What if the World Championship match is decided in blitz games?

by New In Chess
11/21/2018 – Two years ago, in New York, Magnus Carlsen defended his title by winning the rapid tiebreaker. Traditionalists will shudder at the thought of blitz games deciding the world championship match in London. GM MAXIM DLUGY looks at the players’ chances if that scenario becomes reality and has suggestions for how Fabiano Caruana, despite a whopping rating gap of 172(!) points, might beat the odds.

ChessBase 15 - Mega package ChessBase 15 - Mega package

Find the right combination! ChessBase 15 program + new Mega Database 2019 with 7.6 million games and more than 70,000 master analyses. Plus ChessBase Magazine (DVD + magazine) and CB Premium membership for 1 year!


New in Chess on the potential tiebreak scenario

As the World Championship between Magnus Carlsen and Fabiano Caruana was around the corner, and the bookmakers’ odds were only slightly in favour of the World Champion, there seems to be every reason to wonder what would happen if, after 12 classical games, we’d see a tiebreaker with a faster time-control. Whether there would be rapid games, as we saw in New York two years ago, when Magnus faced Sergey Karjakin,  or whether we would even go to blitz games if the rapid games yield no decision either.

NIC special issueWhile Magnus Carlsen also excels in rapid and blitz chess, Fabiano Caruana’s strength clearly lies in the classical time-control. Does this mean that Magnus has the pleasant option of just coasting to the tiebreaker, as he did against Karjakin once he had equalized the score with two games to go? Is the champ such a clear favourite in the faster time-controls that his challenger will have to steer clear of the tiebreak at all cost and put all his efforts into the first 12 games?

On the face of it — yes! At the time of writing, the difference between the two players in the classical time-control is 12 rating points, while the difference in the rapid ratings is 91 points, and the gap in the blitz ratings even a whopping 172 points!

From my discussions with former World Champion Anatoly Karpov, who had the privilege of talking to Arpad Elo, the inventor of the current rating system, I learnt that the difference in ratings is just an indication of the expected score between players in the long run, and in no way an indication of a foregone conclusion. In fact, Prof. Elo was annoyed when FIDE decided to use his system to include and exclude players from certain events, merely on the basis of these numbers — he never meant it to be used that way. 

Notwithstanding this, we live in a practical world, and it does seem to us that if a player is 170 points lower, he is the underdog, even if we shouldn’t expect the top dog to automatically score 7½ points in a 10-game match, as predicted by the rating system. This doesn’t bode well for Caruana, who it seems has shown serious weakness in playing blitz against the top players in the world. His recent 20½-6½ loss to Levon Aronian in the Chess.com championship created a sensation, while his struggling in the blitz part of the Grand Chess Tour has become the norm, rather than the exception.

Does that mean that Magnus only needs to tie the match, split the four rapid games and finish Fabiano off in the two-game blitz playoff? I don’t think it will be that simple. First of all, as any odds player knows, the positive expectation of results plays out very well over the long haul, but in a short match, anything can happen. Even if Fabiano’s expectation in a two-game blitz match is half a point out of 2, it’s very easy to beat the odds, either by catching your opponent with an opening novelty and winning one game or simply by managing to draw both games, which would take the match to a totally unpredictable Armageddon shoot-out. The nervous tension in those two blitz games would be as high as with a throw-in in the final seconds of a basketball game with the shooter’s team down one point. Here, too, the ability to control your nerves will be the main factor.

To showcase what can happen when time runs short, I would like to direct the reader to the following incredible game, which may hopefully demonstrate the kind of situation that could arise in the final countdown. This blitz game, played last year between Fabiano and Magnus in the Paris leg of the Grand Chess Tour, demonstrates how a lack of time can make the best players in the world come up with seemingly random moves.


The Blitz Whisperer column, by Max Dlugy
Republished with kind permission of 
New in Chess magazine

Free 50-page Carlsen-Caruana Match Special

special World Championship issue includes a preview by Jan Timman and an interview with Sergey Karjakin about his 2016 match with Carlsen (who confesses to overlooking dedicated training in rapid in advance of the match)

About New In Chess Magazine

New In Chess is read by club players in 116 countries.

  • 8 issues a year
  • 800 pages a year of the very best in chess
  • game annotations by the best players on the planet 
  • on-the-spot tournament reports
  • columns both serious and hilarious
  • in-depth interviews
  • 'Secrets of Opening Surprises' 
  • chess history that comes to life
  • an honest GM opinion on the latest chess books
  • exclusive columns by Judit Polgar, Nigel Short, Hans Ree and others
  • the digital edition is available for iPad or iPhone, for Android tablet or phone and for Windows desktop and notebook (Windows 8 and 10)

The magazine has regular contributions from all the world’s best players: Magnus Carlsen, Wesley So, Fabiano Caruana, Anish Giri, Hikaru Nakamura, Vladimir Kramnik, Levon Aronian, Vishy Anand, Maxime Vachier-Lagrave, Sergey Karjakin, Veselin Topalov, Alexander Grischuk, Judit Polgar and Hou Yifan. Editor-in-chief is Dirk Jan ten Geuzendam.

New In Chess (NIC) was founded in 1984 and appears eight times a year. It is read by club players in 116 countries. A yearly subscription for eight issues costs €79.99.
Discussion and Feedback Join the public discussion or submit your feedback to the editors


Rules for reader comments


Not registered yet? Register

acepoint acepoint 11/26/2018 08:12
»Does that mean that Magnus only needs to tie the match, split the four rapid games and finish Fabiano off in the two-game blitz playoff?«

Isn't it a maximum of 5 x 2-games blitz?
JamesColeman JamesColeman 11/25/2018 11:30
@Keshava - absolutely, I actually meant the author of the article; he gives the distinct impression that just drawing two blitz games would send the match into Armageddon, which it would not.
Keshava Keshava 11/25/2018 09:31
@JamesColeman, of course you are correct. I was making a theoretical point - few people believe that a tiebreaker, should it occur - would get to the ridiculous Armageddon scenario.
dumkof dumkof 11/23/2018 11:27
I suggest head and tail, as a tiebreaker. It's a simple, fast and result oriented way to ruin classical chess.
No need to lose time with blitz and Armageddon games.
JamesColeman JamesColeman 11/22/2018 11:32
All valid points for Fabi’s chances but I get the impression you don’t realise there are potentially 10 blitz games scheduled, (5 pairs of games) so extremely unlikely to reach an Armageddon scenario.
Keshava Keshava 11/22/2018 08:14

Your idea of the World Champion having draw odds is one that has merit - that is why it has been done in the past. However, when this was in place the challenger at least had 24 games to try to wrest the crown away from the champion. The top players don't want that anymore because they have a more full calendar. However, perhaps if they had a few less rest days (like one every 3 - 4 days) - they could schedule some more classical games (still less than 24) in the same number of weeks. That would give the challenger more of an opportunity to try to win against a champion that had draw odds.

I agree with lajosarpad's answer to your question.
lajosarpad lajosarpad 11/22/2018 07:44

Interesting thoughts. I agree with you that determining who the world champion is should be done using only classical games. Playing safely and waiting for the opponent to blunder is a fairly safe approach for both players, unless the outcome becomes unacceptable for a player in the case of a tie. In our specific case the players are less terrified of playing the playoff than to take risks in the games. Who should have the draw odds? I say the world champion, since he earned the title before. You say the winner of the candidates and I think the reason behind this idea is that the winner of the candidates has shown an outstanding performance very recently. A third option would be to let the higher rated have the draw odds. As we can see, there are lots of options. About the main point we agree, however, on the detail of who should have the draw odds we disagee. Nonetheless your idea makes sense and I think it would be an improvement in comparison to the current system. Drawing odds makes sure the match is alive, somebody has to prove something.


I think you misunderstood Keshava. He/she did not propose to have a potentially infinite event. The proposal was to get rid of the tiebreaks, invite the world champion to the candidates and the winner of the candidates would have draw odds. The players at the final would be the world champion and the challenger. The challenger would be the best performing player at the candidates, besides the world champion, of course.


I agree.


The system should determine who the best player is. In the case of a tournament, there are issues like player A playing very well against player B and not as well against player C. I think the world title should be determined in a match. The current match is very interesting in my opinion, but I would like to avoid a rapid or blitz tiebreak.
fgkdjlkag fgkdjlkag 11/22/2018 05:30
Of course another "solution" is to give the champion draw odds, but how satisfying is it that he could only demonstrate being equal to the challenger?
fgkdjlkag fgkdjlkag 11/22/2018 05:29
JackCrabb has identified the kernel, which seems to go unnoticed - the high frequency of draws. The first reason is not correct, the players have taken risks in the 9 games thus far, to take more risks is to play objectively weaker moves. The second reason is partially correct, but it is not that middlegame positions are drawish, rather the quality of the opening moves by both sides is so high that it leads to relatively even middlegames. But there is a bigger reason for the draw frequency - there is a 3 point elo difference between the players. That is combined with extremely high opening preparation. This combination of facts is why there is no easy solution. To use blitz/rapid is not agreeable to all because there are separate blitz and rapid world championships. To extend the match does not fundamentally change the high draw frequency. The only solution to the current method is to fundamentally change the rules of chess.

@petaQ, I am not sure what your proposal is trying to solve; the players now are the undisputed #1 and #2 players, so it suggests that there is not a problem with the qualification process, although I have proposed to return to an interzonal-like system with geographic qualification to make chess more popular and accessible to everyone.
nbeqo nbeqo 11/22/2018 05:13
If all 12 games end in draws then i's about time to reduce the Classical time control to 1 hour per player for the entire game.
jsaldea12 jsaldea12 11/22/2018 12:20
Let me correcr myself: How about this modified playoff after 12 classical drawn:
(a) 30 minutes each player, two games, If drawn,
(b) 20 minutes each player, two games. If drawn
(c) 10 minutes each player, two games.Only if drawnc will Carlsen claims outright the crown.
Ajeeb007 Ajeeb007 11/21/2018 11:46
I agree that blitz and classical are two different games and a classical match should be decided with classical games. Also, deciding a match for the world championship by playing blitz or rapid just devalues chess and the title. However, given that FIDE doesn't really care about how trivial chess is made to appear to the general public why don't they just resort to their 1983 Smyslov-Hubner tie-break procedure and settle the match at the roulette table? It too is a different game and one in which neither player is known to excel above the other so it's a a more equitable way to settle it. Sorry Maxim but a 172 point difference is definitely significant in any context.
petaQ petaQ 11/21/2018 11:00
Why not continue with a matchplay belt, in Classical, Rapid, Blitz, whatever formats are deemed worthy and more consistent/popular than the old Chess960 WC match at Mainz, in a similar vein to boxing. Some simple purse put up by FIDE for a match every 1.5-2 years as we have now, you could even have wide elo or world ranking based divisions so that all the strong 2600+ monsters have a prestigious title to fight for without being swamped as they would in matches against the supreme.

Change the WCC to reflect the strength in depth of the modern game. If you want to enfranchise the elite, take the top 45 (based on current 2700+ strength players), and have three interzonals, with 2/5/8 etc in the first, 3/6/9 etc in the second, 4/7/10 etc in the third, some way to equalise the strengths across all three while separating the top players. Keep the WC out for now. 15 player concurrent all play alls, could be held every third year instead of Wijk or some other tournament, in that same place, to keep running costs down. Top two from each go through automatically, you could add the highest scoring third place player too to keep the drawmeisters on their toes. They go to the Candidates tournament, 8 player double round robin as now, with the WC. If you want to privilege him on the basis of his title, give him 0.5 headstart, and make number of wins the first tiebreaker. Take the top 4 from that and hold a second stage, final tournament, 4/5 round robin, which is still only 12-15 rounds. Every elite player gets their chance like in the old zonals, but there's no punishing 3 year long cycle, it fits within a supertournament structure, the WC has to earn his place in the final, and with multiple rounds between elite players, you should see plenty of victories as well as having a very well deserved champion whose participation next cycle is assured.

tomohawk tomohawk 11/21/2018 10:05
Something got truncated in the commented game. http://www.chessgames.com/perl/chessgame?gid=1879490 it evidently went for another dozen or so moves.
JackCrabb JackCrabb 11/21/2018 04:30
Indeed, a tie-break seems quite likely. However, I'm pretty sure that for two reasons most people in the chess community dislike the idea of a QP / blitz decision.
1) we don't need that, as we already know that at present, Carlsen is by far the world's best QP and blitz player.
2) QP and blitz are categories different from proper games - just like marathon on the one and 100/400m sprint on the other side.
I think Keshava is right to demand that "classical chess should be decided by classical chess games." On the other hand, Justjeff has a point when stating that we don't live in Soviet times anymore: nowadays you can't book a suitable venue for an undefined period of time. Moreover, today the contestants' diaries are much more packed than in the Kasparov-Karpov era.
Maybe it helps to have a look at the root of the problem: the high probability of draws. As I see it, there are mainly two reasons:
1) in WCH matches, at least in this one, the players are driven by the fear that a single loss would probably decide the match, therefore they avoid to take risks.
2) because of long, meticulous home preparation the players don't start thinking before drawish middle-game positions are reached.
I'm afraid we can do little against 1), as suitable measures such as a draw fine (e.g. each draw reduces both players' prize-money for, say, $10,000) wouldn't probably get a majority in the deciding bodies.
But problem 2) could be overcome – by scrapping home preparation in the tie-break ! My suggestion: after a 6:6 in the regular games, the match won't be decided by QP and blitz, but by two proper games, and if they end 1:1, another two, a.s.o. The point: in these tie-break games, the players won't have free choice of openings. Instead, before the match, both camps together with the referees define a draw bowl of, say, 12-16 openings, and immediately before a tie-break game, the opening to be played will be drawn. Ideally, then, the opening would be unfamiliar to both sides.
Marselos Marselos 11/21/2018 03:26
After 24 games, 10 games (1 hour per player) could be the solution.
I agree: 12 games are few.
treetown treetown 11/21/2018 03:20
If it comes down to a rapid/blitz off for the championship, then the title match should be held more frequently. That is one good thing about the world knockout approach being used for the women right now. It builds to a final set of games. Not ideal and not like the great matches of the past 1948-1985(?) but in the computer era, that may be where we are heading.
Justjeff Justjeff 11/21/2018 01:01
@Keshava is the venue secured for an undetermined amount of time? Are the closing ceremony folks okay with sticking around to the end? Volunteers and paid support staff? There are some good reasons for having it not go on forever. The first Kasparov vs. Karpov match had to switch venue and could only go on for as long as it did because it enjoyed state sponsorship.
Keshava Keshava 11/21/2018 12:47
What if the World Championship is decided by Armageddon? That is much worse! Better to let them play blitz until one of them errs from exhaustion. It would be ridiculous for Carlsen to lose his title because he could only draw an Armageddon game with White! The solution? Invite (not require) the champ to play in the Candidates. The winner of the Candidates has draw odds in the World Championship match. That way the classical World Championship is decided entirely by Classical Chess Games. We already have a Blitz World Champion - classical chess should be decided by classical chess games!