lichess.org
Donate

Analysing Armageddon Games: Time Usage, Scores and Sharpness

ChessSoftware Development
Taking a closer look at armageddon games from Norway Chess

While I was looking at the classical games from the recent Norway Chess tournament, I thought it might be interesting to take a closer look at the armageddon games. I had so many open questions, like which color scores better and if there is a way to look at the risks both sides are taking.
So in this post I’ll look at the time usage for both colors, the scores of the colors and the players, and finally if White is taking more risks than Black.

The Games

There is no unified time format for armageddon games, so it can be quite difficult to compare them. Since I wanted all the games to be played with the same time format and under similar conditions, I decided to only look at the armageddon games from the Norway Chess tournaments starting in 2021. These games are played with 10 minutes for White and 7 minutes for Black.

Time Usage

Thinking about the time situation gives rise to the question of how top grandmasters are using their time in armageddon games.
To answer this, I looked at the average time left for both sides in all armageddon games and got the following plot.
White starts with a 3-minute time advantage, but Black closes the time gap from move 1. After 20 moves, White is around 2 minutes ahead on the clock, and by move 40 the time difference is just under 1 minute. White keeps a small time advantage for the remaining moves.
So White is playing a bit slower than Black, using their extra time to think. But at the same time, White tries to always preserve a time advantage over Black, even after 60+ moves.

Scores

Another important aspect of armageddon games is the balance between the two colors. To quantify it, I calculated the scores for both colors and White scored 46.6% and Black 53.4%. So Black has a slight advantage in the armageddon games, but it’s similar to White’s advantage in classical chess.
Another interesting thing to look at is the armageddon score in comparison with the classical score of the players.
Note that not all players have played the same number of games, so some of the strong performances (e.g. the armageddon score from Anand) are just taken from a couple of games.
Amazingly, Carlsen scored over 80% in the armageddons, even though he has played in every tournament I looked at. It’s also surprising that Firouzja’s armageddon score is so low. I would have expected that he scored much better in armageddon. But in 2021 and 2022 he scored only 1 win in 7 games.

Sharpness

Finally, I wanted to see if White is taking more risks than Black in the armageddon games.
This is much harder to quantify than the time usage or the score. My attempt to quantify this is to look at the change in sharpness after each move. The idea is that when one player takes risks, it increases the sharpness in the position and so the sharpness change will be positive. On the other hand, when one side is playing safe, the sharpness will decrease and hence the sharpness change will be negative.
I decided to look at the average sharpness change for every move to see how the games developed. Also, this makes it easy to exclude long, drawn-out endings where there aren't many opportunities to make the game sharper.
It's amazing to see that there is already a big difference in the sharpness change after only 5 moves. The gap only gets bigger when the games go into the middlegame and it's clear that Black is trying to calm things down since the change in sharpness is negative for almost every move. This shows that White and Black are approaching armageddon games differently.

Conclusion

I found the analysis of the armageddon games fascinating. It was interesting to see how the players use their time and also how the colors score in armageddon. But the most striking part was the change in sharpness.
Let me know what you think of the different stats and if there is something else you’d like to take a closer at.
If you've enjoyed this post and would like to see more posts like this, check out my Substack.