“Offense sells tickets, but defense wins championships”. An exciting offense may indeed sell tickets (as the 2016 Raiders did, attracting their highest crowd average since 1995), but does defense really win championships? Some hacks put it forward as assumed fact without question (as we ourselves did in our 2016 Finals Preview). Others claim it’s a self-perpetuating myth. So, which is it? Do elite defensive teams really outperform elite offensive teams? And if so, why? Let’s take a look at the Melbourne Storm to find out.
The Melbourne Storm always win
Ok. It may be a slight exaggeration to say the Melbourne Storm always win. However, they certainly perform pretty darn well, having not missed the finals since 2002 (with the notable exclusion of 2010, in which they were forced to play for no points. And for the record, if they had played for points, they’d have finished 7th*). So, what makes the Melbourne Storm so consistently good? (And no, wise-guy, the answer isn’t paper bags full of money. And a boat.) The answer, it turns out, is their defense.
In order to test the hypothesis that elite defensive teams outperform elite offensive teams, we looked at the finishing positions of the four best performing offenses and defenses since 2008 (we chose 2008 because that’s as far back as the handy nrl.com sortable ladder goes). For 2016, we used our VOA Total Offense/Defense rankings, for earlier seasons we were forced to use points scored/conceded (we acknowledge that we consider this a flawed metric of offensive or defensive competency, but in the absence of other data, it’s the best we could do). The results are as follows:
|Elite Offenses||Elite Defenses|
|Finishing position range||1st-12th||1st-8th|
|Won the Comp||6/9 times (67%)||8/9 times (89%)|
|Missed the Top 4||16/36 times (44%)||8/36 times (22%)|
|Missed the Top 8||7/36 times (19%)||0/36 times (0%)|
As you can see from the above table, on average, elite defenses do in fact outperform elite offenses. Only once in the last 9 years did a team manage to win the competition without an elite defense (and had Ben Hunt caught the ball, it may have been none from nine). That isn’t to dismiss the value of a great offense entirely – it’s worth noting that of the 8 premiers who featured an elite defense, 5 of those also had high-powered offenses (highlighting the benefit of having a well-rounded side).
Rather, the point is that, on average, elite defenses produce better results. For example, in the last 9 years, an elite defense has never missed the Top 8. This is interesting for two reasons. Firstly, elite offenses can’t say the same (they miss the finals with semi-regularity). Secondly, it helps explain the continued success of the Melbourne Storm. Though the Storm have featured an elite offense in just 5 of those 9 years, they’ve featured an elite defense every year. And it’s on the back of that defense that the Storm have enjoyed continued success.
Why sometimes points just ain’t enough
So we’ve established that yes, defenses do indeed win championships (or in the very least, they place teams in a strong position to contend for them). What we don’t yet understand is why that’s the case. In an attempt to solve that particular riddle, we looked at all the offensive and defensive results of 2016’s best performers in their respective categories. The results point towards a possible explanation.
|Elite Offenses||Elite Defenses|
|Avg Points Scored/Conceded||24.67||15.34|
|Range of Points Scored/Conceded||6-60||0-48|
|Range in best 90% of matches||12-60||0-26|
As you can see in the above table, elite offenses score a lot of points, and elite defenses concede very few points (no groundbreaking news there). What’s of interest to us though, is the range of results for each.
The range of offensive output for elite attacking teams was 54 points (between 6 and 60), whereas the range of points conceded for their defensive counterparts was 48 points (between 0 and 48). Slightly smaller for defensive sides, but the difference is marginal. However, if we discard the worst 10% of performances for each side (since anyone can have a bad day, for various reasons), we begin to uncover something.
- 90% of the time, the range of points scored by elite offenses was 48 points (between 12 and 60). However;
- 90% of the time, the range of points conceded by elite defenses was just 26 points (between 0 and 26).
This is significant, and may point towards why defensive teams enjoy more consistently good results.
Consider this: when an elite offense plays a terrible defense, they may score something like 60 points. However, when they run into elite defenses, they can be contained to just 12 (or even less). What this means is that the teams with elite defenses can be required to score just two or three tries in order to win, against even the most exciting of offenses.
On the other hand, the difference between a good day and a bad day for an elite defense is much smaller. Yes, they can shut-out bad offenses (and between them they did exactly that 10 times last year), but more importantly, they can almost always contain an elite attacking side to 26 points, and often even less. Which means that regardless of their own attacking ability, they’ll typically be competitive in more matches (as they don’t get blown out).
In practice, this proved to be exactly how it played out. The following are the points scored by elite offenses vs Melbourne (the league’s best defense) compared to against Newcastle (the league’s worst defense):
For comparison, now let’s compare the variation in points conceded by elite defenses vs Canberra (the league’s best offense after 26 rounds), and Newcastle (the league’s worst after 26 rounds):
From these two tables we can see the manifestation of the above theory:
- Despite an aptitude for hammering the Knights, when elite offenses ran into the stingy Melbourne Storm defense they were generally well-contained (3/5 times they were held to just 2 tries or less); however
- When facing the high-octane Canberra Raiders, the elite defenses fared better – they were able to contain the Raiders to 26 points or less most of the time (and to three tries or less 50% of the time).
In practice, it makes sense that the ability to run all over Newcastle doesn’t necessarily translate to September success. After all, Newcastle aren’t typically playing in September (sorry Knights fans). Yes, it’s important to put away rubbish teams (we explained why here), however whether you beat them by 30 points or 50 points is neither here nor there. What matters is the ability to stop your opponent (regardless of their attacking aptitude) from running up the score against you, thereby keeping the match competitive and giving you the best chance to win. And that’s why defense wins championships.
*For our purposes, we’ve treated the 2010 Melbourne Storm as coming 7th throughout this article, since this where they finished in terms of wins.
Postscript: It’s worth mentioning that it’s likely that gameflow is a small, but non-negligible factor in the inferior results for high-scoring teams. Bad defensive teams are more likely to be losing during matches, and then necessarily have to score more points in order to compete. Or to put it differently, while scoring a lot of points won’t make you lose matches, if you’re losing matches you’ll need to score a lot of points.