Talk:Candidates Tournament 2020–2021

Coronavirus impact

Hi, Adpete! I'm the guy who cooperated with you on the Coronovirus section a number of times and you did great by correcting my last edits. Also, I saw you were not quite satisfied with "Altered playing conditions" section. I actually suggest to change it to: "FIDE and Russian authorities reaction" instead. It will adequately reflect all the current and future changes as it is a broader definition. As you can see, this changes very quickly now. --2601:1C0:CB01:2660:5485:E39B:BE62:3C20 (talk) 02:43, 12 March 2020 (UTC)

In my opinion: there is no point mentioning Russian authorities in the title unless they are also in the text. If the Russian authorities do something to change the tournament conditions, we can change the title then. Adpete (talk) 11:45, 12 March 2020 (UTC)

P/S: I even found an article in Russian (from 2-3 days ago) from Yekaterinburg where the hosts downplay the coronovirus factor and do not understand the move of Radjabov to withdraw. - I'm talking about official jury member - Sergei Shipov (he is a jury together with Judith Polgar)--2601:1C0:CB01:2660:5485:E39B:BE62:3C20 (talk) 02:47, 12 March 2020 (UTC)

Sources in Russian:

• https://www.1news.az/news/teymur-radzhabov-okazalsya-prav-tuchi-nad-turnirom-pretendentov-sguschayutsya

(analysis of Radjabov withdrawal and FIDE critics) March, 10

• https://www.sports.ru/chess/1084065906.html

(Sergei Shipov (the host country jury on the tournament) criticizes "hysteria" and Radjabov for his decision (March 8,) - this article clearly shows Russian authorities not full understanding of a very quickly changing situation - they still think here Europe is a safe place to come from. --2601:1C0:CB01:2660:5485:E39B:BE62:3C20 (talk) 02:55, 12 March 2020 (UTC)

There is an article on chess.com outlining the medical precautions being taken with the players and staff. Would this be relevant to include? I don’t want to detract too much from the tournament itself. Here is the link: https://www.chess.com/news/view/coronavirus-testing-at-fide-candidates-chess-tournament BobertWABC (talk) 06:37, 16 March 2020 (UTC)

round by round descriptions

I wrote descriptions of the 1st 2 rounds, then decided to remove them. I decided it was too much effort to write then to a good standard; they only exist in partial form (if at all) for previous Candidates, e.g. in Candidates Tournament 2016 people gave up after round 6; and no one is going to care in a few weeks' time, except for the critical games. So instead I have put in links to daily reports, in a Round by round reports section. I like using Chessbase links, because their articles stay for the long term (the Chessbase links still work from 2004), but I encourage editors to add links to other good sites too. Adpete (talk) 08:53, 19 March 2020 (UTC)

Ranking

Why in the standing section does the table show the first three with the same rank (1-3), but 4 and 5 are individual when they have the same score, shouldn't they have the rank 4-5? — Preceding unsigned comment added by Bensational (talk • contribs) 01:34, 21 March 2020 (UTC)

In the events of a tie by points, the tiebreak criteria are applied as described below the table. First, individual encounters: none of the leading three players have faced each other, neither had players 4 and 5. Next is the number of victories. All leading players have a score of +1-0=2, ranking them equal. Meanwhile, Caruana has one victory (and one loss), but Grischuk has no victories (and no losses), placing Caruana ahead. (Above question and this answer are accurate as of after round 3) 88.200.136.72 (talk) 05:12, 21 March 2020 (UTC)

Ok, thanks, what separates 7th and 8th? — Preceding unsigned comment added by Bensational (talk • contribs) 07:05, 21 March 2020 (UTC)

Next in the criteria list, namely Sonneborn–Berger score. Giri has his draws against MVL and Wang (who have two points each), giving him total score of 2, while Alekseenko's draws are against Nepomniachtchi (two points) and Grischuk (1.5 points), giving him a total score of 1.75. (It so happens that for the three leading participants even Sonneborn–Berger score is equal at 2.25 each, so their tie remains unbroken.) 88.200.136.72 (talk) 08:23, 21 March 2020 (UTC)

I think players on the same score should have the same rank. That is how any news site would report them. However, it might be good to include the tie breaks in the table, like Candidates Tournament 2014. Adpete (talk) 09:54, 22 March 2020 (UTC)

I added in the SB scores. It's kinda hacky as far as the module goes (I used away goals in place of SB), but it works. 2601:2C4:C400:8EA0:FC71:F76:5CFA:4ECB (talk) 00:36, 23 March 2020 (UTC)

I also added in head-to-head. At first, I gave it its own column, by using the rw_NAME option, but this was a bit hacky. So instead, I used the built in hth_NAME, which places a note attached to the points total. I think that a separate column might be preferred in chess (since points don't differentiate so much), but not so much that we go with a hacky option. Especially since if you leave out "|rwpoints=0" you can end up with some errors (the formula for points uses 1*wins + 0.5*draws + rwpoints*rw). I will leave it up to a discussion here for what the final decision is on how to display head-to-head information.2601:2C4:C400:8EA0:FC71:F76:5CFA:4ECB (talk) 02:54, 23 March 2020 (UTC)

Also, should we be using the "|pts_first=true" option, so that points come before all of the tiebreakers? 2601:2C4:C400:8EA0:FC71:F76:5CFA:4ECB (talk) 02:59, 23 March 2020 (UTC)

Good job! I think tie breaks are better off "out of the way" on the right of the table, as at Candidates Tournament 2014. That probably means using a different template. That would also allow us to get rid of the "Pld", "W", "D" and "L" tables - we don't really need them, they are more relevant to soccer when not all teams have played the same number of games. However, if that is too much work, what we have is better than nothing. Also, I still think players on equal points should be shown as equal ranking, so e.g. now (after Rd 5), Wang, Grischuk and Caruana should all be placed as 3-5; and that is how they are placed at the official site.[1] Adpete (talk) 06:13, 23 March 2020 (UTC)

This template already has options to suppress the columns you don't want. Adding the following

|pts_first=true
|hide_played=yes
|show_draw=no
|show_loss=no

along with the earlier hacky HTH coumn gives

Rank	Player	Pts	W	HTH	SB		NEP	MVL	WAN	GRI	CAR	DIN	GIR	ALE
1	Ian Nepomniachtchi (RUS)	3.5	2	0	8		—		1	½	½		1	½
2	Maxime Vachier-Lagrave (FRA)	3	1	0	6.5			—		½	½	1	½	½
3	Wang Hao (CHN)	2.5	1	0.5	5.25		0		—	½		1	½	½
4	Alexander Grischuk (RUS)	2.5	0	0.5	6.5		½	½	½	—		½		½
5	Fabiano Caruana (USA)	2.5	1	0	6.25		½	½			—	0	½	1
6	Ding Liren (CHN)	2	1	0.5	4.75			0	0	½	1	—	½
7	Anish Giri (NED)	2	0	0.5	5		0	½	½		½	½	—
8	Kirill Alekseenko (RUS)	2	0	0	5.75		½	½	½	½	0			—

Updated to match(es) played on 22 March 2020. Source: Official website
Rules for classification: 1) points; 2) head-to-head score among tied players; 3) total number of wins; 4) Sonneborn–Berger score (SB); 5) tie-break games.

Note: Numbers in the crosstable in a white background indicate the result playing the respective opponent with the white pieces (black pieces if on a black background).

Apparently, previous Candidates have already specified that we will not use notes, so the hacky hth column has been restored. I added a comment not to remove |rwpoints=0, since it has already been removed once, and if it gets removed, the points totals are wrong. In the mean-time, I have posted a comment in the module talk thread asking for an option for a separate hth column. Who knows if this will be created though. 2601:2C4:C400:8EA0:FC71:F76:5CFA:4ECB (talk) 15:01, 23 March 2020 (UTC)

I think the table above is better than the one currently in the article [2], because the one in the article has unnecessary columns. But really I think we should do the table manually; these templates are written for sports like soccer and simply are not suitable for chess with multiple tie breaks. But I am not volunteering, so the soccer table is better than nothing.

You still haven't answered my point that the people on the same score should be given the same rank, because (a) tie breaks at this point make little sense, and more importantly (b) that is what is done by reliable sources such as the official site [3]. I would really like to see that change unless someone can give me a good reason to override the official source. Adpete (talk) 00:13, 24 March 2020 (UTC)

p.s. I see an editor has now removed the tie breaks from the table. I think there is some benefit in including them so they should stay; I just think they should not affect the rankings. Adpete (talk) 03:13, 24 March 2020 (UTC)

Tie breakers

Having an edit war over tie breakers is stupid. And yet, it seems that this is happening. At least one editor does not want the tie breakers to be used mid-tournament, and so does not want tie breaker information to be included in the standings. Other editors do want tie breakers to be used mid-tournament, and so do want the tie breaker information to be included in the standings. I tried to come up with a compromise such that tie breaker information would be included in the table, but the rank on the left side of the table would not take those tie breakers into account. However, trying to come up with a compromise without discussing it with either of the parties involved never actually works. Which brings me to the fact that these editors are not discussing this in the talk thread, as they should be. And so here we are. 2601:2C4:C400:8EA0:3857:8234:FA7D:CA3C (talk) 00:08, 25 March 2020 (UTC)

Yes, that is exactly the compromise I suggested above. Unless the edit warring editors are prepared to give reasons here, they should be reverted. My reasoning for not doing rankings is (a) they fluctuate too much (in particular the head-to-head scores) to be meaningful except in the last 2 rounds or so, and (b) the most important reliable source, the official site, puts players on the same score as the same rank. But I am happy for them to be included because it is of some interest, in particular for tracking number of wins and SB tie break (though to be meaningful, SB tie break should probably be extrapolated on the assumption that all remaining games are drawn, perhaps calling it something like "extrapolated SB"). Adpete (talk) 01:19, 25 March 2020 (UTC)

Wow, I didn't know there was a talk page for each wikipage. That's cool. I'll read all discussions tomorrow. Joserobjr (talk) 01:35, 25 March 2020 (UTC)

To be fair, chess.com does include tiebreaks in their rankings, updated round by round. So I can see it being included. I don't care myself which way we end up doing things, as long as there is no edit war going back and forth. That doesn't work at all. 2601:2C4:C400:8EA0:3857:8234:FA7D:CA3C (talk) 02:07, 25 March 2020 (UTC)

Actually, I (the person who wanted to include tiebreakers in this editing war) don't mind ignoring the tiebreaker rules in this current tournament situation. Of course, it doesn't make much sense to analyze the HTH score, when three or more people are tied, and some have played with each other more times than others have. However, I would like the tiebreakers to be analyzed, when there were not three or more players tied, or when the tournament was in the final stretch. Because in these situations, the players involved in the tie would have played all their games, and or would have a decisive HTH score (one player won the other, and no more to be done by the losing player, except to overcome him in points) , or they would have a tie HTH, which can never again be used as a tiebreaker rule between these players, that is, in which case we could use the following tiebreaker cases (like the number of wins) without any prejudice in the analysis. Can anyone tag the other editor involved in this editing war, in this discussion? I haven't edited wikipedia for a long time (something I did a lot of years ago, in Portuguese-speaking wikipedia), and I don't remember how to tag someone anymore. Joserobjr (talk) 10:00, 25 March 2020 (UTC)

Again. An edit war is stupid. Additionally, information has been lost during the edit war, and someone is trying to use SB score ahead of wins, even though the tiebreak rules are clear. Come here to talk instead of reverting please. 2601:2C4:C400:8EA0:3857:8234:FA7D:CA3C (talk) 15:33, 25 March 2020 (UTC)

To bolster Adpete's point, FIDE itself is not applying tiebreak criteria mid-tournament. The vandalistic WP:NOTHERE IPs have no ground here, especially when the table clearly links to the Official website via the |source= parameter. CaradhrasAiguo (leave language) 15:36, 25 March 2020 (UTC)

To be fair, at one point the chess.com rankings from round 6 were included in the source, and those do use tiebreaks. But those aren't updated until hours after the games end. 2601:2C4:C400:8EA0:3857:8234:FA7D:CA3C (talk) 15:38, 25 March 2020 (UTC)

Also, there are two IPs who haven't been showing up in the talk thread, but I think that one of them was trying to add SB scores faster than I was, since I was admittedly kinda slow on that. So I don't really fault him/her for that. 2601:2C4:C400:8EA0:3857:8234:FA7D:CA3C (talk) 15:45, 25 March 2020 (UTC)

Scratch that. The second IP just participated in edit war without visiting talk thread. 2601:2C4:C400:8EA0:3857:8234:FA7D:CA3C (talk) 15:53, 25 March 2020 (UTC)

The Ukrainian IP is literally using a Round 6 article to cite Round 7 standings, all the while insisting they're right. Let that sink in. CaradhrasAiguo (leave language) 16:12, 25 March 2020 (UTC)

What if we add a table like that:

Rank		Player	Stats				Matches
Current	Projected		Pts	HTH	W	SB	MVL	NEP	CAR	GIR	WAN	GRI	DIN	ALE

Joserobjr (talk) 16:09, 25 March 2020 (UTC)

The module makes things a bit easier. We would have to make a whole new module or template. If you want to do that you can, but I'm not putting in that work. 2601:2C4:C400:8EA0:3857:8234:FA7D:CA3C (talk) 16:15, 25 March 2020 (UTC)

Giri and Wang Hao have the same score, same S-B and drew with each other. Do we give Giri 4th place on the basis that he drew with black against Wang Hao? MaxBrowne2 (talk) 08:16, 26 March 2020 (UTC)

One of the IP users that has not shown up in the talk thread changed Giri to be first because of his higher FIDE ranking. Nobody else seems to have cared, and that is as sensible of a decision as any. NHammen (talk) 14:34, 26 March 2020 (UTC)

In the "Results by round" section, remove the dates for Rounds 8 onward and replace with "TBD"

In the "Results by round" section, remove the dates for Rounds 8 onward and replace with "TBD" — Preceding unsigned comment added by 2601:645:C200:8D20:5136:B75:1802:1EEA (talk) 21:31, 28 March 2020 (UTC)

Done, thanks. Adpete (talk) 03:36, 30 March 2020 (UTC)

Sports table has a chess module that I am going to try

We are currently using the Sports table WDL module. I am going to try the Chess module instead, to see if that might work better. The benefit is that if you enter the match result, the wins and points are calculated automatically. This module used to have an H2H=0 changes to H2H is mdash problem, but I fixed that in the module code. I will also try to add SB score to the module.

Rank	Player	Pld	Score	H2H	Wins	SB		MVL		NEP		CAR		GIR		WAN		GRI		DIN		ALE
1	Maxime Vachier-Lagrave (FRA)	7	4 1⁄2	1	2	15.25		—		1		½			½		½	½		1			½
2	Ian Nepomniachtchi (RUS)	7	4 1⁄2	0	3	14.25			0	—			½		1	1		½		1			½
3	Fabiano Caruana (USA)	7	3 1⁄2	1 1⁄2	1	12.25			½	½		—			½	½			½		0	1
4–5	Anish Giri (NED)	7	3 1⁄2	1 1⁄2	1	11.25		½		0		½		—			½	½			½		1
4–5	Wang Hao (CHN)	7	3 1⁄2	1 1⁄2	1	11.25		½			0		½	½		—			½		1	½
6	Alexander Grischuk (RUS)	7	3 1⁄2	1 1⁄2	0	12.25			½		½	½			½	½		—		½		½
7	Ding Liren (CHN)	7	2 1⁄2	1⁄2	1	8.25			0		0	1		½		0			½	—		½
8	Kirill Alekseenko (RUS)	7	2 1⁄2	1⁄2	0	9.25		½		½			0	0			½		½		½	—

Updated to match(es) played on 25 March 2020. Source: Official website Chess.com
Rules for classification: 1) points; 2) head-to-head score among tied players; 3) total number of wins; 4) Sonneborn–Berger score (SB); 5) tie-break games.

Note: Numbers in the crosstable in a white background indicate the result playing the respective opponent with the white pieces (black pieces if on a black background).

NHammen (talk) 22:50, 4 April 2020 (UTC)

Yessss! I got SB score to automatically calculate! NHammen (talk) 00:09, 5 April 2020 (UTC)

Rank	Player	Pld	Score	H2H	Wins	SB		MVL		NEP		CAR		GIR		WAN		GRI		DIN		ALE
1	Maxime Vachier-Lagrave (FRA)	7	4.5	1	2	15.25		—		1		½			½		½	½		1			½
2	Ian Nepomniachtchi (RUS)	7	4.5	0	3	14.25			0	—			½		1	1		½		1			½
3	Fabiano Caruana (USA)	7	3.5	1.5	1	12.25			½	½		—			½	½			½		0	1
4–5	Anish Giri (NED)	7	3.5	1.5	1	11.25		½		0		½		—			½	½			½		1
4–5	Wang Hao (CHN)	7	3.5	1.5	1	11.25		½			0		½	½		—			½		1	½
6	Alexander Grischuk (RUS)	7	3.5	1.5	0	12.25			½		½	½			½	½		—		½		½
7	Ding Liren (CHN)	7	2.5	0.5	1	8.25			0		0	1		½		0			½	—		½
8	Kirill Alekseenko (RUS)	7	2.5	0.5	0	9.25		½		½			0	0			½		½		½	—

Updated to match(es) played on 25 March 2020. Source: Official website Chess.com
Rules for classification: 1) points; 2) head-to-head score among tied players; 3) total number of wins; 4) Sonneborn–Berger score (SB); 5) tie-break games.

Note: Numbers in the crosstable in a white background indicate the result playing the respective opponent with the white pieces (black pieces if on a black background).

Version with |disp_fractions=no and |match_style=colorbox , which is the options I have chosen for the article. disp_fractions determines whether the points and H2H are displayed as fractions or decimals, and match_style can be 'tall', 'colorbox' or 'none', which affects the match table. NHammen (talk) 00:33, 5 April 2020 (UTC)

@NHammen: I am looking to implement this on the other tournaments dating back to 2013. How does one implement the reverse colors? As in, say, Wang Hao draws with the white pieces against Ding in Round 8. Should the code be |match_WAN_DIN= ½++1 or |match_WAN_DIN= ½+;+1? CaradhrasAiguo (leave language) 02:22, 5 April 2020 (UTC)

|match_WAN_DIN= ½+1

The idea is |match_WAN_DIN= W+B where W is the score as W and B is the score as B NHammen (talk) 02:52, 5 April 2020 (UTC)

Nuts. |hide_played=yes doesn't work with the Chess version of the module. Guess I need to port some stuff over from WDL.

@CaradhrasAiguo: So I looked back at the older tables. For 2016 and after, they used WDL, so if there are any features from WDL that you want (other than the ability to hide number of games played) let me know. For before that, some of the features of those tables might be difficult to replicate. I especially like the way that they have their match table. It's just a lot of effort to edit that, even if it looks neat. NHammen (talk) 03:05, 5 April 2020 (UTC)

D'oh! That W+B somehow eluded me. I think the Chess sub-module reduces the room for manual error and is neater within matches, whereas the boxes previously had to be hard-coded in. CaradhrasAiguo (leave language) 03:08, 5 April 2020 (UTC)

@CaradhrasAiguo: @Adpete:, my changes to Module:Sports table just went through, so now we have a better crosstable for Chess. Because this was an update to the module, any page that was using Module:Sports table/Chess has automatically updated to the new crosstable. Pages that used the WDL style still have the old match table. Because this crosstable is better than the other options that were available, I think I shall start removing some of the code for the old crosstable styles which is currently unreachable, instead of altering it to be reachable. Other than that, I think I now have most of the features that we need for chess, unless we want some functionality for matches with more than 2 games between each pair of players, which would take a bit of work. NHammen (talk) 18:01, 8 April 2020 (UTC)

Excellent work with hiding the H2H between players with different scores! I already made sample code for each of the previous four Candidates tournaments here. I think the days of triple- or quadruple-round robin tournaments (e.g. the 1959 Candidates) are over, so that can be considered to not be a priority. CaradhrasAiguo (leave language) 18:12, 8 April 2020 (UTC)

@CaradhrasAiguo: I notice that in the article for Candidates Tournament 2014 that there is a rating column, which you don't have in your version of the table, as well as a few visual features that are different. If you want to include those features, you can include these options in the table for 2014:

|solid_cell=lightgray |hide_played=yes |show_rating=yes |rate_date=March 2014

|rating_ANA=2770

|rating_AND=2709

|rating_ARO=2830

|rating_KAR=2766

|rating_KRA=2787

|rating_MAM=2757

|rating_SVI=2758

|rating_TOP=2785 NHammen (talk) 19:15, 8 April 2020 (UTC)

I am indifferent to adding ratings to the candidates tables, because the monthly FIDE rating going into the tournament is included in the #Qualification section of each tournament. For Tata Steel, Sinquefield, etc., fine. CaradhrasAiguo (leave language) 21:45, 8 April 2020 (UTC)

Given that there are no more changes that need to be made to the code, I spent the past day writing documentation for the module. So now Module:Sports_table/Chess has documentation. Yay! NHammen (talk) 22:04, 9 April 2020 (UTC)

I haven't had a detailed look, but that is looking really good. Thank you for your work! Adpete (talk) 00:01, 10 April 2020 (UTC)

Candidates Tournament 2020?

Hi, Adpete and whom it may interest, as the tournament becomes biennial, wouldn't it be appropriate to move the title to "Candidates Tournament 2020-2021"? We have official confirmation of the FIDE about it.--2601:1C0:CB01:2660:28F2:47EA:261F:5510 (talk) 08:06, 21 October 2020 (UTC)

I am in no rush to change it, because "Candidates Tournament 2020" is probably easier to search for. But we probably should. Adpete (talk) 03:44, 27 October 2020 (UTC)

Adpete is right. We should wait until the Candidates Tournament resumes and probably even ends well (coronavirus-wise). Then, we have 2 working options:

• Option A - to rename the page to Candidates Tournament 2020-2021 or something of that kind. We can also add "redirect" for both pages titles leading to the same page - this is the most appropriate way to do, I think. Also, saving some Wikipedia space which is not indefinite.
• Option B - to create another page for Candidates Tournament 2020-2021 with both the pages linked with "Main article" blue links. In this scenario, we have to justify why we need 2 pages as "redirect" solves all the "search issues" easily for the Wikipedia readers. Personally, I'd prefer moving the title with redirect but let's wait until this pending tournament is over.

2601:1C0:CB01:2660:51D4:499C:A977:FB8 (talk) 03:59, 2 November 2020 (UTC)

Definitely option A, because it is a single tournament. Wikipedia will make the redirect (from "CT 2020" to "CT 2020-21") automatically when we do this. I do not care if we do the rename now, or later. Adpete (talk) 04:37, 2 November 2020 (UTC)

Probably, it is a better idea to survive until spring 2021 and see that tournament start=)). Stay safe and well — Preceding unsigned comment added by 2601:1C0:CB01:2660:B5CA:D2BC:E8EE:825A (talk) 22:46, 2 November 2020 (UTC)

Actually after more thought, it is clear the event is not going to complete in 2020, so why not rename it to Candidates Tournament 2020-2021 now? Candidates Tournament 2020 is now clearly wrong. There is a faint chance something weird will happen (like it being abandoned, or further postponed to 2022), but I think that is very unlikely; and if that does happen, it is no great problem to rename it again. Actually for me, the biggest issue is how to do the hyphen: is it 2020-21 or 2020–21? Adpete (talk) 08:47, 3 November 2020 (UTC)

Adpete, yes it is a good idea to do it now. I'm not an expert in hyphens, but 2020–21 looks a bit better to me personally. Also, I think the best way to do it is to check any other biennial tournaments or events and see how they managed it in the past on Wikipedia. 2601:1C0:CB01:2660:246B:C255:10B4:BE61 (talk) 18:42, 5 November 2020 (UTC)

I found the answer: we should use "2020–21". The policy is MOS:RANGES and you can see examples at all the NBA seasons, e.g. 2019–20 NBA season. Adpete (talk) 23:23, 5 November 2020 (UTC)