17

I've tried finding statistics on the results of competitive armageddon games played at a high level (say, GM level), but to no avail. Evidently, these statistics must exist, as for example FIDE changed the time controls in the official rules from 6 minutes-5 minutes to 5 minutes-4 minutes in preparation for the Carlsen-Karjakin WC match (thus indicating that they felt that 6-5 gave black too much of an advantage compared to 5-4).

Basically, does white really have a ~50% win rate in GM-level armageddon chess games?

BCLC
  • 1,892
  • 1
  • 17
  • 46
ATLPoly
  • 642
  • 4
  • 10
  • 1
    I am also not able to find any statistics online, but it seems apparent that black has a massive advantage especially at high level games where draws are fairly common. Personally I feel like there needs to be a massive time difference, like 5 minutes to 2.5 minutes for white to have a 50% chance of winning. If you even look at the quality of games of GMs at full time control and rapid, they are almost the same. Time is just not nearly as important when GMs have memorized almost everything. – Matthew Liu Nov 08 '18 at 22:47
  • 2
    Welcome to the site, great first question! – Glorfindel Nov 09 '18 at 08:41
  • 1
    @Matthew Liu I have to strongly disagree here. While GMs can play nearly identically in the opening, blunders later on in the game have a much higher chance of appearing. Also, shuffling pieces while trying to flag your opponent in an equal position is always a factor. I'd say 5-4 are fair odds for super GMs, but at the master level it favors White. – Inertial Ignorance Jan 05 '19 at 03:49
  • @Inertial Ignorance A 5-4 minute advantage is practically nothing at GM level. Flagging your opponent in a dead drawn position is not really an option because there is a delay. Look at http://universalrating.com/ratings.php. If the blitz gap is only 28 then how can a 1 minute time odd do anything? Carlsen can play with 5 minutes to the opponents 90 and still beat everybody else. Also people berserk on lichess all the time in short time controls with no delay where flagging is a huge factor. – Matthew Liu Jan 07 '19 at 16:10
  • If both opponents use 4 minutes of time, you're now down to a 2 minute vs 1 minute situation. What argument do you have that supports a super GM with 1 minute being on even footing with a super MG with 2 minutes? Even if there is a 2 second delay (for Armageddon I think there isn't one), you can keep your opponent on the brink for flagging for a very long time while you press. GMs are good at doing this. Finally, Carlsen is able to beat amateurs with 5 minutes vs 90 minutes, not other super GMs. – Inertial Ignorance Jan 07 '19 at 18:52
  • 2
    I disagree nobody has any interest in answering this question. It involves considerable research, as time controls aren't tracked precisely by FIDE, but people are interested. – aschultz Aug 04 '19 at 04:37
  • 1
    No statistics that I know of since armageddon games are pretty rare. – Inertial Ignorance Sep 28 '19 at 22:50

3 Answers3

12

The earliest Armageddon games I can find go back to the Women's World Chess Championship 2001, and the FIDE World Championship in 2002, which GM Ruslan Ponomariov won. This is probably a fairly complete list since Armageddon really only lends itself to knock-out tournaments or matches, and the question did ask primarily about GMs and Armageddon. There might be the occasional country championship that goes to tiebreaker, which I did not find, similar to when GM Irina Krush lost to IM Anna Zatonskih in 2008 for the U.S. Women’s Championship in a controversial finish.

So, while these stats may not be 100% complete, they can give a good idea about the overall statistics.

I am amazed by the results of this: Out of the 48 games I found, 24 were 1-0, 15 were 0-1, and 9 were ½-½. If you look at that both the 0-1 and ½-½ count as wins for black, then the result was exactly 24-24. Seemingly quiet a fair tiebreaker system.

Here is the data I used to compile this estimate.

FIDE World Chess Championship 2002

This was the only FIDE World Championship to use Armageddon during the period of when Kasparov broke off with the title until it was unified in 2006. Prior to this, in 1998 and 2000, they continued playing additional blitz games, and in 2005, FIDE reverted to a double-round-robin tournament.

GM Zoltan Gyimesi defeated GM Stuart Conquest 4-3 in round one. Gyimesi had white 1-0. GM Emil Sutovsky defeated GM Alonso Zapata 4-3 in round one. Sutovsky had white 1-0. GM Lazaro Bruzon defeated GM Liviu-Dieter Nisipeanu 4-3 in round one. Bruzon had white 1-0. GM Jaan Ehlvest defeated GM Ilia Smirin 4-3 in round three. Ehlvest had white 1-0. GM Alexey Shirov defeated GM Veselin Topalov 4-3 in round four. Shirov had black 0-1.

Chess World Cup 2005

GM Yuri Shulman defeated GM Vadim Zvjaginsev 3.5-3.5 in round one. Shulman had black ½-½. GM Predrag Nikolić defeated GM David Navara 3.5-3.5 in round one. Nikolić had black ½-½. GM Sergey Erenburg defeated GM Zviad Izoria 4-3 in round one. Erenburg had black 0-1. GM Evgeniy Najer defeated GM Shakhriyar Mamedyarov 4-3 in round two. Najer had white 1-0. GM Yuri Shulman defeated GM Vadim Zvjaginsev 3.5-3.5 in round two. Shulman had black ½-½. GM Sergei Tiviakov defeated GM Oleg Korneev 4-3 in round two. Tiviakov had white 1-0. GM Loek van Wely defeated GM Alexander Moiseenko 4-3 in round two. van Wely had white 1-0.

Chess World Cup 2007

GM Michael Roiz defeated GM Varuzhan Akobian 4-3 in round one. Roiz had white 1-0. GM Konstantin Sakaev defeated GM Nikita Vitiugov 4-3 in round one. Sakaev had black 0-1. GM Zhou Jianchao defeated GM Andrei Volokitin 4-3 in round two. Zhou had white 1-0. GM Kiril Georgiev defeated GM Rustam Kasimdzhanov 4-3 in round two. Georgiev had white 1-0.

Chess World Cup 2009

This year, by some accounts a mistake, the tiebreaker was changed to four rapid games, then up to five pairs of blitz games, and only then, Armageddon. Not surprising, but there were no Armageddon games, however, GM Varuzhan Akobian did defeat GM Pavel Tregubov 9-7, just one game short of Armageddon.

Chess World Cup 2011

Starting this year, they went back to the previous tiebreak format, but with two shorter 10+10 rapid games added after the 25+10 rapids. GM Yuri Drozdovskij defeated GM Alexander Motylev 5-4 in round one. Drozdovskij had white 1-0. GM Leinier Domínguez Perez defeated GM Igor Lysyj 5-4 in round three. Domínguez Perez had white 1-0.

Chess World Cup 2013

GM Evgeny Tomashevsky defeated GM Alejandro Ramírez 5-4 in round one. Tomashevsky had white 1-0. GM Julio Granda defeated GM Hrant Melkumyan 5-4 in round one. Granda had black 0-1. GM Daniil Dubov defeated GM Ruslan Ponomariov 5-4 in round two. Dubov had white 1-0.

Chess World Cup 2015

GM Gabriel Sargissian defeated GM Mateusz Bartel 4.5-4.5 in round one. Sargissian had black ½-½. GM Michael Adams defeated GM Viktor Láznička 5-4 in round two. Adams had white 1-0. GM Hikaru Nakamura defeated GM Ian Nepomniachtchi 5-4 in round three. Nakamura had black 0-1.

Chess World Cup 2017

GM Levon Aronian defeated GM Maxime Vachier-Lagrave 5-4 in the semi-finals. Aronian had white 1-0.

Chess World Cup 2019

GM Anish Giri defeated GM Evgeniy Najer 5-4 in round two. Giri had black 0-1. GM Yu Yangyi defeated GM Nikita Vitiugov 5-4 in the quarter-finals. Yu had white 1-0.

Women's World Chess Championship 2001

Note: I could not confirm the format with 100% accuracy. I am reasonably guessing that this used the same format as the men, and the women in 2008, which I could confirm. In addition, the brackets had the 4-3 results as in the men’s tournaments.

GM Zhu Chen defeated GM Maia Chiburdanidze 4-3 in the semi-finals. (no data)

Women's World Chess Championship 2004

GM Peng Zhaoqin defeated IM Joanna Dworakowska 4-3 in round one. Pend had black 0-1. WGM Ketino Kachiani defeated WGM Nguyễn Thị Thanh An 3.5-3.5 in round one, Ketino Kachiani had black ½-½.
IM Tatjana Vasilevich defeated WGM Xu Yuanyuan 4-3 in round one. Vasilevich had black 0-1. IM Lilit Mkrtchian defeated then WGM Monika Soćko 4-3 in round one. Mkrtchian had white 1-0. Then WGM Kateryna Lagno defeated then untitled Ekaterina Atalik 4-3 in round two. Lagno had black 0-1.

Women's World Chess Championship 2006

None.

Women's World Chess Championship 2008

IM Elisabeth Pähtz defeated WGM Ilaha Kadimova 4-3 in round one. Pähtz had white 1-0. GM Monika Soćko defeated WGM Sabina-Francesca Foisor 4-3 in round one. Soćko had white 1-0.

Women's World Chess Championship 2010

Then IM Harika Dronavalli defeated then IM Mariya Muzychuk 4-3 in round three. Harika had black 0-1 (on time in a lost position!).

Women's World Chess Championship 2012

This was the first year of the extended tiebreaker with the rapid 10+10 games added. Then IM Mariya Muzychuk defeated WGM Cristina Adela Foișor 4.5-4.5. Muzychuk had black ½-½.

Women's World Chess Championship 2015

GM Tatiana Kosintseva defeated WGM Mary Ann Gomes 5-4 in round one. Kosintseva had black 0-1. Then WGM Aleksandra Goryachkina defeated IM Lilit Mkrtchian 5-4 in round one. Goryachkina had white 1-0.

Women's World Chess Championship 2017

IM Anastasia Bodnaruk defeated WGM Mitra Hejazipour 5-4 in round one. Bodnaruk had white 1-0. GM Tan Zhongyi defeated GM Anna Ushenina 4.5-4.5 in round two. Zhongyi had black ½-½. IM Nino Khurtsidze defeated then IM Nino Batsiashvili 5-4 in round two. Khurtsidze had black 0-1. GM Tan Zhongyi defeated GM Harika Dronavalli 5-4 in the semi-finals. Tan had black 0-1.

Women's World Chess Championship 2018

GM Kateryna Lagno defeated WGM Natalia Pogonina 5-4 in round three. Lagno had white 1-0.

US Women’s Championship 2008

Then IM Anna Zatonskih defeated IM Irina Krush 3-2. Zatonskih was black, and white flagged, 0-1.

https://en.chessbase.com/post/us-women-s-armageddon-reactions-from-our-readers

World Chess Armageddon Series 2019

Nepomniachtchi -Karjakin 0-1 Radjabov-Karjakin ½-½ Radjabov-Dubov 1-0 Radjabov-Kramnik ½-½

https://www.chessbomb.com/arena/2019-world-chess-armageddon-series

PhishMaster
  • 32,477
  • 4
  • 105
  • 177
  • 6
    Great effort! I noticed there are no Carlsen games mentioned, so here are a few of his Armageddons that come to mind: Norway chess 2019 -> Carlsen-Anand 1-0, Aronian-Carlsen 0-1, Carlsen-Mamedyarov 1-0, MVL-Carlsen 1/2-1/2, Carlsen-Ding 1-0, So-Carlsen 1/2-1/2, Caruana-Carlsen 1-0. Grenke 2015 -> Carlsen-Naiditsch 1-0. Also World Blitz playoff 2006 -> Grischuk-Svidler 1-0. – Ellie Dec 05 '19 at 20:25
  • 1
    @Phonon Thank you. That took hours to do just that. First, researching every game in those tournaments to see who went to Armageddon, and then database searching every match. I, simply, did not remember his Armageddon games. Thanks for adding those. – PhishMaster Dec 05 '19 at 20:56
  • 2
    Thanks for your time. It was our top voted unanswered question. We are doing it even better This is very nice because new users feel they can freely ask a question and receive an answer. –  Dec 06 '19 at 10:47
  • 1
    Nice work. I don't know how much time that would require, but if it is practical, could you precise which games where 6'-5', 5'-4', or a time-control with increments ? – Evargalo Dec 06 '19 at 11:22
  • 1
    @Universal_learner Thank you for accepting my answer for the bounty. – PhishMaster Dec 06 '19 at 13:19
  • 1
    Very nice work, @PhishMaster! Just a note though: armageddon games at the GM level go back further than 2002, for example this one between Ivanchuk and Kramnik played in 1994. – ATLPoly Dec 07 '19 at 02:00
  • Great effort +1. Amazing there's no database or detailed compilation of Armageddon games and we seem to be making the first. My own data-points added below. – Mobeus Zoom Apr 18 '21 at 04:16
3

Just wanted to augment the excellent answer above with some further examples:

Smirin vs Anand, 1994, 0-1

Mamedyarov vs Lputian, 2004, 1-0

Ye Jiangchuan vs Ni Hua, 2004, 1-0

G. Sargissian vs Tiviakov, 2004, 0-1

Radjabov vs P.H. Nielsen, 2004, 1-0

L. Dominguez vs V. Malakhov, 2004, 1-0

Hamdouchi vs Kudrin, 2004, 1-0

P. Smirnov vs Aronian, 2004, 1-0

Dreev vs Sakaev, 2004, 1-0

Ni Hua vs Vladimirov, 2004, 1-0

Nisipeanu vs Kharlov, 2004, 0-1

L. Dominguez vs Radjabov, 2004, ½-½

Anand vs Nakamura, 2015, 0-1

Aronian vs Vachier-Lagrave, 2017, 1-0

Paikidze vs A. Wang, 2018, 1-0

Karjakin vs Giri, 2019, 1-0

Aronian vs Yu Yangyi, 2019, ½-½

Vachier-Lagrave vs Mamedyarov, 2019, 0-1

Harikrishna vs Karjakin, 2019, ½-½

Dubov vs Nakamura, 2020, 1-0

Carlsen vs Firouzja, 2020, 1-0

Caruana vs Carlsen, 2020, 0-1

Aronian vs Carlsen, 2020, 0-1

Nakamura vs Carlsen, 2020, 0-1

So vs Vachier-Lagrave, 2020, ½-½

Nepomniachtchi vs Radjabov, 2020, ½-½

Carlsen vs Dubov, 2021, 1-0

Giri vs van Foreest, 2021, 0-1

Mamedyarov vs Nakamura, 2021, 0-1

This adds 24 examples of which the results for White are +14 -10 =5. Combining this with PhishMaster's +24 -15 =9 and Ellie's examples (+6 -1 =2) we arrive at +44 -26 =16 or, with Black's draw odds, 44-42 in favour of White. Therefore, the tiebreaker seems pretty balanced but if anything, actually slightly advantages White.

(As an aside, Carlsen's record is Armageddon in the games we have assembled so far is 12-1.)

Mobeus Zoom
  • 2,279
  • 8
  • 19
-1

The answer is no, as theere is data. But it has not even been aggregated. Nobody is known to have done meaningful statistics on that data.

Armageddon is somewhat arbitrary as it tries to balance the effect of time versus the probability that black could draw or win so as to have a definitive play off winner instead of a long long string of drawn games.

But it is still a stretch to say that it truly represents the 'winner' of a tournament or a match at a much longer time control.

Rewan Demontay
  • 17,514
  • 4
  • 67
  • 113
edwina oliver
  • 3,881
  • 10
  • 29