PDA

View Full Version : Justin Tan in 14th IGB Dato Arthur Tan International Open Chess Championship



Leonid Sandler
03-09-2017, 01:45 AM
With one round to go in 14th IGB Dato Arthur Tan International Open Chess Championship being held in Kuala-Lumpur former Victorian,International Master Justin Tan with 6 points from 8 games is sharing first place with another ten players.

To score his last GM Norm Justin needs to win against Chinese GM Zhou Jianchao(2597).

The last round game starts at 11-00 am AEST.

Good luck Justin!

http://www.chess-results.com/tnr297801.aspx?lan=1&art=9&fed=AUS&wi=821&snr=16

Garrett
03-09-2017, 07:27 AM
How exciting !

Looks like the game might be live at http://www.datchesscenter.net/live/

Yesterday's games are there now.

Max Illingworth
03-09-2017, 10:00 AM
Very cool! Justin is not the only Australian excelling in Malaysia - Albert Winkelman is also performing way above his rating.

Andrew Hardegen
03-09-2017, 04:16 PM
How exciting !

Looks like the game might be live at http://www.datchesscenter.net/live/

Yesterday's games are there now.

The final round game Tan - Zhou Jianchao is showing as a win for Justin. But the final position after 44...Kd5 looks like a straightforward draw.

Frank
03-09-2017, 05:20 PM
The final round game Tan - Zhou Jianchao is showing as a win for Justin. But the final position after 44...Kd5 looks like a straightforward draw.

Chess-Results http://www.chess-results.com/tnr297801.aspx?lan=1&art=9&fed=AUS&wi=821&snr=16
shows draw

Metro
03-09-2017, 05:22 PM
Name Tan Justin
Title IM
Starting rank 16
Rating 2450
Rating national 1800
Rating international 2450
Performance rating 2561
FIDE rtg +/- 14,3
Points 6,5
Rank 12
Federation AUS
Ident-Number 0
Fide-ID 3206882
Year of birth 1997


Rd. Bo. SNo Name Rtg FED Pts. Res.
1 80 21 GM Neelotpal Das 2441 IND 6,0 w
2 37 103 Hakim Nuh 2096 INA 4,5 s 1
3 19 69 Pranav V 2248 IND 5,0 w 1
4 10 43 CM Gukesh D 2365 IND 4,5 s
5 9 37 GM Gonzales Jayson 2378 PHI 6,0 w 1
6 3 14 GM Fang Yuxiang 2477 CHN 7,0 s
7 7 32 IM Iniyan P 2404 IND 6,5 w 1
8 3 4 GM Tran Tuan Minh 2547 VIE 7,0 s
9 2 2 GM Zhou Jianchao 2597 CHN 6,5 w

Kevin Bonham
03-09-2017, 05:23 PM
Misses norm by half a point despite scoring 3/5 vs GMs.

Kai
04-09-2017, 09:56 AM
Thanks Leonid, Max, Garrett.

The organiser decided to use a tiebreak based on number of wins, to determine prizes: http://www.chess-results.com/tnr297801.aspx?lan=1&art=4&wi=821

So, in the end, Justin ended up with 12th prize (and GM Zhou Jianchao with 11th prize), despite the two of them having the highest RP in the group. Whereas the players who had performed more poorly (and playing and defeating lower rated opponents) ended up with higher prizes. (At one point, GM Venkatesh, who had been having a very poor start and thus played no-one rated above 2277, ended up ranked no.1). All the players with 2500+ RP ended up with the lowest prizes!

The ranking almost has an inverse relationship to RP as a result of this tiebreak (in effect, also rewarding those with the most losses and encouraging Swiss gambits). The top prize for the 6.5 group went to IM Iniyan, whom Justin defeated in R7. Using this tiebreak system to determine prizes seems to me to defy common sense.

NickC
04-09-2017, 06:52 PM
The organiser is clearly following the Recommended Tie-Break Systems shown in section G of the FIDE Competition Rules Annex3: Tie-Break Regulations (https://www.fide.com/component/handbook/?id=187&view=article), which for Individual Swiss Tournaments where all the ratings are consistent consists of the following tie-breaks in descending order:

Direct Encounter
Most Wins
Greater number of games with Black
Average Rating of Opponents Cut 1
Buchholz Cut 1
Buchholz
Sonneborn-Berger

Unfortunately this system of tie-breaks performs poorly, in fact the worst of 22 tiebreak hierarchies tested by Jesper Norgaard in his excellent analysis of tiebreaks in the Tiebreaks thread (http://www.chesschat.org/showthread.php?14182-Tiebreaks/page13). Which is hardly surprising, as Direct encounter rarely applies in large Swiss tournaments (since it requires that every player on a particular score has played each other) and Most Wins tends to favour 'Swiss gambit' players.
Jesper's analysis suggests that of the commonly available tie-breaks, the Progressive/Buchholz/AverageRatingofOpponents hierarchy is the most predictive but given that for an unknown reason Progressive currently cannot be applied according to the 2016 FIDE Arbiter's Manual (http://arbiters.fide.com/images/stories/downloads/2016/Arbiters-Manual-2016.pdf) (see page 148 if you're interested), at Norths we now use DirectEncounter/BuchholzCut1/AverageRatingofOpponentsCut1 which has produced more intuitive rankings.

Regardless of the tiebreak hierarchy, in my view they should only be used to break ties for indivisible prizes such as a trophy, and prize money should be shared equally between all players on the same score.

Kai
04-09-2017, 07:29 PM
The organiser is clearly following the Recommended Tie-Break Systems shown in section G of the FIDE Competition Rules Annex3: Tie-Break Regulations (https://www.fide.com/component/handbook/?id=187&view=article), which for Individual Swiss Tournaments where all the ratings are consistent consists of the following tie-breaks in descending order:

Direct Encounter
Most Wins
Greater number of games with Black
Average Rating of Opponents Cut 1
Buchholz Cut 1
Buchholz
Sonneborn-Berger

Unfortunately this system of tie-breaks performs poorly, in fact the worst of 22 tiebreak hierarchies tested by Jesper Norgaard in his excellent analysis of tiebreaks in the Tiebreaks thread (http://www.chesschat.org/showthread.php?14182-Tiebreaks/page13). Which is hardly surprising, as Direct encounter rarely applies in large Swiss tournaments (since it requires that every player on a particular score has played each other) and Most Wins tends to favour 'Swiss gambit' players.
Jesper's analysis suggests that of the commonly available tie-breaks, the Progressive/Buchholz/AverageRatingofOpponents hierarchy is the most predictive but given that for an unknown reason Progressive currently cannot be applied according to the 2016 FIDE Arbiter's Manual (http://arbiters.fide.com/images/stories/downloads/2016/Arbiters-Manual-2016.pdf) (see page 148 if you're interested), at Norths we now use DirectEncounter/BuchholzCut1/AverageRatingofOpponentsCut1 which has produced more intuitive rankings.

Thanks, but in this tournament, direct encounter wasn't even applied. The first line was most wins.

As I said, Justin defeated the top 6.5 prize-getter (Iniyan) in their direct encounter. Indeed, 3/4 of the bottom place-getters (12-14) defeated place-getters 5,6,8 in their direct encounters.

If direct encounter was applied as the first step, it would at least make some sense. In this case, it wasn't. So I don't know what system was applied here.

Yes, the system you apply at Norths would produce the most sensible ranking, but in any case, I agree with your view on sharing of prizes.

NickC
04-09-2017, 08:57 PM
Hi again Kai,

At the Chess Results link (http://www.chess-results.com/tnr297801.aspx?lan=1&art=4&wi=821) you provided for this tournament, the three tiebreaks listed at the bottom of page are:

Direct Encounter
Most Wins
Most Black

As per my last post these are the first three tiebreaks in the FIDE recommended hierarchy for Individual Swiss Tournaments where all the ratings are consistent, which is why I suspected that is what the organiser used.

Note that Direct Encounter only makes a difference if every player on a particular score has played each other, so in this tournament it never had any impact since this condition wasn't met for any score (as can be seen by the 0 for every player in the TB1 column). Therefore the first tiebreak to have any practical effect is Most Wins, which as you noted previously often produces perverse rankings.

However, the bigger issue is using tiebreaks to divide prize money for players on the same score, which we both agree is a bad idea!