MOSCOW, Russia - Mahjong analyst Vitaly Novikov shares the idea of an overall MCR players ranking. An innovative ranking system and detailed database of over 1,000 Chinese players now allow to "meet together" the results of Western and Eastern tournaments at one place.

Next generation 'level' ranking system 

In late 2017 it was decided to review the ranking system (RS) in Russia. The weakest point of previous versions of the RS (incl. old EMA RS) was "assigned" tournament coefficients (like MERS 2 even for 16 players tournament) which may not reflect real tournament complexity. The most teasing was the class of RS where average level of participants calculated formally may be a constituent part of tournament coefficient. In order to run such kind of RS entirely new technical level of data management must be achieved since under new system any player's result in a single tournament will influence further calculations.

Nevertheless, all technicalities have been surpassed. All almost 300 tournaments in EMA database have been processed. As the result, all 1000+ players were given some "level" (see below) and ranking. Stepping forward the new problem arises: can we add also known Asian tournaments, some of which have more that 200 players? Yes, almost 10 major known (to be correct, results of which are available in table format) tournament have been added to analysis (see table below). We speak about so called Paiwangsai (牌王赛, "Kings") and Dashisai (大师赛, "Masters") tournaments and several more tournaments organised by Mahjong International League. Not speaking of all MCR world championships. The major problem which have been solved was the avalability of results for some tournaments in photo, not in a tables. Despite that fact ALL Chinese names have been deciphered and placed to database. We have 1,000+ Chinese players names written both in characters and pinyin.

'Level' ranking

So, what is "level" and how does this "level" ranking system work? A player's strengths is assessed formally in terms of steps of a ladder, altogether there are eight "kyu" and twelve "dan" levels. Kyu levels start from 0 and go till 1750 with 250 points between kyus. Dan levels start from 2000 and go up with the distance of 500 between levels. So, D1 (as a contraction of "Dan 1") corresponds to 2,000, D5 -- 4,000, D9 -- 6,000 etc. In any tournament a player receives personal result with is equal to "basic rank" * "tournament coefficient" (BR*TC). For instance, 1st place in tournament with 76 players and TC=2.70 will bring 1000*2.7=2,700 points while 26th place will bring (76-26)/(76-1)*1000*2.7=1,800 and 51st place -- only 900. 2,700 points means that this player fulfills the norm of D2 (2500 points), 1,800 brings fulfilment of q1 (1st kyu or 1,750 points), 900 -- q5 (5th kyu or 750 points).

All levels from 250 to 2,000 are "permanent", whenever a player surpasses by personal result any norm in that range that kyu or D1 will "stick" forever. All dan level are "double-layered". Firstly, in order to go up any norm fulfilment brings only one "achievement" (coded as "+"). For instance, player with D1 (2,000) gets personal result of 3,700, then what is new level? Since player fulfills the norm 3,500 then it means 1,500 points of "surplus" from the current level, ultimate level is D2+++ or D3+ (here each "++" means 1 level up). Secondly, any level of D2 and higher is given for 365 days. Whenever a player cannot "confirm" his / her dan level within that period a player receives "-" for each year. Accordingly, "--" means 1 level down, while "+-" = "".

So, for any tournament one can calculate the average strengths of players as an average of points corresponding to player's "levels" (neglecting "+" or "-"). That component will be very low for novice tournament but can be very high for "masters" tournament. For the record, the highest value of this component equal to 3.65 has been recorded for MCR EMA tournaments ## 269, 271 and 283.

Top ranking lists

After processing of 310 tournaments one can find almost 2,800 (!) names. Since there is an obvious "accessibility" issue meaning that some tournaments are not present in analysis (like many "medium-size" Asian tournaments, any local European like Italian "itinerary" ones etc.) only top ranking lists will be presented below.

Players' rating have been processed based on ranking system adopted in Russia: 25% of score is taken from player's level, 75% is based on top 5 tournaments within last 3 years (weighted sum, weights fall by 0.08 each 3 months, from 1.00 to 0.12 in a range of 3 years). So only 5 "big-and-fresh" tournaments matters.

Top 100 players as of April 01, 2018