that same period.

For those new to the site, or need a refresher on the Eye Test Expected Points (ETXP) model I developed two seasons ago, the full explainer is here. The quick version is that we divide the field up into 4×4 metre squares, with a point value based on the historic probability of scoring from that location. We add every one of those instances together and we’ve got a game expected point total, which we can then compare to actual results and see who is over or under performing based on field position and the quality of that possession.

These sort of trends can be easily seen in the ETxP charts I’d been producing which showed a teams cumulative expected points over the course of 80 minutes. A flatter line indicates low quality possession and field position, and a steeper line shows when and how much strong field position a team had. It’s a great way of visualising the flow of possession and field position over the course of a game. Spoiler: Souths have had some very flat second half expected point lines.

But before we get to looking at their games by expected points, we’re first going to is to check their play the balls inside an opponent’s 20 metres, as well as how many times their foes played the ball inside Souths’ 20 metre zone, specifically for second halves.

Thanks for reading The Rugby League Eye Test. Subscribe below to receive new posts in your inbox

In Round 1, the Bunnies ended up 27-18 winners over the Sharks with a net +5 play the balls inside their opponents 20 during the second half.

In the second half of their Round 2 game that number swung wildly as Souths gave up 26 tackles inside their own 20 to Penrith. They only managed 3 themselves, resulting in a net -23 tackles inside 20, and it was reflected on the scoreboard with a 16-12 loss.

Round 3 was much of the same, but on a smaller scale as Souths played the ball 12 times in the Roosters 20 metre zone, but gave up 20 in their own red zone, for a net of -8.

That makes up a total net -31 tackles inside an opponents 20 metre zone for the past two games for Souths in the second half alone. Once you add in the Cronulla game, for the season they’re at -26 in second halves, only ahead of the Gold Coast Titans at -38.

What is probably most impressive from these numbers is that Souths weren’t punished more on the scoreboard, which we can see if we look at their ETxP numbers for the past two games.

Under expected point (ETXP) totals, Souths should have been behind on the scoreboard by 29.04-10.28 to Penrith on the back of the 26 tackles inside the Rabbitohs own 20 metre zone that the Panthers had in the second half. A reminder the expected points is what an average team should have scored with the same possession and field position across the course of a game.

At half time of that game, Penrith would have been ahead 11.04 to 6.72 on expected points, despite leading the actual game 10-0. The glut of quality possession for the Panthers that was referenced above in the second half should have resulted in more points, but Souths managed to hold on and make a game of it.

For Round 3 it was slightly different in that Souths streaked away early to a 10-0 lead on the back of two quick tries without a lot of sustained field position. And despite being on top for the opening quarter of the game, Souths were only ahead on for about 6-7 minutes around the 20 minute mark as we can see in the following chart.

The issue for this game, like the game prior, was that other than a set after half time Souths struggled to move the ball into strong field position and had to again rely on some late heroics to put themselves in a position to win.

One thing to keep in mind is that ETxP generally correlates highly with tackles inside 20. That’s with good reason – the majority of tries are scored late in tackle counts whilst teams are in good field position.

This is why the Bunnies expected point total creeps up slowly in this game – Keon Koloamatangi’s try was scored from about 10 metres out, but it was on the second tackle, which has a much lower probability of producing a score than a play on the 4th or 5th tackle, hence a lower expected point value for that play.

The correlation between ETxP and the number of play the balls inside 20 metres is quite high, which can be seen below in the plot of ETxP v play the balls inside 20 in every NRL game from 2016 to 2021.

The r-squared (basically how well the line fits the data) for this chart is 0.83, and as it runs on a scale from 0-1 this confirms what I had mentioned before that it correlates highly. If you have the ball more close to the line, the more likely you are to score. Hardly rocket science.

But unlike expected points, the number of actual points scored does not correlate well with tackled inside 20 metres, which we can see here.

The r-squared for this plot is much closer to 0 than it is 1. Possession and strong field position doesn’t always equal points, as evidenced by the safety first Dean Pay coached Bulldogs who barely threatened the try line no matter how many times they had the ball in their opponents red zone. Or you have teams like the Melbourne Storm who were (prior to 2023) threats to score from anywhere on the field and thus would consistently outperform their expected scores. But you can’t score if you’re not giving yourself the best chance to do so.

And why are Souths struggling to get downfield or fortify field position?

I’ve noted before that completion rate is a bit of a myth (it is), but Souths are currently second last in the NRL in completion rate with 69%, only ahead of the Tigers at 68%. They’re also averaging the third most errors in the competition at 13.3 per contest, behind two winless clubs in the Tigers (15.3) and the Eels (13.7).

Again, this shouldn’t be a surprise for Bunnies fans as I noted they consistently had the highest error rate (possessions per error) in the NRL in 2022, generating one about every 30 touches of the ball.

This has continued into 2023, with Souths making an error every 29.32 touches of the ball, only ahead of the Gold Coast at 29.0 and right in line with the Tigers at 29.36.

Interestingly the two “safest” teams with the ball this season are the Dragons at an error every 52.93 touches and the Warriors (!?) at one every 48.26 possessions.

It’s very early in the season and whilst there’s plenty for Souths to work with when they have the ball, but the positive side is that their defense has been keeping them in games. In both the Penrith and Roosters losses, their actual points conceded was below the expected opponent scores. The Panthers scored 45% fewer points than expected against Souths, whilst the Roosters score of 20 was 19.5% lower than expected.

And whilst they haven’t enjoyed great field position as those games progressed, the very late tries they did score pushed their actual totals above expected for both games as well. I noted their early season struggles in 2022 and they overcame those, and with 21 games remaining there’s still plenty of time to remedy it. They even had issues last year in the opening round scoring on a considerably worse Brisbane side, posting just four points against an expected 17.82.

Having some of their forwards return from injury and not losing a starting middle in the opening minutes of a game would probably help, but like last season fixing unforced errors would be the lowest hanging fruit.

Just three teams affected by the 11 day concussion stand downs

Last week, friend of the site and headgear connoisseur NRL Physio posted a fantastic thread about the NRL’s new concussions rules including a mandatory 11 day stand down period.

The introduction of this stand down period resulted in the expected overreaction that “players could miss two games” after a concussion, but if you read the thread linked above the chances of that happening are miniscule.

If you didn’t see Briens great therad, I’ll give you the cliff notes here. Looking at each team individually, there’s just over 400 instances of teams playing a game this season (17 teams x 24 games). Of those 400 instances, I found there were only three occasions where a team would play two matches within in 11 day period following a game.

There were another 35 instances of teams playing two games in 12 days, so you can clearly see why 11 was the magic number. Of course, those who play Origin could increase the number of occurrences of this happening but given the way players are routinely stood down after Origin that’s probably less likely to happen, and would only be affecting a small number of players (the unlucky few out of 34 who suffer a head knock during the three games).

Back onto the three occasions I mentioned above, they were Parramatta in Rounds 6-8, the Warriors in Rounds 8-10 and Manly in Rounds 21-23. The

Manly’s two in 11 days following a match are relatively benign – the furthest they travel is to Wollongong in the middle of their trio of games. Their games are 23/7 at home v Cronulla, 29/7 away against the Dragons and 3/8 against the Roosters at the SCG.

Parramatta’s is slightly more taxing. It starts Monday April 10 against the Tigers in their traditional Easter clash. They then meet the Bulldogs at home on April 16, before backing up 5 days later against the Broncos in Darwin of all places. Eels fans will be mortified at the thought of Brad Arthur’s bench rotations in a mid-April game in the north where temperatures could easily be 28+ and 80% humidity.

But the most appalling of all of these is the stretch of games the Warriors have in late April/early May. Starting on ANZAC day they face the Storm in Melbourne. Five days later they’re back in New Zealand to meet the Roosters at Mt Smart. Six days later they’re playing Penrith in Brisbane as part of Magic Round. That caps off their two games in 11 days.


Another six days later they’re back in Sydney to play the Bulldogs on May 12. That’s four games in 18 days and backing up the ANZAC day clash with three in 17, with a significant amount of travel in between. They do end up with the bye for Round 12 and don’t play again until May 27, but it’s a brutal stretch of travel and matches that should have been avoided.

I get there’s commitments with traditional ANZAC day games but given all that the Warriors did for the game during the last three seasons, surely the NRL could have done them a solid and spread out those games a bit more?

If you enjoyed this post please consider supporting The Rugby League Eye Test through a donation via one of the links below.