Реклама
Your Ad Here
Ако не искаш да виждаш тези реклами Стани член!

DancesportInfo Rating system

Why rating?

Wherever you go in the dancing community, one of the most discussed subjects is always 'Who is better than whom'. However, there has never been an adequate way to compare couples with each other. Existing ranking systems often do not agree with popular perception, and are based on very simple rules. This leaves many questions open to discussion. Who (in their time) was better, Paul Killick and Hanna Kartunnen or Bryan Watson and Carmen? Paul and Hanna were first in the ranking; however Bryan and Carmen were winning with them in the actual competitions. Similarly, is Cocchi better than Kongsdale? You can see on our forums the wide spectrum of opinions on the subject.

A very similar discussion applies to competitions, where people argue which competition it is more worth it to go to. Many times, we have heard from competitors that they are or are not going to a competition, because it is or is not 'good'. It is the general opinion in dancing circles that the British Open in Blackpool is *the* competition to go to. However, which is second? Is the Russian Open better than the German Open? Is the International better than the UK Open or not?

After some long deliberation, and many requests, we have decided to develop, and soon release, a rating system which could be applied to, and hence allow to compare, all dancing couples around the world.

What is advantage of unified rating system?

A world-wide rating system will allows us and possible other users, both now and in the future, to provide measurable statistics in a wide spectrum. For example:

  • Being able to rate events, one could decide that reaching the semi-final in one event maybe be better than winning another.
  • We will be able to create a graph representing the change of a couple's abilities over time. We will be able to provide a better comparison than the existing one, of two or more couples, even if they did not take part in exactly the same events.
  • We can also go back in time, limited only by how back our data goes, and find out any couple's rating for that day.
  • By providing an event rating, we can stimulate competition organisers to improve the quality of an event, and convince more dancers to take part in the events.
What is rating system?

Our system is not supposed to replace any national or international ranking system already existing, but rather to complement them. Our rating system is based on the ELO chess rating system, used as a standard worldwide, and suggested by some of the members on our forum. It has, however, been modified (patent pending) to cater for both large and small competitions.

Take note also that this is a rating system, and not a ranking system. The difference is that a rating system describes the current strength or quality of a couple, rather than it's position in the chart.

In our opinion, it is nonsensical to compare, let's say an Amateur Ballroom couple with a Professional Latin couple, hence there are seperate and different rating tables for each style and group.

The basic principle is that we assume that a couple's rating increases as it wins with other couples, and decreases when it loses. However, it increases more when it wins with a stronger couple, and less if it wins with a weaker couple. The basics of the system are described here.

The rating system for couples allows us to construct a rating system for events and competitions. However, there are several factors which have to be considered when trying to calculate the rating for an event. In our formulae, we try to take into account most major factors which affect a dancer's perception of the quality of an event or competition.

We rate any competition both as a whole, and by considering all the events of different styles and groups (e.g. Amateur Latin is separate from both Professional Latin and from Amateur Ballroom). When considering different events, we do not differentiate between Rising Star events from Open events, if they have the same group and style.

A more in-depth description of this system can be found here.

As much as we have tried to make this a reflection of quality of both dancers and competitions, we are aware that it is not possible to define all factors of dancing as a mathematical formula, and therefore this rating system might not represent some peoples' perception of quality. In the same vein, this rating system does not represent the subjective opinions of the Dancesport Info team.

How can dancing community help?

We are looking for people to cooperate with us to further improve our rating system. We need beta testers to look at the results of the rating, to point out possible mistakes, and to inform us of their subjective opinions, which may lead to a better formula for rating. At the same time, we are looking for mathematicians, who could the mathematical side of the rating system. Finally, our system depends on a large database of accurate results, and will improve as we get more data. Hence as always we are looking for contributions from the whole of the dancing society and cooperation from competition organisers and scrutineers.

If you wish to take part in the testing and improvement of this beta release, please e-mail us for further details.


Couple rating

A competition is considered as a sequence of individual 'duels' between each possible pairing of two couples. Let us consider the following example:

We have three couples - A, B and C - competing in an event. Couple A enter the competition with a rating of 2000, couple B enter with a rating of 1500, and Couple C with 1700. In the final, A was first, B was second and C third. Couple A will then gain in rating because it won with B and C. Couple B will gain points for winning with C, but lose points for losing with A, and Couple C will lose points for losing to couples A and B. However, B will gain more points from winning with C, than lose from losing to A, because B's original rating was lower than C's. Conversely, C will lose more points from losing to B than to A, for the same reasons.

The amount of points gained or lost in each 'duel' is calculated according to the ELO system.

The total gain or loss of points by a couple is based on the sum of all the results of the individual duels, and normalised by the number of people competing, to avoid too great an influence of a big event. Only couples whose original rating is in the same range are compared against each other, as there is no point in penalising a non-experienced couple for losing against the world champions. In the first event they part in, their first rating will be the average rating of the couples which ended up in the same round, and if this is not possible to calculate (e.g. all the couples are new), the following initial values are assigned to the couples:

  • Деца: 600
  • Юноши: 900
  • Младежи: 1200
  • Аматьори: 1500
  • Сеньори: 1200
  • Over-50: 900
  • Професионалисти: 1800

Event and Competition rating

An event is considered better if:

  1. Stronger couples (couples with a higher rating) compete in it
  2. There are more couples
  3. There are couples from more countries

To calculate the rating of events of a certain style and group (of which there may be only one in a competition), we first create a list of all competitors from these events. We sum their ratings, and normalise them by the number of competitors. Then, this is increased proportionally to the number of countries taking part. The final rating is then normalised to avoid too great a difference between event ratings, and multiplied by a constant to make it more readable.

To calculate the rating of a competition, we use a similar formula; however we take into consideration all the dancers who took part in this particular competition.

Please be aware that this ranking for events does not reflect does not include any subjective factors, such as traditions, the venue or even the quality of the floor.