4NCL and Chess Results dot com

[Joey Stewart]

OK, so to clear this up...

Swiss-Manager is perfectly capable of doing the pairings. In fact, we use it to do the team pairings in the Swiss in Division 4. It is also capable of doing it in the other divisions, except the 4NCL uses customised pairings for Divisions 1-3, in order to guarantee 1 white and 1 black in each of the first weekends. The customisation goes further. Teams in each division are divided into top and bottom half based on last year's standings, and for three of the first four weekends, teams will play one top half and one bottom half opponent. The final weekend will have two matches against teams in the same half, and one match against teams in the other half. This needs to be done manually. There's no reason why this couldn't be plugged into Swiss-Manager manually though.

As has been noticed, some of the 4NCL is put on Chess-Results, but this is done entirely manually, in order for the games to be broadcast more easily on Chess.com. Lichess is far more flexible in this sense - it doesn't need that.

The reason the league cannot be run on Chess-Results/Swiss-Manager is that the rules of the league are not programmed into it.

If $10 billion is coming my way, I am sure I can put it to good use.

Why does the manager need to know rules - Aren't simple names and numbers sufficient for it to do its job?

Also, while the subject of 4ncl and Swiss manager is raised, why does it continue to use ecf ratings despite the actual games being fide rated instead - does it get confused by players having two ratings and only pick the highest number (which is invariably the ecf one) or is it a UK thing to try and make our tournaments look more attractive to foreign players by making it appear our fide ratings are way higher then they actually are?

[Alex Holowczak]

OK, so to clear this up...

Swiss-Manager is perfectly capable of doing the pairings. In fact, we use it to do the team pairings in the Swiss in Division 4. It is also capable of doing it in the other divisions, except the 4NCL uses customised pairings for Divisions 1-3, in order to guarantee 1 white and 1 black in each of the first weekends. The customisation goes further. Teams in each division are divided into top and bottom half based on last year's standings, and for three of the first four weekends, teams will play one top half and one bottom half opponent. The final weekend will have two matches against teams in the same half, and one match against teams in the other half. This needs to be done manually. There's no reason why this couldn't be plugged into Swiss-Manager manually though.

As has been noticed, some of the 4NCL is put on Chess-Results, but this is done entirely manually, in order for the games to be broadcast more easily on Chess.com. Lichess is far more flexible in this sense - it doesn't need that.

The reason the league cannot be run on Chess-Results/Swiss-Manager is that the rules of the league are not programmed into it.

If $10 billion is coming my way, I am sure I can put it to good use.

Why does the manager need to know rules - Aren't simple names and numbers sufficient for it to do its job?

I don't know what this means.

Also, while the subject of 4ncl and Swiss manager is raised, why does it continue to use ecf ratings despite the actual games being fide rated instead - does it get confused by players having two ratings and only pick the highest number (which is invariably the ecf one) or is it a UK thing to try and make our tournaments look more attractive to foreign players by making it appear our fide ratings are way higher then they actually are?

I don't know what "it" is.

For a FIDE-rated tournament, you can use whatever means you like to generate your starting rank list. The objective is to get the best people at the top, and the worst people at the bottom. I once used Lichess Rapid Ratings to pair a tournament, because it was the most accurate rating available. (The first tournament back after covid, and all the previous stages had been played on Lichess, and they were Rapid games - the other kids' OTB ratings were 18 months out of date.)

[John Upham]

Why does the manager need to know rules - Aren't simple names and numbers sufficient for it to do its job?

By "rules" (and this surely is fairly obvious) we mean the pairing rules for a Team Swiss competition. The 4NCL league competition uses additional rules not part of the FIDE rule set. These additional rules have yet to be programmed into the code for Swiss Manager. Hence it cannot be used.

However, chess results could be employed to publish team lists, pairings and standings but getting them there would require a great deal more effort.

[Roger de Coverly]

By "rules" (and this surely is fairly obvious) we mean the pairing rules for a Team Swiss competition. The 4NCL league competition uses additional rules not part of the FIDE rule set. These additional rules have yet to be programmed into the code for Swiss Manager. Hence it cannot be used..

According to AlexH it isn't the Swiss pairing rules for division 4 that are the problem, but the long standing conventions on how the pairings should work for the all play all divisions. In both division 3s, I think the varying strength of teams usually thwarts the attempt to face one top half team and one bottom half team each weekend. In practice other than the season's results, it's difficult in those divisions to determine at the start of the season which are the nominally stronger teams.

The alternation of colours is a useful feature. In the earlier days of the 4NCL, it didn't happen to quite the same extent, so some captains would swop players within the then 80 point rule to alternate colours.

[Tim Spanton]

By "rules" (and this surely is fairly obvious) we mean the pairing rules for a Team Swiss competition. The 4NCL league competition uses additional rules not part of the FIDE rule set. These additional rules have yet to be programmed into the code for Swiss Manager. Hence it cannot be used..

According to AlexH it isn't the Swiss pairing rules for division 4 that are the problem, but the long standing conventions on how the pairings should work for the all play all divisions. In both division 3s, I think the varying strength of teams usually thwarts the attempt to face one top half team and one bottom half team each weekend. In practice other than the season's results, it's difficult in those divisions to determine at the start of the season which are the nominally stronger teams.

The alternation of colours is a useful feature. In the earlier days of the 4NCL, it didn't happen to quite the same extent, so some captains would swop players within the then 80 point rule to alternate colours.

Why not get captains to submit team lists, and then go by average ratings?

[LawrenceCooper]

Why not get captains to submit team lists, and then go by average ratings?

One concern about that was shown in an early 4NCL online season where a team submitted a questionable squad list of players resulting in them getting a much higher seeding than they otherwise would have done. Admittedly the exception shouldn't shape the rule and it's less likely to be an issue with established 4NCL captains so it could be, and maybe is, an option to be used.

[Nick Burrows]

Early season registration lists are a very unreliable guide to team strength. Although not perfect, I think the way it is done, based on last seasons finishing positions is a more reliable system. Perhaps simply asking captains to estimate rating average over the season would be an improvement. Any captain who gets in significantly wrong is publicly flaggelated before the start of round 11.

[Roger de Coverly]

Although not perfect, I think the way it is done, based on last seasons finishing positions is a more reliable system.

In the two division 3s, the teams are reshuffled every season. Also with promotions and relegations you don't have many teams that competed in the same division, so relative rankings are guess work really.

[Alan Walton]

What is the objective of all this?

I personally I have no issues with the current process

1) As a captain maintaining the registration list isn't onorous
2) Submitting team lists is very easy
3) Results are refreshed as quickly as possilbe on the day itself
4) League tables are normally available once all results have come through (though Div4 has been delayed, presumably due to the triangular match)

So I presume you want to use chess results for other purposes? But people seem to be worried the 4NCL don't use it, so surely the simple answer is someone volunteer to do the work if it is that important to you

[Nick Burrows]

Although not perfect, I think the way it is done, based on last seasons finishing positions is a more reliable system.

In the two division 3s, the teams are reshuffled every season. Also with promotions and relegations you don't have many teams that competed in the same division, so relative rankings are guess work really.

Fair. I was thinking more of the top 2 divisions. I'm pretty sure a captains estimate will almost always get very close to a teams true strength.

[Tim Spanton]

Early season registration lists are a very unreliable guide to team strength. Although not perfect, I think the way it is done, based on last seasons finishing positions is a more reliable system. Perhaps simply asking captains to estimate rating average over the season would be an improvement. Any captain who gets in significantly wrong is publicly flaggelated before the start of round 11.

I'm not suggesting using registration lists; rather making the pairings after team lists are submitted on the day.

[Alex Holowczak]

In both division 3s, I think the varying strength of teams usually thwarts the attempt to face one top half team and one bottom half team each weekend. In practice other than the season's results, it's difficult in those divisions to determine at the start of the season which are the nominally stronger teams.

Division 2 and 3 have been very competitive in the recent past whether in the top or bottom half, but when Division 1 had Guildford and Manx in their pomp, and the gap between them and the teams at the bottom made it rather more relevant. But even then, you may recall the season when we experimented with 2x6, playing a Scheveningen for the first 6 rounds, then a RR for 5 rounds, and after 3 weekends six teams would have 9+ points and six teams would have 3- points.

[Alan Walton]

I'm not suggesting using registration lists; rather making the pairings after team lists are submitted on the day.

So are you saying you just want to look at pairings on chess results instead? Therefore why would the 4NCL want to duplicate effort if it is not required, the 4NCL site is perfectly usable to get updated pairings, match and board results as they happen

[Tim Spanton]

I'm not suggesting using registration lists; rather making the pairings after team lists are submitted on the day.

So are you saying you just want to look at pairings on chess results instead? Therefore why would the 4NCL want to duplicate effort if it is not required, the 4NCL site is perfectly usable to get updated pairings, match and board results as they happen

I'm not saying anything about chess-results - I'm talking about evaluating the strength of teams when making round-one pairings

[Nick Burrows]

I'm not suggesting using registration lists; rather making the pairings after team lists are submitted on the day.

You are suggesting rather than pairings being known in advance, they are only released on the day? I've always considered the opportunity to prepare in advance one of the uniquely attractive features of 4ncl chess.