Welcome to the Science of Sport, where we bring you the second, third, and fourth level of analysis you will not find anywhere else.

Be it doping in sport, hot topics like Caster Semenya or Oscar Pistorius, or the dehydration myth, we try to translate the science behind sports and sports performance.

Consider a donation if you like what you see here!

Did you know?
We published The Runner's Body in May 2009. With an average 4.4/5 stars on Amazon.com, it has been receiving positive reviews from runners and non-runners alike.

Available for the Kindle and also in the traditional paper back. It will make a great gift for the runners you know, and helps support our work here on The Science of Sport.

Thursday, June 30, 2011

Home ground advantage: International vs local differences

Home ground advantage: The effect of international travel on performance

So today is the third and final installment in the series on home ground advantage, and it's going to be a visual and quicker to read one (promise!).  That will be followed tomorrow by a video recap of the whole topic of home ground advantage, which will also be the "world debut" of our Science of Sport Video series (but more on that later), and then we hit the 2011 Tour de France for three weeks of hopefully great, and interesting (and dope-free), cycling.

In the previous post, I looked at the size of home-ground advantage in Super Rugby, and my (admittedly) brief analysis showed that:
  • Home ground advantage exists - 61% of matches in the Super 14 were won by the home team
  • The odds of a team winning an away match were 0.64 times lower than the odds that they'd win the home match (eg: If the odds of winning at home are 80%, then the chances of winning away are 52%)
  • Home ground advantage is worth 9.5 points.  This is calculated by looking at all home records and comparing them to all away records.  On average, teams win by 4.8 points at home, but lose by 4.7 points away, and so home ground advantage is worth 9.5 points
  • The teams with the strongest relative home advantage are the Lions, Cheetahs and Brumbies.  The teams with the weakest home records are the Stormers and Chiefs
The travel factor: As yet just out of reach, but international vs local analysis sheds some light

Now, what we need to consider is how travel might impact on this.  I'll say upfront that this is a question I'm going to pursue in much more detail in future.  The data I have so far don't allow me to answer the specific question about travel...yet.

However, what we can do, using historical records, is look at the following two questions that get indirectly at the issue:
  1. "What is the probability of the away team winning a match in its own country, compared to winning a match outside its country?"  
  2. "What is the size of the home ground advantage when a team travels away WITHIN its country compared to playing an away match overseas?"

In other words, there are two types of away matches in Super Rugby - those where a team plays a LOCAL opponent (the Johannesburg-based Lions travel to Cape Town to play the Stormers), and those where a team plays outside its country (the Lions travel to Auckland to play the Blues, for example).


These questions are relevant because they allow us to start seeing what the effect of travel may be.  Two limitations to the above questions are important.  First, this method doesn't allow us to see the acute effects of time-zone changes, because teams play those away matches for four to five consecutive weeks, so you have to include "time on the road" as well as travel.

Also, there are scenarios where being the home team still involves travel.  Imagine, for example, that the Lions have just played four away matches in Australia, and then fly back to South Africa to play the Bulls.  In that case, the Lions are the traveling team, even though they may be at home against local opposition.

This is a level beyond the analysis I've done so far.  However, as I said, I'll definitely look at it in future.  The results I present below thus only answer the question of "International vs Local" matches away from home - it's an indirect measure of the travel effect.  It's also a relatively small data set - with time, I'll build it up and go back many more years to strengthen it.  However, I think it reveals some interesting truths, but it's by no means final!

So let's look at those "truths"...

The odds: How likely is a team to win outside its own country?

So, the main implication of the above figure:
  • Home-ground advantage is considerably lower when you play against a team from inside your own country - the home team then only wins about 54% of matches (remember, the tournament average is 61%).  The odds of winning away are 0.85 times those of winning at home - not bad at all.
  • When a team goes overseas, the home team wins 64% of matches. Now, the odds of winning away are cut in half compared to winning at home.  
  • There, winning away matches outside your own country is far LESS likely than winning them in another country - home ground advantage counts for more when the visitors must travel internationally
  • I repeat the stat that says that since 2000, no team has won a knock-out match having had to travel to another country.  That may change this weekend, but it's 0 from 34 now, and that says a lot
To repeat, this doesn't mean it's exclusively due to travel (it may be "homesickness" if a team spends a full 5 weeks on the road).  And there are matches in the above data set where the traveling team is actually the local team, as I mentioned above.  But overall, it says that playing overseas has a significant effect on the likelihood of a team winning.

The size of home-ground advantage against local and international opposition

Next, we take the same approach I did yesterday, looking at every single team's home and away record to see how large their relative advantage is when playing at home.

First, the figure below shows points differences and average scores for all teams when playing against LOCAL opposition (from the same country) either home (top) or away (bottom panel):

So, you can see that most teams have winning records at home (green bars) and losing records away (red).  The black rings show examples of teams with winning records both home and away, while red rings show losing records home and away.  Some teams, like the Reds (see arrows) have a winning record at home, but lose away.  The difference between the home points difference and the away points difference is the home-ground advantage, but I'll get to that shortly.

Next, we look at the same 14 teams, this time home and away against international opposition:

Once again, some teams have winning records both home and away (like the Crusaders, ringed in black), while others lose both.  Interestingly, only two teams have winning records outside their own country.  It's quite clear from these two graphs that the average result being overseas is worse than being away within your own country.  And the red rings this time indicate the more common pattern - a winning record at home, but a losing one away.  For example, the Brumbies average a 27 - 20 win at home, but a 17 - 24 loss away.  Again, the difference between these two will tell us the value of being at home.

That value, the "home-ground advantage" is summarized in the graph below.

So, the top panel shows the relative home-ground advantage when playing against LOCAL teams, the bottom panel the advantage when playing international teams.  Bizarrely, the Bulls actually perform better away from home against local teams - if you go back up to the first figure, you see that their home record against other SA sides is a 23-21 win, whereas their away record is a 29 - 19 win.  I'm pretty sure there's nothing in this - it's the reason why the dataset needs to be expanded.  But they are the exception - all other teams far better at home against local sides.

Similarly, against international sides, everyone but the Chiefs do better at home against international teams.  You'll recall from yesterday that the Chiefs actually have the lowest home-ground advantage - this is why - they average a 3 point win at home, a 2 point win away (the error is due to rounding up/down).

However, the real significant fact is this:
  • Average home-ground advantage when playing against local teams is + 6.3 points
  • Average home-ground advantage when playing against international teams is +10.8 points
  • Therefore, home-ground advantage is increased when playing international teams.  To the tune of 4.5 points, which is the value I would attribute to "geography", in the sense that this is what it shows.  This is summarized in the graph below.

Once again, I must stress that this is doesn't accurately quantify the effect of travel, but rather of playing internationally or locally, and it uses a small sample size. To get the full value, one needs to a) go back further, all the way to 1996, and b) track results as a function of local or international travel, time away from home, direction of travel and home-ground advantage.

But it does start to indicate that the travel is a burden on teams in Super Rugby.  That being away from home, the probability of winning matches is much lower when you are overseas than in your own country - perhaps this is travel-fatigue, perhaps it is culture-related, perhaps it is related to home-sickness and motivation.  Whatever the reason, the early results suggest an affect.  Once again, I'd stress that this competition is unique in this regard.  Factor in the altitude (for some teams) and Super Rugby may have a lot to teach us about home-ground advantage!

All in good time, of course! That's a lesson to resume again in the future!

The Tour de France:  Three weeks of cycling coming up

And it's a wrap for home-ground advantage.  I realize that maybe the rugby focus was not relevant for many, but as long as it inspires some thought, thank you for reading!

The Tour de France is up next, and it's going to be all systems go!  The race starts this weekend, and you can expect analysis and thoughts as the race unfolds.  If there are any specific questions, or information, or power output values, or anything else, please don't hesitate.

The Video Series, launch imminent

And finally, an exciting (we hope) announcement.  Our main objectives with The Science of Sport are to share our insights and opinions (we mouth off on sport and science!) and to communicate scientific concepts, usually as applied to sport. The key is communication.

And so, we are always on the lookout for ways to improve how we communicate science.  That's why we created the Facebook page and our Twitter feed, so that we can provide links and thoughts more consistently than time allows us to post here.

But, one avenue we've never explored is video.  The thinking is that video lends itself to communicating science more effectively.  It's graphic, and we can talk through concepts rather than write them.  Maybe it's also more personal. 

And since we do a great deal of speaking and presentations (Ross in particular, being in an academic environment), we thought it a good idea to start publishing videos of our presentations and the topics we cover on this site!  That we, we can talk through ideas, and maybe condense your time a little.  We'll still keep going on the writing, don't worry, but the video hopefully brings an added dimension to the Science of Sport.  Please share and distribute!

The videos will go up on YouTube, they will be embedded here. We start with a video of a presentation on Home-ground advantage (which I will give internally at the Sports Science Institute tomorrow).  I know it's not a topic that many of you will relate to, being about rugby, but it's a start and in the future, I think we'll try to do this much more often, for topics ranging from the Pacing Strategy talk to Talent and 10,000 hours.

That will come shortly, let us know any feedback and how we might improve on our mission!