Chez Nous: Top 14 Orange 2013-2014



Earlier this week, a friend shared a link to François Valentin’s Planet Rugby post, French rugby, where the locals rule.

François observes:

The Top 14 is a championship in which the locals rarely lose. So far this season, visiting sides have won only 12 percent of the games

I thought I would follow up on this assertion given my interests in home advantage and probabilistic approaches to winning.

Top14 Orange 2013-214

One hundred and forty-five games games have been played in the 2013-2014 Top 14 season. There have been 121 home wins, 19 away wins and 5 draws. This gives approximately 13% (13.10%) of away wins in this season’s competition.

At the end of Round 21, the Top 14 Orange points table is:


My record of these games is:

T14 HA

The legend for this table is:


The results have surprised me.

  • 77 of the 145 games played this season have followed my ranking model.
  • 12 of these 77 wins have followed my expectation of a ‘routine’ away win by a higher ranked team.
  • There have been 68 counter intuitive results (56 home wins and 7 away wins by lower ranked teams, 4 home draws and 1 away draw by lower ranked teams)

The Tournament leaders, Toulon, have: lost away games seven times to lower ranked opponents, drawn away against a lower ranked opponent once; and have won 10 of 11 games at home including a defeat of Clermont (a higher ranked team from the previous season). Toulon has lost one game at home to a lower ranked team (Grenoble) in Round 16 by one point.

This is Toulon’s profile for the season to date:


After 21 rounds of the 2013-2014 Tournament, the home and away winning profiles of the teams are:


The two teams that joined the competition for the 2013-2014 season, Brive and Oyonnax, have not won away from home but have won 9 games and 8 games at home respectively.

François notes in his Planet Rugby post:

Clermont haven’t lost since November 2009 at Marcel-Michelin and every extra match they win becomes an additional source of motivation. Losing a home game would by a massive blow to the morale of the team and the entire town.

The next section presents some comparative data from across the English Channel in the Aviva Premiership.

Aviva Premiership 2013-2014


Ninety-six games have been played in the 2013-2014 Aviva Premiership season. There have been 51 home wins, 43 away wins and two draws. This gives approximately 45% (44.79%) of away wins in this season’s competition.

At the end of Round 16 of the Aviva Premiership, the Table is:



My record of these games is:

Aviva games

  • 67 of the 96 games played this season have followed my ranking model.
  • 32 of these 67 wins have followed my expectation of a ‘routine’ away win by a higher ranked team.
  • There have been 27 counter intuitive results (15 home wins, 11 away wins by lower ranked teams, 1 away draw by lower ranked teams)

The Premiership leaders, Northampton, have: lost one away game to a lower ranked opponent, drawn away against a higher ranked opponent; and have won all 8 games at home including defeats of two higher ranked teams (from the previous season).

This is Northampton’s profile for the season to date:


After 16 rounds of the 2013-2014 Tournament, the home and away winning profiles of the teams are:

Aviva HA


I enjoyed François Valentin’s Planet Rugby post. It set me off on a research journey that has been really helpful in considering probabilistic models of success.

Many teams aspire to a home fortress in sport. Top 14 Orange illustrates how this occurs in a particular cultural context. The pattern of games in this competition raises some important questions about tacit knowledge and how seasons are constructed.

I am a very strong advocate for all teams being able to win at home in front of their own supporters. The Aviva Premiership appears to achieve this whilst still valuing away game success. The Top 14 experience appears quite different.

I think there are some fascinating research opportunities here in pursuit of a taxonomy of winning performance.

Photo Credits

Match Rugby Top 14: Racing Metro 92(RM92) vs. Perpignan (USAP) (Christophe Cussat-Blanc, CC BY-NC-ND 2.0)

Try Saving Tackle (Peter Dean, CC BY-NC-ND 2.0)

Home Advantage at the Olympic Games: 1988-2012

On Monday this week I wrote a post titled Overwhelmed.

Back in 2010 I wrote about Home Ground, Home Advantage.

A comment by Danielle Woodward on my Overwhelmed post sent me off a journey back to 1988.

I thought I would look at all the host cities from Seoul to the present day and consider the impact on the host nation’s performance in the Games immediately before, at the host venue and the following Olympics to look at patterns of performance.

I did not go back to 1984 because of issues about boycott.

I have used the excellent London 2012 Games web site as my source of truth for the medal results for Seoul, Barcelona, Atlanta, Sydney, Athens, Beijing and London.

The pattern for total medals won at each Olympic Games from Seoul (1988) to the present is:

The data for the graph are (host year in bold):

Total Medals USA China GB Australia Korea Spain Greece
1988 94 28 24 14 33 4 1
1992 108 54 20 27 29 22 2
1996 101 50 15 41 27 17 8
2000 97 58 28 58 28 11 13
2004 103 63 31 49 30 19 16
2008 110 100 47 46 31 18 4
2012 104 88 65 35 28 17 2

The pattern of Gold Medal success (and position on the Medal Table) from Seoul (1988) to the present day:

The data for the graph are (host year in bold):

Gold Medals USA China GB Australia Korea Spain Greece
1988 36 5 5 3 12 1 0
1992 37 16 5 7 12 13 2
1996 44 16 1 9 7 5 4
2000 37 28 11 16 8 3 4
2004 36 32 9 17 9 3 6
2008 36 51 19 14 13 5 0
2012 46 38 29 7 13 3 0

The position on the Medal Table (based on Gold Medals won) from Seoul (1988) to the present day:

The data for the graph are (host year in bold):

Medal Table USA China GB Australia Korea Spain Greece
1988 3 12 13 16 4 26 41
1992 2 4 13 10 7 6 21
1996 1 4 53 7 10 13 18
2000 1 3 10 4 12 27 18
2004 1 2 10 4 9 21 15
2008 2 1 4 6 7 15 84
2012 1 2 3 10 5 21 75

There are some fascinating patterns of performance in these tables. Between 1988 and 2012 all host nations saw improvements in their medal table status. Within sixteen years from Atlanta to London, Great Britain improved by 50 places. Korea has shown a very interesting trend. After a dip in overall ranking in 1996 and 2000 Korea has returned to the top 5 nations in 2012. At present their performance curve is concave whilst Australia’s is convex. The country least affected by a home Olympics is Greece.

At the time of hosting a home Olympic Games in the 1988-2012 time period, all nations recoded their best gold medal performance. The USA (2012), Australia (2004) and Korea (2008, 2012) have beaten their home gold medal haul since hosting a Games. (I acknowledge Andrew Read’s comment on this post. Andrew points to the changes in world sport brought about by the demise of East Germany and the transformation of the Soviet Union. In 1988 the Soviet Union led the medal table (132 medals, 55 gold) and East Germany were second (102 medals, 37 gold). The Unified Team led the medal tally in Barcelona in 1992 (112 medals, 45 gold) but the USA were second ahead of a unified German team.)

Photo Credit

Ki Bo Bae

Weather, Risk, Analytics

I have been looking at probabilistic approaches to success in sport performance.

Whilst researching some ideas around rule based behaviour I came across this advertisement.

My interest was piqued and I sought out WeatherBill.

David Friedberg is the company’s Chief Executive Officer. David was with Google, where he joined as one of the founding members of the Google’s Corporate Development team. He managed a number of strategic projects for Google, including identifying and leading several of Google’s largest acquisitions. He has served as a Business Product Manager for AdWords. He has a degree in Astrophysics from UC Berkeley.

Siraj Khaliq is the company’s Chief Technology Officer. Siraj worked at Google in multiple technical lead roles, from the company’s distributed computing infrastructure to the high-profile Google Book Search project and other offline content search initiatives. Siraj has an M.S. degree in Computer Science from Stanford University, and a B.A. (Hons.) in Computer Science from the University of Cambridge, England.

WeatherBill offers insurance policies that allow farmers to protect themselves from “losses caused by Mother Nature.”  There are “no claims to file, no adjustment needed—if bad weather happens, WeatherBill will send you a check automatically, within 10 days of the end of your policy period.”

I listened with interest to David Friedberg’s discussion of WetherBill’s use of local data to assess and manage risk. He was a guest on Radio National’s Bush Telegraph.

I followed up with a visit to a post about WeatherBill’s use of Google Analytics.

WeatherBill has released news about recent investment in the company.

I am very interested in WeatherBill’s model for assessing and managing risk. I think it has some important resonance with discussions about home advantage in sport and more general discussions about performance in relation to ranking. The key appears to be rich local information in the context of a global system.


Photo Credits

Farmland 1


Tasmania Landscape