More local derbies to spark interest was one of the features of this year’s Super Rugby competition, but has it lead to more tries than inter-conference matches?
Unless you are from stereotypical thinking in the northern hemisphere, rugby is about scoring tries. I wanted to know if the derby format impacted on tries.
The concept (and hype) for Super Rugby derbies were that each team in a conference plays each other twice, home and away. While not all teams play each other, there were an equal number of games played in the derbies and inter-conferences, eight a-piece (exception was the round two, Hurricanes versus Crusaders).
By averaging the number of tries in derbies and inter-conference round robin games, I found that some teams had a different number of tries in derbies compared to the number of tries scored in inter-conference matches.
Looking at Australian teams first, the Brumbies have scored the same amount of tries, an average of 2.0 in derbies (i.e. Australian teams) versus an average of 2.0 tries playing inter-conference (i.e. New Zealand and Australian teams) matches. Similarly, the Force averaged 1.6 tries in derbies versus 1.5 for inter-conferences.
The Rebels and the Reds averaged fewer tries in derbies than inter-conferences. The Rebels averaged 1.4 tries in derbies and 2.4 playing inter-conference games; and the Reds 2.4 tries in derbies and 3.3 for inter-conferences.
The Waratahs have a different outcome as they average more tries playing derbies, 3.4 compared with 2.6 tries in inter-conference matches.
The results for the South African and New Zealand teams show in every instance an average of fewer tries scored in derbies versus those scored in inter-conferences.
Comparing the average number of tries in derbies versus inter-conferences for South African teams: the Bulls averaged 2.3 tries versus 2.8; Cheetahs 2.3 versus 3.3; Lions 2.0 versus 2.5; Sharks 1.9 versus 3.3; and Stormers 1.9 versus 2.3.
Comparing the average number of tries in derbies versus inter-conferences for New Zealand teams: the Blues averaged 2.0 tries versus 3.4; Chiefs 1.6 versus 2.5; Crusaders 2.7 versus 3.4; Highlanders 1.5 versus 2.4; and Hurricanes 0.9 versus 3.3.
To compensate for the fact that not everyone played everyone, games were played in different weather, different time zones, different players, different refs and the fact I’m using averages, I applied 95 perc ent confidence intervals.
These are statistical formulas that, if you are unfamiliar, provide a range that you would normally expect another occurrence to fall within 95 percent of the time.
When comparing two groups, in this case the range of derbies tries versus the range of inter-conference tries; you are looking for whether there is an overlap between them. An overlap means no statistical difference. No overlap means a statistical difference.
The only team with a statistical difference was the Hurricanes whose tries in derbies range was 0.2-1.5 versus inter-conference range of 2.0-4.5. Since there is no overlap (i.e. there is a gap between the 1.5 and 2.0) there is statistical difference. Despite the Sharks average of 1.9 tries in derbies versus 3.3 in inter-conferences the ranges were 0.9-2.8 versus 2.0-4.5 respectively, so an overlap meaning not statistically significant.
So did the first year of the Super 15 Super Rugby produce more tries in derbies than inter-conferences overall? No. Did derbies produce fewer tries than inter-conference matches overall?
Yes, but in only the case of the Hurricanes was it statistically significant, and that might have had to do with other things off the field than on it.
RuckinGoodStats
Roar Rookie
Too hard for me to do live, but nice challenge. I do know 80% of tries are scored with the first 4 phases and the scrum is the starting point of most of the tries, lineout the second most starting point, its not broken play or from a game of foce back is my point. I'll have a look aas there are less games this weekend, but more media and people chasing me for data. I'll post it here rather than do an article???
Katipo
Guest
And generally the most spectacular tries involve multiple passing movements rather than repeated one-off phases. It's what we used to mean when we said 'running rugby' lots of running and passing. I'd like to see the statistics on how many one-pass phases compared to multi-pass phases per team (note the difference: passes per phase, not phases in possession)... I wonder which teams pass-the-ball the most and how that relates to tries scored/success? But hey, that's a big job. I ain't gonna do it. Would be interesting to know though. perhaps a quick count of the semi-finals passing is in order?
RuckinGoodStats
Roar Rookie
Cheers. I wish commentators would use them as well. That's the advantage of the roar. I've pitched to main stream media this and one I've got one Cruden v Carter using difficulty on place kicks and it was too stats heavy...
RuckinGoodStats
Roar Rookie
I'm pretty sure in one of my articles on the roar I've said Lies Darn Lies and statistics. You points are quite correct.
RuckinGoodStats
Roar Rookie
The NZ conference was the only one that had a statistical signficance using 95% CI, but didn't have the word count to put it in. And cheers for the comments.
RuckinGoodStats
Roar Rookie
Cheers. Wasn't sure about doing it, but am happy that people wanted it.
soapit
Guest
almost tries also would be good. theres a difference between an open game that has no tries through good cover defence and scramble and a grind where forwards bash into each other for 3 rucks then a game of force em backs until a theres a knock on. dont necessarily need tries but a bit of running around with ball in hand is essential.
spaldo
Guest
Good to see stats with significance. Should be compulsory whenever figures are mentioned. More please
sheek
Guest
While rugby is about scoring tries, it's more so about cleverly crafting tries against a well organised defence. i reckon even I would get sick of seeing 7-8 tries being scored EVERY games. Like anything else in life, tries are meaningless, unless they were worked hard for. So the number of tries scored in local derbies can be misleading. How many of the tries were well executed, catching the defence unawares? That would be a better guide. Also, the intensity of the contest, & closeness of the score line. These are also important factors.
Charles Plowdog
Guest
Lies, damn lies and statistics. I can't remember who said it, and it may be true. But more tries do not necessarily make more interesting games - and note I said 'necessarily'. But why I like Rugby Union more than Rugby league and AFL [apart from the fact that it is a more multi dimensional game (than RL) and allows for greater diversity of sizes and shapes, and has scrums and lineouts (AFL)] is that it is an international game played in almost 100 countries a small number of which are really good. Therefore to me the loss of games against SA and NZ sides is a weakness in the competition, not a strength. I can't see the situation where Rugby can compete with the tribalism and local support of league and AFL. But they cannot, in turn, compete with the internationalism of Rugby. If the Rebels beat the Brumbies, well that is not very good (particularly as I am a Brumbies supporter) but when the Rebels beat the Bulls in SA, well that will be magnificent for them: no matter how many tries are scored!
sixo_clock
Roar Guru
Tough read RGS, cheers, The figures below are the averaged differentials from the 2011 season. ld = local derby, o/s = overseas Oz ld Oz o/s NZ ld NZ o/s RSA ld RSA o/s 16.88 37.06 11.38 26.63 12.19 33.06 What it demonstrates is that the local derbies were closer affairs and therefore more exciting to watch. Your work also highlights that although SANZAR wants to create more tries (and more entertainment) with the bonus point system they may not have it right just yet, four tries is a big ask in a tight competition.
Westius
Guest
Very happy to see statistical significance here! -- Comment left via The Roar's iPhone app. Download The Roar's iPhone App in the App Store here.