Rugby League World Cup 2026 - A small data visualisation project about the England Men's team

England Match 1: 

Because of poor decisions by International Rugby League I couldn't watch the match anywhere so was listening to it on the radio. Truly, Robbie Hunter-Paul is an international treasure. 

Not entirely sure why the English press were saying it was a shock, other than them never giving Tonga the credit they are due. If they'd said the first Tongan try was a shock, that would be true, it was against the run of play but after that it was a lot more even. 

I'm a Saints fan, spiritually, I prefer what McDermott was trying to get this England team to do rather than the kind of rugby Shaun Wane had them playing. 

On the other hand, Wane looked at the players he had and played them in a system that worked to their strengths. If you're going for quick and clever, why are you not playing Mikey Lewis (never mind Jake Connor who didn't even get picked)? George Williams and Harry Smith have their uses, but quick and clever isn't it. 

I don't blame McDermott for some of the errors because the England players appeared to develop poor decision making skills every time they got into the Tongan 20 meters, but there was some oomph lacking. I don't know if McDermott's made that choice because he doesn't think this front row can bring the biff, and they're big units but they don't scare me the way say Burgesses plural, or Elliott Whitehead or James Graham did. (Challenge, go a whole thing without mentioning Jamie Graham, glorious failure). So fine, you aren't doing wham bam rugby, you're running the ball instead, but then you have to involve the wingers, especially if one of them is Dom Young. And that seemed to be missing. 

So England now have to rely on beating France and Lebanon and hoping the other results go their way. I'm going to be really upset about either England or PNG not making it, I fear. 

Data visualisation project: 

Much like the last World Cup, I'm doing a small data visualisation project to see if I can spot patterns in player selections and how the team performs. 

One game doesn't give much data, but here is what I have so far. 

England's point-scorers by name and time Very little there other than Savala is the chosen kicker. 

Because there were only two tries for England, the chart of which players are on the pitch when England score is uninformative. The one for "who is on the pitch when England concede" looks like this. You can see Thompson and Havard coming on and off, and Jez Litten being given the last 20 minutes. 

The two equivalent bar charts for which players were present for point-scoring and point-conceding moments are below. You can see who were rolled on as subs, King, Nsemba, Litten and O'Neill. According to nrl.com King and Nsemba swapped on for each other, and obviously Litten and O'Neill are a straight swap. 

The bar chart of who was on for point-concessions is more interesting.  

The larger number of events reflects both the result and the number of 2-point penalties Tonga scored. In the chart you can see the backs who didn't get subbed off, and the forwards who were. 

The two matrix charts of who was together when England scored and conceded are below. There's not yet enough data for the equivalent network graphs. Because England only had 4 point-scoring moments so the chart is simple, with the players who were together for all 4 points in purple and the ones present for fewer in orange. The one exception is Max King, who is in very pale yellow where he crosses over with Nsemba and O'Neill, suggesting they didn't play together as often when England scored. It makes sense for Nsemba because he was subbed on for him, and O'Neill got subbed off for Litten at around the same time.

Then there's the chart for when England conceded: There's a distinct split between the dark purple, the players most often together when England concede, section, and the second darkest (the darker orange). The latter includes O'Neill, Walters and Nsemba, who were substituted on and off rather than remaining with the main group throughout. Ben Currie has the palest line because he only came on at minute 68. 

Harvard and Walters were the players present for relatively fewer point-conceding moments, being in the darkest purple section for scoring but not for conceding (yes, I know I'm being nice about Wigan players. Yes, it feels wrong). 

It'll be interesting to see what patterns develop over time.

Rugby League World Cup 2026 - Women's Group Stage Network Diagrams

Here are the network diagrams for the group stages for the Women's Rugby League World Cup 2026. 

First, let's look at the non-labelled version of the diagram.  

Which teams are which? France are completely on their own which doesn't bode well for them. 

The other two teams sticking out at the bottom are England and Wales, which also doesn't bode brilliantly. 

Of that top cluster of 5 teams, Fiji, Samoa, New Zealand, Australia and Papua New Guinea, Papua New Guinea are the team sticking out slightly to the right. 

The other thing I've noticed is that while, in the equivalent men's diagram there is a mixture of club teams that the players are from, in the women's most of the national teams have one club team that most players are from (Papua New Guinea are the only team where that's not really the case). 

The community view also does something fun.  

Normally, there's the same number of communities in this view as national teams, or maybe fewer when the teams are deeply enmeshed due to links through club teams. This time, the North Queensland Cowboys are an extra community, I think because they link 5 of the national teams. 

Because of France being separate, it's really hard to identify the central team, particularly the central national team because of the way they are clustered. Central club team is probably Sydney Roosters. 

The Roosters are also the club team with the most representatives, alongside RC Lescure-Arthes XIII, with 12. They are followed by Gold Coast Titans, Wigan Warriors and Canterbury-Bankstown Bulldogs on 11, and Cronulla-Sutherland Sharks on 10.

Rugby League World Cup 2026 - Men's Group Stage Network Diagrams

If you would like a nice, in-depth and reasonable preview of the World Cup, may I recommend this from the Maroon Observer. 

Here I am all about pretty pictures and bemoaning there only being 2 Saints players at the World Cup and one of those spent most of the season injured. 

What do the network diagrams look like?  

The two teams that are sticking out are Papua New Guinea and France. Teams stick out if their players play for clubs who don't have a lot of other participants playing for other national teams. If the previous pattern that sticking out = weaker remains true, this means Tonga have been stuffed over by the draw and have a harder route to qualify for the semifinals. They would need to beat one of Samoa or England to qualify, while England and Samoa only have to beat Papua New Guinea and France. No diss on Papua New Guinea intended, because I really want them to do well. 

I think Papua New Guinea are pulled out because so many of their players play for London Broncos who don't have any other representatives while the Cook Islands players are spread over more club teams so the weight of any individual team doesn't pull them out in the same way. 

The club team with the most representatives are the New Zealand Warriors and the Sydney Roosters who both have 13. They are followed by the mysteriously un-city-ied Dolphins and Penrith Panthers on 12, and Manly Warringah Sea Eagles, Melbourne Storm and Toulouse Olympique on 11. 

Because 8 of the teams are clustered together, Fiji are the team in the geometric centre of the graph, while Tonga and Australia are probably the "gravitational" centre. 

Hull Kingston Rovers and South Sydney Rabbitohs are the club teams closest to the physical centre. 

The other thing that strikes me is a suspicion that the NRL must be relatively evenly matched because the Australian squad aren't dominated by one club team, unlike the way say France are by Toulouse Olympique. (Until the first match, I promise not to comment on the excessive number of Wigan players in the England squad. Brian McDermott, you'd better have a good reason for not starting Jez Litten.) 

The one upside to there only being ten teams is that it's relatively quick to make the diagrams so I hope to have the women's and, data permitting, the wheelchair teams graphs ready shortly.

Formula 1 2026 - The Bahrain Grand Prix in Malaysia

Because F1 2026 makes very little sense. (Although I love that we're back at Sepang. It may be the only track I like because of how pretty the grandstands are.)

Due to the delay between the formation lap nonsense and the actual start, the BBC only have from the second start onwards on BBC Sounds so I have had to resort to watching the Channel 4 highlight show. (I prefer the radio coverage, whole post on why coming soon.)

You know it's going to be one of *those* races when people go off on the formation lap. I frequently make jokes about having seen x, y or z before but I have never seen half the field prevented from functioning by computer failure before. "Starting procedure suspended," indeed.

And then people off in formation lap 2.

I did enjoy the BBC commentators chortling over the work the FIA had to do to ignore its own rules.

Bloody heck at Piastri nearly hitting Albon in the pitlane. It seemed to be Albon's day for it because I am really not sure Stroll should have been released from his pitstop like that.

George Russell only has bad luck.

The gaps between cars, even after the track became dry, were very small, and, because it's a race track, built for racing, there were overtakes. It's almost like they should race in places designed for it, not around regular roads.

Formula 1 2026 - Azerbaijan Grand Prix

It turns out me not watching races doesn't always equal a Ferrari win (busy looking at the Bayeux tapestry, report eventually).

Baku historically has either had an interesting qualifying or an interesting race. Qualifying was exciting, so I wasn't expecting much from the race. I was right not to. I was fascinated that the press all went with "Baku Chaos Bonanza", which is a very blogger take, but this race wasn't, except the Red Bull psychodrama. (And no, BBC, Max did not sound calm.)

BBC suggesting that driver to driver comms would be interesting. Probably, but also more beep than word. It would also end in thumping - see also, Norris and Colapinto in this race. I would also like to bless Mark Priestley for saying "Albon = alright" was the most important thing after his accident.

The other odd thing was the soft tyres lasting longer than the medium. Even on a track with little wear, that is odd. Were the sudden rash of incidents around lap 30 - the Norris spin and the Albon crash - due to tyres wearing out because people were waiting for a safety car and just stayed on their first set of tyres too long?

I hate the new men's rugby league world cup format as much as I hate all Montagues

I was preparing for the usual data visualisation (for the 2021/2 example please see here). Then I looked at the format for the men's tournament. 

Please assume I am here from here on in:

via GIPHY

Dear RFL, I had been nice about you for so long, and you provide me with this nonsense. (Full details here) 

Okay, due to everything, they've had to drop down to 10 teams. I can deal. I will ignore that going begging bowl in hand to the NRL ought to have solved that but okay. 

10 teams, you do two groups of 5, top 2 go through, all's fair in love, war and rugby league. 

But no, they give us a mutant set up with 3 groups, 4 teams in group A, 3 teams each in groups B and C. 

And then, group B only play the teams in group C while C only play B. 

The top two teams in Group A go through. For Groups B and C, though, the two groups are put together in one mega-group, with the top two teams from the combined results going through. And yes, it is possible for both the qualifiers from the group B/C table to come from one group. 

My, very simple, vision gives all teams 4 games. This gives teams 3 games. One of Tonga, Samoa and England will be going home after 3 games. 

In group A, Fiji are the team most duffed up by the draw. I mean, I will be crossing all that is crossable for them to beat either New Zealand or Australia but ... they've been given the shark pool. 

In group B/C, Papua New Guinea have been done over, while, as I said, at least one of Tonga, England and Samoa isn't getting a knock out match. I am more annoyed on Papua New Guinea and Tonga's behalf than England's. 

We're also skipping that, of the co-hosts, Papua New Guinea host 2 games, New Zealand host 1, and Australia get 15 games. Truly well distributed. (I do like that they're making France vs Lebanon and Tonga vs Samoa a double header. It may be the only thing I do like about the way this has been organised.) 

Trying to be positive, Papua New Guinea vs Samoa in Port Moresby ought to be a cracker in terms of atmosphere, ditto Samoa vs Tonga as the game that probably decides who qualifies out of group B/C. That's an excited rating of 2/15 for the group stages (England games, by definition, are not exciting, they're a torment). 

It's going to be an unholy mess, they've almost stacked it so one of the co-hosts can't make it through to the knock out stages and, barring a miracle, Australia can play the reserves in their last group stage game without it affecting anything.

Are the scrutineers picking on my driver? - Update after the Italian Grand Prix

The scrutineering data for the Italian Grand Prix is straightforward. Somehow, despite being the first car out, Leclerc isn't the least tested. 

The χ² for the race is 1.38, so any differences aren't statistically significant. 

For the season to date, it looks like this:  

The "tested less than expected" cars are definitely further from expected than the "over-tested" cars. I think that's because the Aston Martin and Cadillac teams have remained what we will politely call "not good", while the rest of the teams have come closer together in terms of results. 

The χ² for the season up to the end of the Italian Grand Prix is 4.32 so any differences aren't statistically significant, but the number is still increasing.

Formula 1 2026 - Spanish Grand Prix

The Madrid Circuit: Of course, if Barcelona has one, Madrid has to have one. Grand Prix in Spain are getting almost as plentiful as they are in the MotoGP.

But the race it reminded me of was Valencia (thanks to L, different L to usual, for helping me twig that it was Valencia I couldn't remember). And the reason it reminded me of Valencia was because Madrid was yet another street circuit where nothing happens.

(So little happened at Valencia that someone literally had to remind me those Grand Prix existed)

I think we just have to be honest about street circuits. If a circuit doesn't have a fair few spots where two and a half cars can get through there will be no overtaking. Circuits that are designed around normal roads will not have that. So if it's a street circuit, there will be no overtaking.

So stop adding more street circuits to the calendar!!!!

Ahem!

I was also amused and slightly charmed by how confused Ted Kravitz was by the existence of mega-sized conference centres like the one in Madrid. Possibly because some of my colleagues have actually been to the one in Madrid and I spend probably too much time in them with work.

The race itself:

The start was exciting, with me going "eep!", "oop!" and similar noises. Then it settled down. The track being that tight meant that any time anyone went for anything, that was the noise I made. But it also meant that almost no one went for anything because disaster was too close.

For once I can't really complain about Ferrari's strategy because they made the most of an opportunity. That curses and pins and voodoo dolls didn't cause another red flag or safety car wasn't their fault. Neither could McLaren be faulted for the unfortunate timing of the virtual safety car with regard to Norris. Mercedes's strategy for Russell was their own fault.

Sainz jnr was heavily hit by the home-town race curse.

It was a bit of a nothing race overall, but what else is expected from a street circuit?

Formula 1 2026 - Italian Grand Prix

I would like to pretend I had something more reasonable to say but ... I hate being a Ferrari fan. Why are they always like this?

I don't, in any way, blame either driver for the first corner - that's racing. I blame the team for not coming up with a working strategy for the one remaining car.

The racing was fun while it lasted, but I trust this has finally ended any belief George Russell has that there is driver equality at Mercedes.

Are the scrutineers picking on my driver? - Update after the Dutch Grand Prix

For the purposes of this series, the main thing for the Dutch GP was Hadjar missing it due to injury. Lawson moved up to Red Bull, and Tsunoda was temporarily back in whatever Toro Rosso are called this year (I think it's Racing Bulls, I can't keep up with the constant name changes). 

I didn't think it would affect the actual race data, because hey, that was still just 22 drivers. 

Dutch GP race data itself: Russell was the most scrutinized compared to expected, followed by Hulkenberg and Lindblad, with Stroll, Ocon, Bottas and Bearman the least scrutinized. 

What I don't understand is why Verstappen isn't on the under-scrutinized list given he was out before the end of the first lap, before Bearman who is on that list. 

Most peculiar. 

Any differences weren't statistically significant, with a χ² of 1.38. 

The thing I thought the Hadjar swap out-Tsunoda swap in would affect was the season-to-date ranking. But what I forgot was that, because it's observed vs expected, Tsunoda will still get a value for the one (or more depending on Hadjar's wrist) grand prix he participated in. 

The thing it changes is the χ² critical value, because Tsunoda's appearance increases the degrees of freedom in the data from 21 to 22. The critical χ² value goes up from 32.671 to 33.924. 

The season to date chart looks like this:  

At the top, Bortoleto still stands out as not being like the other cars up there. 

At the bottom, it's still the cars that don't finish a lot, with Stroll and the two Cadillacs down there. 

Tsunoda's car has been tested exactly as much as expected after 1 race, and so has Piastri's after 12 races. 

The χ² value continues to go up, it's now 4.03, still nowhere near significant but rising. I think the cars being tested less than expected are the stronger contributors because their standardised residuals have larger values. 

We also didn't have a random check of a top 10 finisher, so the evidence is really piling up that that is a truly random check that only occurs sometimes and the early-season pattern was just a quirk of the numbers.