TOURNAMENT SEEDINGS “How’s it work ?”

Comments From Brian Davey – MARCH 2003.

It was approximately a year ago, Duncan Constable and I had been progressively developing a player-ranking database for use when seeding tournaments. I decided to try an experiment and distributed a list of our top twenty ranked players in ranking order from the database and gave it to eight players who were either in the top twenty or had played against the players in the top twenty. I invited the eight players to return the list to me with their own ranking order, the outcomes were: -

1. None of the eight players agreed with our list from the database and,
2. None of the eight players agreed with each other and,
3. Six of the eight players ranked themselves higher than what the original list had them.

My takeaway was that you will never please or get agreement from everybody about this and secondly, we needed to ensure that our database was influenced by some logic and not peoples opinions.

Since then, Duncan and I have put a considerable effort into improving the database and make it a workable tool that is effective in ranking our tournament players, we believe that we have progressively been achieving that. The purpose of this article is to offer you an explanation as to how some of the rankings and matches happen at the respective tournaments, and also offer you a partial insight into the variables we have to consider and some we don’t consider.

During the article I will often refer to “we”, that being Duncan and myself, it is worth pointing out that Duncan voluntarily contributes many hours into this, he is continually maintaining and making enhancement improvements to our database that also executes the tournament draw.

Effectively, we have a database that has 386 active tournament players on it. Each player is ranked from number 1 onwards, and this is adjusted after each tournament after consideration of the results.

It is often commented that we should include pennant results in our rankings; we don’t for the following reasons,

• Pennant grades are ranked by the team members as a collective, not as individuals, it is not uncommon to find stronger individual players in lower pennant grades, and equally not uncommon to find less stronger players in higher pennant grades. The upshot of this is that the individuals rarely, if ever play each other in pennant, yet will cross paths often in tournaments.
• Of the 816 players who play VMSA pennant, only 210 of them are regular tournament players.
• The standard difference between competitions is a constant variable we can’t monitor, e.g – Grade 2 in Bendigo midweek is not as strong as cup grade 2 in VSF. Also, some pennant competitions have handicap scoring. Our current crop of tournament players compete in, VMSA, State grade, VSF, SSL, North West, and several regional pennant competitions, it would be huge task to monitor the results of all of these pennant matches each week.

Some of the pennant competitions have allowed us to have team entry copies, these are used as a reference, however, pennant results have little if any impact on our ranking list.

It is often commented that, I was in section 3 at Sale, which I won a trophy in and now I have gone down to section 4 at Ballarat

• Unfortunately, it is common for most people to interpret a section as a grade, IT IS NOT ! Each section is a group of sixteen players. Section one is the top ranked sixteen players that have entered that appropriate tournament, section two the next group of sixteen and so on.
• The more players that enter a tournament, the higher the section number you will be grouped in, (unless you are ranked in the top sixteen players). I will use myself as an example, my ranking number in 99, if every player on our ranking list entered the same tournament, then I would be in section 7, (99 divided by 16). Generally speaking, I play in either section 2,3 or 4, which is determined by the number of entrants.
• Using the recent Sale tournament as an example, I was in section 3, My grading number is 99, and I was seeded at number 1. The number 16 seed in section 3, had a ranking number of 129, again the number of entrants determined this. At the Ballarat tournament, due to the large number of entries, I would probably be a lower seed in section 4. Unless I improve my ranking from 99 between now and then.

Gee, Brian is seeded at number 5, above Ray at seed 9, yet in their pennant team, Ray is higher, how does that work..

• The major contributing factor to these anomalies is contributed to by your availability to play on a Friday night. It is critical that the tournament completes at least 55 matches on a Friday evening, due to the fact that not all entrants are able to play on a Friday evening, we adjust some of the seedings to accommodate players who can play each other on a Friday, and players who can’t, play each other first thing Saturday.
• We still have to factor in times, some players can arrive at 7.30 pm Friday, some at 10.00 pm, hence the alterations. We only make these adjustments within a section.
• The other frustration is some players pull out of the tournament at the eleventh hour; this is almost impossible to cater for. The regular offenders of this practice have been noted.

We do consider players returning or recovering from injury, or who have not played for a while, however, if you are in this category, it is critical you advise us. The most effective way to do this is to note it on your entry form. We also watch as many matches as possible, so as to familiarise ourselves with who is who and who is hitting the ball well, or not so well. Unfortunately, we don’t consider the toll you may place on your body during the course of a tournament, self-inflicted performance dehancenet activities are your choice. Don’t try to kid us on this one, Duncan and I are the experts in this area !!!!

A measure that we have applied to ourselves is to monitor the number of second round matches that have a 3/0, 3/1 or 3/2 results, regardless of whether they are a main event or plate event. We are noticing a decrease in the number of 3/0, second round results, which indicates that the matches are more even; therefore the rankings and seedings are having the desired effect.

The RANKING – the whys, where-fores, what ifs, butts and maybes.

Contrary to some players opinion, the tournament ranking process is not governed by the laws of witchcraft or astrology. Detailed below is a commentary to explain how the performances of a player is accessed at a tournament. Each player that enters a tournament has a ranking number, and that is adjusted after each tournament they play in. If a player does not play a tournament for two years, his name and details are archived, should they ever play another tournament, their ranking will be updated after considering current details. We make a similar consideration for players who may be returning from serious injury. In any individual tournament, players are sorted into sections of 16 players (assuming there is sufficient players in quantity and standard to fill that section, otherwise a smaller group is formed.) The first round matches are played from a higher seed v lower seed key:- Seed 1 v 16, seed 2 v 15, seed 3 v 14, seed 4 v 13, seed 5 v 12, seed 6 v 11, seed 7 v 10, seed 8 v 9.

Each player is rewarded for their wins by allocating points in accordance with the table below:-

At the end of the tournament, a players ranking is adjusted (higher or lower) and the following criteria is factored into the adjustment.

• The final placing (1st to 16th) as determined by the points per win allocation.
• The strength of the section, which is determined by the gap between the ranking number of the highest and lowest ranked player, i.e.:- Player seed 1, is ranked at 60, player seed 16 is ranked at 110, the gap is 50 ranking places (110 – 60 = 50). This gap number is divided by 16, and becomes our weighting factor for that section, in this example the weighting will be 50 divided by 16, which is a weighting of 3.1.. This figure is applied to adjust your ranking up or down by 3.1 places subject to where you finished in the section.
• Points won from a walkover are not considered when we make a ranking adjustment.
• Some basic common sense……. Yes Duncan and Brian do have his capability.

By considering the criteria above, we seek to “smooth” the changes in rankings and hopefully allow more competitive and enjoyable matches.

In closing, can I invite your comments, it is important that we know how we are doing, and what you are expecting. Remember, a tournament is about good squash, good fun and good friends.

Yours in squash…………………….Brian. (Mobile 0417394314 / Home 95463078 ).