The April premiership

By the time you read this, it should be April, which means the NRL is on the cusp of handing out the April Premiership, the third most important, and the least existent, trophy of the season.

The term “April premiers” seems to have a few variations in meaning. Typically, it’s a team that starts particularly well and is either on top of the table sometime in April. The premiership can be awarded either at the start or end of the month, depending on the individual. The title also connotes fading badly on the run home. Its awarding can either come out of a semi-humourous hope of rivals that the club will eventually come back to the pack or out of past experience.

Embed from Getty Images

St George Illawarra have embodied this in recent years, having gotten out to lightning starts, 6-0 in 2018, 6-1 in 2017 and 6-2 in 2015, only to fade in the back half of the season. In those seasons, the Dragons finished seventh (15-9), ninth (12-12) and eighth (12-12) on the ladder respectively. Remember in 2017 when they just needed a win over the Bulldogs to make the finals and couldn’t do it? The Cowboys took their place in the top eight and went to the grand final. That’s pure April premiership right there.

Those kinds of incidents stick in our memories but I don’t think they’re actually that common. I think whichever team is on top of the ladder by the end of April will probably still be there or thereabouts come September and thus “winning” the April premiership should be considered a good thing. Let’s take a closer look.

mb banner 3

I’m going to define the April premiership as the first eight rounds of the season, which sometimes dips into May, and I’m going to limit the dataset to all NRL seasons, 1998 to 2018. Our first stop compares the number of wins in the first eight rounds with the number of regular season wins for that team:

april premiership wins.PNG

It shouldn’t be much of a surprise that there’s a decent correlation (coefficient of correlation = 0.45) between early season form and whole of season success. The more games you win early, the more games you’ll probably win overall. This relationship only strengthens as the season continues, rising to 0.72 by round thirteen.

That tells us that the April premiers aren’t likely to fall off their perch atop the league but this graph does show the range of possible outcomes. A number of teams have registered six wins in the opening eight rounds but failed to win another six games in the following eighteen rounds to get themselves into the finals. By the same token, some teams get to the end of April close to .500 and then go on a tear to finish on 20-4, about the best you can do in the face of the general chaos underlying rugby league. Interesting that the only 8-0 starters, the 2012 Storm, didn’t finish at the top of the ladder that year but did win the premiership.

The worthwhile takeaway is the spread of teams above and below the twelve win mark. In the NRL era, teams with eleven wins or fewer are typically playing September footy thanks only to salary cap penalties, which was the case for the Raiders in 2002 (10-13-1) and Titans in 2016 (11-12-1). Strangely in 2002, the Dragons also qualified for the finals with 9-12-3 and the Eels in sixth place with 10-12-2, which is a nod to the dominance of the ’02 Bulldogs. The only “legitimate” examples I could find where the 2004 Raiders and the 2007 Broncos, both finishing on 11-13 and both wiped out in the first week of the finals.

What the graph doesn’t tell you is how frequently teams with a given record in April go on to make the finals. That’s why we have this table:

finals probability.PNG

The way to read this is at any point during the April premiership (rounds 1 to 8), you can take your team’s record and find their likelihood of making the finals. If your team starts 4-2, you go down the first column to 4 and the bottom row to 2 and where they meet, you see 80% of teams that had that record went on to make the finals:

finals probability highlighted.PNG

This is across the entire NRL history, so includes the top ten finals in 1998, several salary cap scandals and a few years between the return of Souths and the introduction of the Titans where there was a bye every week. Teams in those seasons might not have played a eight games by round eight. A case in point would be the 2006 Rabbitohs, who started 0-13 but had an early bye, so show up in this table under 0-7. I probably don’t need to tell you that they didn’t make the finals and in fact finished 3-21 with the wooden spoon.

Overall, it mostly reinforces the idea that being April Premiers is pretty good. If you can get seven wins in your first eight games, you are almost certain to finish in the top eight. With six or five wins, you are still extremely likely.

You might notice a few quirks. Moving from a 6-1 record to 6-2 (i.e. losing a game) actually improves your finals chances from 83% to 86%. Or going from 5-1 to 6-1 (i.e. winning a game) reduces your chances from 94% to 91%. There’s a reason for that. Here’s how many examples we have of each record:

finals numbers.PNG

At the extremes, we see the sample size issues clearly. In the 5-1 to 6-1 change, for example, we have 16 of 17 5-1 teams making the finals but only 10 of 11 6-1 teams, which is a lower proportion but it’s really only one team that has missed out. That team was the 2008 Titans, who were 6-2 at the start of May and only 10-14 four months later. We also see that the number of 1-0 does not match the number of 0-1. You can blame that pesky bye for that too.