Hi Gerald

At the moment the intention, if we go the EBU NGS road, is to continue to fix each player's grade on January 1st.   Quarterly might be possible, but it would be a nightmare to expect players and organisers to cope with grades changing on a daily basis, even with online entry.  There may be more than three grades.   There may be no grades.

As far as tournaments go, I believe the intention is to leave it to the host club.   So the day's events could be Open / Intermediate / Junior as we currently know them, or it could be say 2-5, 6-8, 9-Q, KA if the numbers justify this stratification.   This begs the question, how do we advertise the events, and how do we take entries ?    I think the preferred solution is to have mostly All Grades tournaments with stratified prize grades.

Masterpoints will continue, so we may still have 5A, 8B, 3B etc tournaments.

But these are good questions that are being discussed.   What do you think ?

regards, Giles

Hi

Yes, the idea is that Rating Points will be discontinued, replaced by the EBU NGS.

Currently a player's Grade is a combination of Rating Points (recent form) and total accumulated Masterpoints (history).   I suspect that this will continue for some time, two types of measurement running in tandem.  Plus Rank titles such as Club Master ... Gand Master.

Yes, we don't actually need grades 2,3,4 ... or Open/Int/Junior, we could just have one number, a 'handicap' percentage.   And then leave it to clubs as to how they use that percentage to stratify tournaments (and prizes).   This is how chess works, and golf.

If we want to have nationally standardised grades with names, then we can certainly have Novice / Junior / Intermediate / Advanced / Open ... whatever.

As far as I know the EBU NGS is not used for grading tournament entries in England.

cheers, Giles

Hi

The EBU handicaps go from about 40% up to 70%.   But it is a normal bell-shaped distribution, so very few players at the extremes.    In NZ most of our members are (national) Junior or Intermediate, BUT most of the tournament players are Intermediate or Open.   So we need to look carefully at how we segregate the events.   Are we comfortable with a Junior tournament with only three tables, alongside an Intermediate tournament with a dozen tables, or do we combine them ?   What if we have a phantom in each event ?   Similarly if we have an Open A alongside an Open B.

In Canterbury and Top of the South regions, we have many 8B All Grades tournaments where we combine all the players, but award prizes in grades, and these grades are flexible and do not necessarily match the national grades.   Usually we do roughly equal thirds of the field based on combined Rating Points.   (Sometimes we do Open/Open, Open/Int, Int/Int, Int/Junior, Junior/Junior for prizes.)

So it can be done, and the players accept this scheme.

I would like the national grade names to match club grade names.   At the moment it's very confusing when "Junior" and "Intermediate" are used for club nights, but don't match the national Junior and Intermediate grades.  And some clubs use Senior for Open.   And internationally Junior and Senior refer to age groups.   But that's a bigger fish to fry.

cheers, Giles

Hi

Yup, all these points are well known and being discussed.  

regards, Giles

Hi Helen

I believe this was discussed at Board level some years ago, Stratified Masterpoints, and rejected.

Yes, if you want to move up from Junior to Intermediate, the best way is to play in Junior-only events and pick up 3 or 4 B points for a win.   You can have a better result, performance-wise, in say an 8B, best Junior pair, and get nothing for it.

- Giles

Another way would be to play at higher levels and improve your bridge then you will earn more points as a result after a year or 2 of play.

I only played 2 intermediate events as an intermediate player because otherwise I know I will eventually move into open territory and be thrown by precise defense that I will never have come across before.

That feature still applies. Most clubs dont know about it though.

Yes, still exists, NZB Manual page E46.

Sorry, section E of the NZB Manual has been updated, it is now defined on page E47.

Hi Gerald

No, it won't be ready for 2019.   I think they're still thinking about options and costs.

- Giles

May I weigh in with some momentous news? Over the Christmas break Bob Fearn has developed a live ratings system based on the EBU model for use with X-Clubs results gathered over the last year or so. If you have heard only vague rumours of X-Clubs - a short potted history:

• In use since 2012 - used to score NZ-Wide Pairs the last three years
• Clubs play the same deals simultaneously - there are two deals for each weekday - AM and PM
• The whole season's deal files, handrecords, travellers provided
• Usually about 150 pairs, some sessions double that
• Results scored across twice a day - 5:30ish and 11pm
• At present 37 clubs involved - 4 non-affiliated - 1 non-publishing - NZB has 114 so we're about 3 in every 10 NZB clubs at this time
• It's FREE!! And it simplifies life
• Also granted free use of Pianola for posting of results - maintaining players' history and stats in graphical form with personalised email once an email address has been registered (X-Clubs home page)
• I, myself, personally, have never understood why EVERY club shouldn't want a piece of the action. Even if just a few in each club are interested there is no impact whatsoever on others. It's completely transparent
• More detail on the X-Clubs site under the About tab

The advent of weekly-updated X-Clubs ratings (XGs) now makes handicapped results possible. This completely removes all quibbles of the sort "Why should I, a Gold GM, wish to compare my results with some beginners?". Now, with handicaps, it is possible for a beginner to compare their results with the GGMs. In fact, the weekly Leagues tables (Example - takes a while to load - big file) give a clear picture of the "cream having risen to the top" when sorted by XG. Check out the NZB ranks and ratings in the top 50 XGs.

It also seems to me that, in comparison with the EBU and any such system that NZB might adopt, an important distinction in that these XG ratings are based purely on C-point club bridge in which the inputs are the result of scoring across a large number of pairs. Analysis has shown that club results may be "out" by +/-10% when compared with the same contracts scored across a large field. You really need about 120 pairs before scores settle down to +/-0.5% in relation to the large-field results. On that basis I would claim that Bob's XG's are about as "pure" as a ratings system could possibly be.

Thanks Michael, and Bob !

It's nice to see this as an example of what we are trying to achieve.   It looks like Bob is using the EBU formulae - most recent 200 boards, decaying per session (?).  Of course it only includes NZ clubs that are using XClub, and doesn't include Teams results (?).   Does it include tournaments ?

regards, Giles

Hi Giles and thank you for your comments. In answer to your questions:

Older sessions start dropping off after 2000 boards - roughly 80 sessions. Older sessions also have less weighting than newer ones.

The database is comprised solely of C-point club matchpoint sessions that are played using the same boards. IMP scoring is not included. Although you can sometimes see tournament results posted on the X-Clubs website they do not get there by the X-Clubs process and are therefore not subject to the XG calculations.

Happy to answer any more questions.

Mike

The scores used are the results of scoring across the many clubs that play the same sessions - click here to see a table I made just yesterday - anywhere between 4 and 15 clubs in 9 different timeslots. The weekly calculation (fresh this morning - see results here (may take some time to load)) uses these scores.

The XG is like a typical club handicap.    “Average” players will have an XG of around 50%, and more experienced players will to have higher XGs, say 55% to 70%.  New XGs are recalculated after every session played and is based on the players existing XG, their partners XG, the average XG of their opponents, and each pair’s actual session score.  The recalculated XG is a weighted average of the session score for the last 2000 deals played, with recent results being given a higher weighting than older results.   

Example 1.   Assume …

                  your current XG = 55%

 your partner’s current XG = 61%

the opponents average XG = 52%

             In this session, the “expected score” for your partnership

=  50     plus       Average of you and your partners current XG

less       Average of the opponents XG

=           (55 + 61)  / 2  - 52  =    56%

If the pair scored 57% for the session, the partnership has scored 1% (57 less 56) above expectation thus the recalculated XG for each player should improve.

Example 2. Assume this pair is playing in a much stronger field where the oppositions average XG is 60%

             The “expected score” for your partnership is now 50  +  (55 + 61) /2  - 60  =   48%  

If the pair scored 57% for the session, the partnership has scored 9% (57 less 48) above expectation thus the recalculated XG for each player will improve.

Example 3.  Assume this pair is playing in a much weaker field where the opposition average XG is 47%

             The “expected score” for your partnership is now 50  +  (55 + 61) /2  - 47  =   61%

If the pair scored 57% for the session, the partnership has scored 4% (57 less 61) below expectation thus the recalculated XG for each player will be decrease.

Bottom line

• if you play above expectations (based on the strength of your partner and your opposition), your XG will increase. Conversely, if you play below expectations, your XG will decrease.
• If you play with a good player and play poorly, your XG will decrease
• If you play with a good player and play quite well, your XG will probably not change
• If you play with a good player and play very well, your XG could increase a little
• If you play with a “poor” player and play quite well, your XG will probably increase
• If you play with a “poor” player and play very well, your XG will increase

As the XG recalculations are dependent on many factors, it is difficult to predict how the XG changes from session to session.   However, a rough rule of thumb is that, for every 1% variation from the expected score, the XG will change by 0.025%.

From the NZB Board Meeting of January 22nd -

"Player Handicap & Grading Project
Handicap project discussed. We need to finalise fixes to the operating platform
before advancing this project further."