Equitable distribution of team strengths within leagues

[John Foley]

Is there a way to measure whether the teams in an amateur league have an equitable distribution of strengths? Typically, some teams have several strong players whereas other teams struggle to put out even moderate players. The resultant matches are often unsatisfactory to both sides. This is relevant to the question about how many divisions to create within a league. Suppose there were 30 teams, then what is the optimum number of divisions? This could be measured in terms of giving players the chance to play the maximum number of same-strength opponents or alternatively to minimise the number of disparate-strength opponents. Five divisions with six teams in each would provide more equal opponents and smaller divisions would allow double headers during the season. Four divisions with 7 or 8 teams in each division would have more disparities, especially with the top division typically having the largest number of teams and the greatest degree of disparity is at the top.

This sort of issue must have been analysed in other sports e.g. football. It may have been distressing to watch Manchester City play West Bromwich Albion in the Premier League just concluded. Clearly, TV revenues make it financially worthwhile to pack the leagues, but amateur chess competitions do not have to obey the same financial imperatives. Any suggestions welcome.

[IM Jack Rudd]

Measuring it is pretty easy: you can use the ECF grading system to figure out expected scores for each side. Working out what to do as a league is trickier. (I'm assuming here that the league is one whose function is only to set the rules, and it does not get involved in picking teams and assigning players to clubs.)

[Roger de Coverly]

Five divisions with six teams in each would provide more equal opponents and smaller divisions would allow double headers during the season.

Personally I think it boring to have to play the same teams, the same people and at the same away venues all the time. With 30 teams go for three single round divisions of 10 teams each. That's particularly the right thing to do if seven or eight of the teams have at least one or two strong players, since it avoids having strong players stranded in a lower division.

[Andrew Zigmond]

You're never going to get a perfect balance of strength in divisions. In the Yorkshire League there are several examples each season of teams having to travel for a match they're going to get a point in at best. Most teams accept this as part of the league experience as long as they're not having to do it every match (and occasionally the scoreline is closer, that's part of the fun).

If the league has promotion and relegation then that goes some way to solving the problem.

[Mike Gunn]

The two leagues I play in (the Surrey League and the Surrey Border League) adopt different approaches. The Surrey League has (currently) 4 main divisions with promotion/ relegation and a nomination system. The advantage of this system is that if your team gets beaten up one year then you get relegated and will get easier games next season. The Surrey Border League runs grading limited divisions. This has the advantage that (apart from the highest graded players) most players can play in a couple (or even 3) adjacent divisions. I find this system gives you more games with people closer to your grade, but the downside is less variety (it's not uncommon to play an opponent one week in division 3 and the same opponent a week later in division 4). However, you do get more closer matches (but at the expense of interest in the promotion/ relegation issues).

The system I would favour is divisions based on an average grade criterion. This gives more flexibility in the selectiion of teams and should also give more competitive divisions and closely contested matches. However, I have failed to persuade anyone else of the merits of this idea, the most common objection being "it's too complicated". I understand one of the Kent leagues has a system based on average grades, but nobody has ever posted their experiences of it on here (as far as I can recall).

[IM Jack Rudd]

The Devon League uses average grades, and it works fine.

The main thing you have to be careful with, as a league, is getting the eligibility rules right. My usual rule would be "grades for team eligibility purposes are the ones players had at the entry deadline; grades for team board order purposes are the ones players had on the match date".

[Alex Holowczak]

The system I would favour is divisions based on an average grade criterion. This gives more flexibility in the selectiion of teams and should also give more competitive divisions and closely contested matches. However, I have failed to persuade anyone else of the merits of this idea, the most common objection being "it's too complicated". I understand one of the Kent leagues has a system based on average grades, but nobody has ever posted their experiences of it on here (as far as I can recall).

The Birmingham Summer League that I organise uses a total grade for the team of 4 players. That has worked well, and no one has ever complained that it is too complicated.

We had a Max 560 division match once where there were three 180 strength players and a player graded 5, so we introduced something to stop that happening the following season.

I once received a comment that "everyone knows juniors are undergraded", so clubs with juniors were choosing juniors to play them in the matches. I was asked what I thought about it, and I'm not sure "delighted" was the answer the complainant was after.

[Kevin Thurlow]

Mike Gunn said,

"The system I would favour is divisions based on an average grade criterion. This gives more flexibility in the selectiion of teams and should also give more competitive divisions and closely contested matches. However, I have failed to persuade anyone else of the merits of this idea, the most common objection being "it's too complicated". I understand one of the Kent leagues has a system based on average grades, but nobody has ever posted their experiences of it on here (as far as I can recall)."

That system favours the bigger clubs, like Guildford (Mike Gunn's club)! The only Kent person who mentioned their system to me said that there were lots of disputes, but a sample of one does not provide a definitive answer. It also relies on grades being right. I can appreciate the humour of Alex's response, (whether frivolity is the way to treat a complaint is another matter), but people don't like playing under-graded opposition.

The whole point of leagues is that some teams win and others lose. Some teams win, some lose, why change the rules to try to stop that happening? When Redhill first got into Surrey Division 1, we regularly lost 7-1 or worse to Guildford, Mitcham etc. Then we got better, now we've got worse. That's life. John Foley's club (Kingston) is struggling at the moment.

There also seems to be a strange Surrey attitude (even from some SCCA "Board" members) that if you are likely to lose an away match, you don't bother to play, which I find appalling.

[Mike Gunn]

If you use the grading limited division system then a small club with one team can put its team in the most appropriate division.

To (slightly) extend Kevin's principle: all systems favour big clubs ...

[Mike Gunn]

(I should have added to the first sentence: "if the average grade system is used".)

[Kevin Thurlow]

"To (slightly) extend Kevin's principle: all systems favour big clubs ..."

True of course - but average grading systems are frequently introduced with a claim that they help smaller clubs.

[John Foley]

John Foley's club (Kingston) is struggling at the moment.

Kingston is indeed struggling - struggling to keep up with the surge in our membership and the enthusiasm of our members. We recently reached the final of the Surrey team knockout in an exciting match v Surbiton which was covered in the Surrey Comet and the Kingston Guardian and BritBase. Indeed Kevin, you awarded the cup to the deservedly victorious Surbiton and made a speech. After narrowly losing the opening match to Ealing, our B team was undefeated in the remainder of the season achieving a well-deserved promotion in the Thames Valley League – and with no defaults. We are running the first ever Kingston RapidPlay on Saturday 28th July for those who cannot make the British Championship at Kingston upon Hull. Our summer chess ladder at the Druids Head is a popular social event attracting players from local boroughs. We play social chess in the pub every Monday apart from Christmas. We are increasing the number of teams entered into local competitions next year. Our AGM was the best attended in years with a dozen officers appointed. We make full use of the web, Twitter, Facebook, WhatsApp and use an internal club app customised by our software team. It's all been quite a struggle.

[IM Jack Rudd]

Average grade leagues certainly help the specific small club I play for, Barnstaple, inasmuch as we can just squeeze into the Devon League's second division with a "one IM and three 130s" team.

[Kevin Thurlow]

"Our AGM was the best attended in years with a dozen officers appointed."

Wow - what are they all doing!?

I meant the results were disappointing in the Surrey League. I'm glad there's lots of enthusiasm.

[John Foley]

"Our AGM was the best attended in years with a dozen officers appointed."
Wow - what are they all doing!?

Off topic, but here goes:

1. President: Ken Inwood
2. Chair: John Foley
3. Treasurer: Adam Nakar
4. Secretary: Alan Scrimgour
5. Fixtures Sec: Nick Grey
6. Liaison Officer: Stephen Moss
7. IT Officer: Ed Mospan
8. Social Media Officer: Genc Tasbasi
   Team Captains
9. Surrey 2: Julian Way
10. Surrey 3: Ed Mospan
11. TVL2: Jon Eckert
12. TVL3: Genc Tasbasi
13. TVX: Dan Oliver
14. Lauder Cup: Paul McGlade
15. TVL KO: Jon Eckert
16. Alexander Cup: John Foley