Saturday’s 18-16 win was the first time St George Illawarra have won a game in Melbourne this century, and the Dragons now have a win over the top two sides in the competition.
That’s something the Roosters haven’t done in five seasons.
Friend of the site JonnyForeigner has his usual tremendous match review up at Storm Machine, which is well worth subscribing to, even if the result didn’t go the way he would have liked.
Of course, when they played a stronger Panthers side outside of the Origin period, the Dragons were thrashed 46-10, and they’ve been very up and down this season. Which is better than their 2023 output, where they were 0-12 in away games, and only won one game after the 10th of June, with that occurring against the Tigers, 18-14 in Wollongong.
Improvement isn’t always a linear path, but there are signs of life being shown that should make Dragons fans positive for the future. They’ve won 10 games this season, including 5 away from home(s) with victories in Melbourne, Penrith and Brisbane. The problem is that they’ve also been extremely bad on occasions as well. Playing at a consistently high level is the next step.
Given these results, I thought I’d look at where head coach Shane Flanagan has fixed in their performances this season, and where they still need to progress.
Last season I wrote how there were two key areas Flanagan needed to improve when taking over the Dragons. The first was their woeful edge defense, and the second was finding some punch through the middle and a backrower who can bend defensive lines.
They haven’t entirely solved the second one yet, which is a bit hard considering their current list, but they’ve definitely improved on the first point.
By tightening up their edge defense, he’s changed how they’re conceding points. The Dragons still far from perfect, or even good though. They’re conceding 25 points per game. and having yielded 100+ points in two games to the Roosters as well as 46 to the Cowboys, 38 to the Dolphins and 44 to the Bulldogs. They’re not conceding them the same way they were last season though.
One way to spot their defensive adjustments is that the average distance a try is conceded from by the Dragons has dropped from 27.0m last year to 18.2 metres this year.
That’s an improvement from fourth worst in the competition to fourth best in just twelve months with limited roster changes. They did this by more than halving the number of tries conceded from their own half (14 to 6) and dropped the number from 50-20m out (21 to 15).
They are still allowing the same volume of tries inside 20m (57 in both seasons), but they’ve cleaned up their defense further out. Part of their try line stagnation stems from the fact they concede the second highest percentage of set restarts on their own line.
46% of all Dragons six agains come 0-10m out from their try line, only trailing the Cowboys at a woeful 53%. That Dragons number is up on last year as well., where it was just 32% (the Cows were still at 50% last season). As noted with the Sharks a few times this season, defending repeated sets on your own line is tough unless you’re fundamentally perfect, and the Dragons (and Cowboys for that matter) are certainly not.
Their opponents this weekend have done even better, with Canterbury going from 23 tries allowed from their own half to just six, and 22 from 50-20m out to also six. Their try line defense has been nearly impenetrable too, with just 38 tries scored inside 20 metres, down from 64 up to Round 22 last year. With the form of both teams based it’s no wonder this weekends game is a sell out.
Delving further into how the Dragons have cleaned up their edge defense, Stats Insiders try location analysis tool shows the Red V have gone from conceding 54% of tries down their right edge to 35% this year. Again this points to the club limiting scoring opportunities out wide, especially down the right side of the field from longer distances.
With the ball there’s still some work to go. One of the lowest average metres per try in the competition at 18.8m, down from 23 in 2023. Just 25% of tries scored by St George Illawarra come outside 20m, 4th lowest in the NRL this year. They’re still very right side dominant, with 49% of tries down that side of the field, only slightly lower than the 55% they posted a year ago. They may be conceding a lot of points on occasion, but not in the same way they were under the prior regime.
What this tightening of their edge defense has done is keep them in games for longer (when they’ve been able to defend their line). Looking at their average margin, on average they’re not more than 2 points out of a game until past the 70th minute, which is better than any season other than 2017/18 in the last nine seasons.
Some of that is due to big wins and big losses cancelling each other out, but wins against the Warriors, Penrith, Manly, Melbourne and Dolphins have proven they can show up in defense when they want to. The next step is consistency, which will come through fixing some of their issues restricting yardage, which we’ll get to shortly.
These up and down performances also show in up in their percentage of minutes leading or trailing, as shown below.
This year they’ve spent 43.5% of the time leading and just 39% trailing. That’s a strong improvement on 54% trailing last year, and they’ve been hovering at 50% or above for last 5 seasons.
The Dragons heat map of opponent play the balls in 2023 had a lot of possession in their own half which looks good.
But they were conceding a lot of points, and it was partly from teams starting sets in their own half because they were scoring regularly and returning the ball from kicks, not because the Dragons were doing a good job of keeping them there.
Their 2024 heat map with the ball is different. There’s a lower intensity in their own half but higher defending their own line. With the reduction in long range tries the club is defending more on their goal line (21% of play the balls up from 17%), which is shown in this seasons heat map above.
This goes along with the drop in long range tries mentioned above. They’re stopping teams scoring from distance, but they’re not stopping them advancing the ball into the red zone, something we’ll get into shortly.
With the ball the difference isn’t as pronounced. Slightly wider in their own half, more evenly distributed with a small shift to more left side play the balls.
With the change in possession distribution and cleaning up of their edge defense, this year they’re conceding only 17% more points than the expected value of the field position their opponents have, a huge drop from the ghastly 67% above expected they were giving away at the same point last season. There is definite improvement here, which is tied into stopping those long rage tries down an edge that we discussed before.
Looking at it over the past nine seasons you can see the downward trend from last season as well. They may not be back to their defensive best of the late 2010s, but they’ve reverted the upward trend of the early 2020s.
Splitting the above chart into performances in wins and losses for 2024, we can see another story emerging. The charts are below, first wins then losses.
In wins this season, the Dragons give up the second least number of points versus expected points at -38%, only trailing the Sharks. Even the Panthers and Bulldogs in their wins aren’t better than 30%, indicating that when they apply themselves St George Illawarra are clearly able to play some very solid defense when they want to. Also note that even in their wins this season, the Eels were giving up more points than expected.
There is some context missing here in that the Dragons also give up the third highest expected point average per game, shown below. Higher on the chart means you concede more field position with a strong probability of scoring a try, and further to the right means you have possession where the probability of scoring is high.
The problem is that in their losses, as seen above they’ve given up 84% more points than expected, third worst in the NRL. Which is why the volume of field position they give up is their greatest concern now they’ve fixed up some of the leaks down the edges of the field.
Either the Dragons show up and look to grind away a defensive victory, or they open up the floodgates and give up a score like the one against Penrith in Round 21. There’s very little in between at the moment, and I’m sure that’s an area Flanagan will be focusing on.
You may also notice that even Penrith and Canterbury allow fewer points than expected even in losses this season. There were only two teams who did that last season, Penrith and Brisbane. Don’t get over excited yet Dogs fans.
The overall work on defense isn’t done yet. They might have stopped the long range tries but their other issue currently is stopping yardage. The Dragons are second worst in the NRL for net metres per run (metres/run gained minus metres/run conceded), only better than South Sydney.
Part of that ties back to their inability to gain metres, as they’re second worst again in the NRL for total metres per run due to being last in pre contact metres. No one is hitting the line quickly at speed for the red and white.
They’re also last in the NRL for post contact metres conceded, and the worst team in the NRL for allowing metres past the play the ball (not run metres). Add those up and you get the situation they’re in with winning yardage battles. Opponents know this as well, which is why they focus so much traffic through the middle of the field.
The other area I mentioned last year that Flanagan will want to improve is having some forwards who bend the line. Last year the best in this regard was Ben Murdoch-Masila who was in the 67th percentile for metres per run of all forwards, meaning the best forward the Dragons had was behind one third of the competition. Most clubs would have one or two players pushing the 80tth
This season, things have improved somewhat, mostly thanks to the addition of Luciano Leilua.
The former Cowboy is in the 98th percentile (among all forwards with at least 20 runs) of metres per run this season at 9.89 metres per carry. Jack De Belin and Jaydn Su’a are only in the low 20th percentiles here, despite being above the 80th percentile in total metres gained thanks to the sheer volume of runs they make rather than the length of them.
They’ve added a bit of strike with Leilua, but the majority of their forward pack is still death by a thousand cuts rather than one or two blistering runs breaking open a line. The rumoured addition of Canberra’s Corey Horsburgh won’t change that much (18th percentile in metres per run), but he will provide some extra depth and fire.
Leilua’s 9.89 metres/run is over one metre per carry better than Murdoch-Masila’s mark of 8.79 metres per run last season, and he’s added a slightly to that number this year as well (8.99m/run). Michael Molo is another who has increased (17th percentile to 77th), but he’s barely met the completed runs threshold and I’m not sold on the sample size. It’s the same story for Toby Couchman, who is 92nd percentile for metres per run but has less than 50 carries all season.
Again there’s more work to be done here, but Flanagan looks like he’s getting as much as he can from a forward pack not entirely suited to the 2024 version of rugby league.
It would be unrealistic to have expect miracles from Flanagan this season, but somehow the Dragons could be playing finals football in a month. The fact they’re still in the hunt for 8th spot with a -90 point differential speaks to how even the competition (overall, not necessarily the games) has been this season. It took Cameron Ciraldo a full season to completely turn around the Bulldogs defense, and like the Canterbury coach Flanagan is setting the groundwork.
St George Illawarra’s run home after this weekend’s tough match up with the Bulldogs is very friendly – Gold Coast (H), Cronulla (H), Parramatta (A) and Canberra (H) to round out the regular season. With a two point gap to the Dolphins & Raiders, and even more to the rest, they just need to win out to claim a spot in the eight. It will most likely depend on which Dragons side shows up, the one that beat the Panthers and Storm, or the one that has conceded 40 points multiple times.
