Monday 30 March 2015

Tight Contest for the Sherlock Holmes of Mahjong


ONLINE March 2015 - This year's Sherlock Holmes of Mahjong contest brought the largest field of competitors yet, and the crowd at the top is still tight; the title could belong to anyone. With four competitors tied for 1st, and five more just one point behind, the last couple of questions will be the make-or-break opportunity for many. The contest is likely to come down to the timing; how fast each player is able to get their answers in. So set your alarm clocks, and get sluthing!

1st place: 1-year VIP membership at MahjongTime and a copy of "Mahjong from A to Zhu"
2nd place: 1-year Premium membership at MahjongTime and a copy of "Mahjong from A to Zhu"
3rd place: 6-month Premium membership at MahjongTime and a copy of "Mahjong from A to Zhu"



Luca! Luca! Luca!

I have three reasons to hate Luca Gavelli.

First of all, he is much more handsome than me. His looks make a girl’s heart meld, and there must be many Italian mahjong ladies who have his picture in their purses.

He is also such a nice guy. We always have a short chat when we meet, and he answers my broken Italian in a much better English.

And last but not least, he is number one in the EMA MCR Ranking, while you have to go down to number 164 or something if you want to find my name (but why would you? See reason number 1.)

The biggest problem with Luca is that Luca has been number one for such a long time.

In the very beginning of the EMA MCR Ranking, back in 2005, I think there must have been another leader of the ranking. I have kept no records of that; might have been Jan Thomassetti from Belgium, or Anders Labich from Denmark. But after that? Each and every new ranking which is presented on the EMA and the Mahjong News websites: Number 1: Luca Gavelli.

No doubt about it - Luca is a terrific player. He ended on sixth position in the first OEMC in Holland, with only four Japanese and one Dane before him. In the second OEMC, in Denmark, he was number 12. In the third OEMC, Baden 2009: third position. Bravo, Luca!

But it is just like Lance Armstrong controlling the Tour de France: it is all so predictable. If, like Luca, you do not play too many tournaments, (and you are a strong player of course), than you can control the ranking for many years.

Another ranking

Look at it the other way. Luca has the highest average of all the 441 players who are listed: 931,07. That’s a great result (1000 is the highest, 0 the lowest possible figure). But behind his name, there are just 6 tournaments. That’s not many, although it is even more than French Pascal Balorin, who holds the runner-up position with an average of 849,65 and 5 tournaments. Anders Labich on position 3 played just 8 tournaments; his average is 822,5.

But then, on number 4: Eveline Broers from Holland. Her average: 817,98, which is fantastic since she played in 17 tournaments. And Otto Myslivec from Austria, who holds the 8th position: he played in 21 tournaments (in three of which he got the first prize), with an average of 777,38.

Now let’s take another ranking. Each year, Alexander Doppelhofer from Austria calculates who was the best player of the past year. First in MCR and, recently, also in Riichi mahjong. Alexander uses only the results of the tournaments in a given year. According to 'Doppelhofers List 2009', Bo Lang from Switzerland was the best player (no wonder, she was the runner-up in the OEMC at Baden).  In the EMA Ranking, she holds just 7th position (still a terrific average: 805,12; which is even better when you know she played in 17 competitions, almost three times as many as Luca did.) By the way, in Dopperhofers List, Luca still has an outstanding number 3  position. But the number 2 is quite surprising: Angela Nyulasiné Lakos from Hungary - the current number 28 in the EMA Ranking, with an average of 692,34 (and 10 tournaments).

Less is better

So, I jump to a conclusion: the less tournaments you play, the better your position in the EMA Ranking. Luca is an outstanding player, but we do not see him often. That is a smart decision, so he can be the leader for a long, long time.

That’s not his fault. But EMA should think out a new system, which is a bit more encouraging for players to subscribe to more mahjong tournaments. After all, that’s one of the most important goals of the association - having the game played.

Oh yes, and I do not hate Luca Gavelli at all. Of course not. I just wanted you to read this column until the end. Apparently, I have succeeded. Thank you for that - and Luca: scusami Sei un vero campionissimo!

Comments (13)Comments are closed
1Monday, 05 July 2010 21:33
Otto Myslivec
Have a look at place 6 - Elisabeth Frischenschlager - she plaid 6 tournamnets and her best result was only one rank 3.
I have 21 tournamnets in the ranking and 3 victories, 2 second and also one 3 place - but I am still behind her!!!!
The system prefers players with only few tournaments. Even I have one more victory I cannot skip one rank.
We discussed the existing system in our club but it's difficult to create a better new one.
2Tuesday, 06 July 2010 06:04
Dear Martin,
I find that LUCA's results are simply fantastic!!
But how does he for ever make finish in the first ones??

Nevertheless, it is normal that you can block the classification EMA with so few tournaments? I do not think.

Why not to take into account the spectacular increase of the number of tournament MCR? Indeed, there was 4 tournaments in 2005 and 16 in 2009. Why not make fall the weight of the 20 % tournaments a year? It would give a 80 % weight for 2009, 60 % for 2008, 40 % for 2007 and 20 % for 2006.

Why not to take into account also the number of players by type of tournament (MERS 1 or 2)?

In conclusion, I also think that it is necessary to make this ranking more dynamic.

I apologize beforehand for grammatical mistakes in English.
3Tuesday, 06 July 2010 08:46
Robert Rijnders
A new ranking system is currently being evaluated (our ranking master is
making 'shadow rankings' to see the impact of the different scenarios). At
the last GA it was decided that we should come up with a different system,
but this is a slow / difficult proces. It's not so hard to come up with
some other system, but any system has (dis)advantages or is not approved /
supported by all players.

Sébastien Berret can explain the favorite new '70/30' system way better
than me, so perhaps he's willing to enlighten us :-) In that system, Bo
Lang would be first, Otto Myslives second, Eveline Broers third and Luca
Gavelli would be in sixth place...

We should be able to take a decission shortly.
4Tuesday, 06 July 2010 12:11
Sebastien Berret
As Robert said it, we work on it since the last GA (July, 2009). Without going into details (it will come when it will be approved ;-), The new ranking could be based on a revision of the life of the tournaments and on a fact of being balanced with the best results, By taking into account the average of best results with the global average (maybe 70 % for the best results against 30 % of global average). And thus, active players having made good results over the last years meet necessarily in better position in the top.
5Tuesday, 06 July 2010 14:41
Otto Myslivec
I think it's time to say a big THANK YOU to Sebastian for his excellent work he has done and he still does for us.
EMA is a very young association and it takes time to find the best (for all players) solution. I know this from chess.
Like in other sports a victory should have a special weight: Marianne has three victories, two of it in Paris, one of the top tournaments in Europa, but she is only ranked on place 34 because of 34 tournaments. Anton with 46 tournaments only one place before.
I am confident that we will get a suitable system in future. ;-)
6Tuesday, 06 July 2010 14:50
Adrie van Geffen
In the process of thinking it may help to start with a distinction between rating and ranking. Rating is an ongoing score of results in some years (like EMA now), ranking should have a start and finish in (probably best) 1 year with a cup to mark it.
7Wednesday, 07 July 2010 01:40
Chris Rowe
Would it not be possible to make some sort of accumulative +/- points system? For example, instead of points given from +1000 to 0 and then take an average, what about +500 to -500 and then take a total?

A purely accumulative system would be bad because it would then heavily bias people who play in lots of tournaments, even if they finish poorly. However, if it's possible to lose points in a tournament as well as gain points, then it would remove that problem. It would mean that players who consistently finish high in many tournaments would occupy the top spots.
8Thursday, 08 July 2010 02:07
I think the biggest problem is that no one wants to program a decay into the ranking, either out of a lack of effort (probably not) or out of a desire to not offend people for not being as present as others. "cap974" had a good idea, of course, building on it within the limits of what's available will be tricky.

If compared to say go ranks on KGS (a movable rating system), for a rank to stabilize, it can easily take over 10 games (say 3-4 events' worth) to no longer be that questionable. But those rangs get recalculated and are partially dependent on future successes of each player's opponents.

"Chris Rowe"'s idea of a negative system could drive away a lot of people because it's *negative*. But there has to be a way to decay a score from a player who stopped 2 years ago to hit zero. If they're that good, they can climb back up again.
9Thursday, 08 July 2010 08:55
Alexander Doppelhofer
It is very difficult to design a new ranking system, but the more difficult it becomes more complicated it is to explain to the players.

There is currently very easy as a newcomer to come in the EMA ranking by playing two tournaments, but if you do not have long played more tournaments, and only have good results, then you are always top (eg: Nusreta Mauthner, at the end July 2010 3rd in the league table will take EMA).

Four years for a ranked list is too long. focus for one year a ranking would be optimal, but not due to the large distance of the tournaments suitable. Since the EMA, said that one can align three tournaments a year from a nation that would restrict it to improve to 2 years. Maybe you could even go a step further and say that if you want to be a top player this tournament must play two of the three + 1 International tournament for its quality to show even better.

I created a separate ranking 2 years ago and there were already so far to surprising results. Martin Rep has contacted me and asked if I can design a new ranking. I sent him mine and he has published them here. I look forward to your comments.
10Thursday, 08 July 2010 15:55
It's a complex and sensitive subject.

During the GA of EMA we (Morten Andersen –Denmark-, Janco Onnink –Netherlands-, Martin Mauthner –Austria- an I –France-) have had the charge of thinking about and propose to EMA a new ranking.

We have tried to make a new ranking that improve 3 points that seems important to EMA :
(1) this system should be representative of the actual strenght of each player
(2) this system should encourage players to play tournaments
(3) this system should be more dynamic than the actual one

The more tournament you play, the more your average is representative, but the less you can make it move... so it seems that it's difficult to have something dynamic AND representative.
And what about players who can't participate to a lot of tournament a year ? Are they automaticly poor players ? No, and Lucas is a bright exemple.

And it's not because you play a lot that you are a good player, so accumulative system are not a good way to rank the player.

We don't have changed the all way of thinking the EMA ranking, transition to a brand new system have to be made smoothly.

For the "dynamicness" (3) of the ranking, we have decided to reduce the fade of the weight of the tournaments to Y years.

For the "representativeness" of the ranking we have choosen to mix the global average (based on all the results) with the average of the X best results in the last Y years.

We hope that by this way, you can try, participating in a lot of tournaments ("encouragingness" (2)), to improve one of your X best results : the more tournaments you play, the more chance you have to improve one of your X best results.

Even if you don’t improve one of your X best results, as the tournament count in the global average, you don’t have play the tournament for « nothing ».

With this system, that we have asked to Sébastien to prototype, we see the recently rewarded players climb up on the top of the ranking.
The best player will be the one who will have the X best results and the biggest global average.

Lucas will still have his chance to be number one, even if he doesn’t play a lot of tournament, and Otto will be rewarded for his exellent performances.

11Wednesday, 14 July 2010 08:14
Bo Lang
I just want to thank all the people who are currently working on their own time to formulate & improve the ranking list. I'm sure it is appreciated not only by me but by players around the world. I am confident that with your combined efforts a significant result will be achieved sooner or later. Though I must admit, I am quite content with the newly revised ranking list for obvious reason. :) So you guys keep up the good work & I'll just keep on playing...
12Wednesday, 14 July 2010 15:41
foschi Marco
dear friends,
thank everything you so much for the time that you devote to all this...
My thought on the new calculation of the ranking is the following:
I find correct to devalue the old results but they should not come to be worth 0, a least contribution they should bring it at least for 5 years.
I think that if the primary objective is to bring more people to play, the players with more competitions should enter to of the bonuses as: possession the possibility to discard or to valorize less a bad competition, obviously after having done a least number of it established and with a devaluation that immediately has handed her even since 100% to 66%, in this way also the players that look at the ranking could be stimulated to play more.
I associate me to the idea of cap974, after 1 year 80% - 2 years 60% - 3 years 40% - 4 years 20% and so by thin to reset him a 5 real year.

13Friday, 16 July 2010 10:03
Alexander Doppelhofer
Hello Marco!

The problem with rankings five years is that the players are superior to tournaments to play, because if I only need five games, then I need only once a year to play a game. Subsequently, then, the players remain in the other tournaments will be unattractive at home or abroad, and the game and it is then only a matter of time until it is resolved.

In every other sport is played in a frame of a year that would not go for financial reasons. A frame of three years, I think not so bad, I would reduce it even to 2 years. The score lives from that student players and their tournaments, the players who have 2-3 years of no tournaments played more are discarded.

As far as I know it has been in Italy this year in addition to the "normal" tournaments and qualifying tournaments for the World Cup. For "we have" in Austria, we use the "normal" tournaments to qualify to us if the Italy makes, then the good players be there and then it becomes even more interesting in the Rangliste.Vielleicht then takes over again, an Italian the ranking.

I've just been away from that with a good result in the next tournament, "Luca" is again front and pass some of it is in the leading positions.
yvComment v.2.01.1

Mahjong News | Copyright © 1997-2015 | About Us | Sitemap | Contact