How about a logarithmic scale for the y-axis?
3 Likes
16 days later (data of 11/12/2025 from Purble):
- 273 games in progress, 291 players involved, 67 groups
- 21 completed games
- 20 players completed all their games, 4 of them with 9 wins
- 3 groups completed all their games
- 2 disqualified players
Here you are:
6 Likes
From a rough eyeball estimate, it seems that games have a half-life around 100 days
3 Likes
The radioactive decay comes back.
We already tried that to estimate the end of the first round
3 Likes
Looks like the exponential decay worked pretty well down to ~20 games last time, which is pretty reasonable. Too early to tell if it’ll work as well this time, but we’ll see
2 Likes
Yes. the first round had a slope on the log graph fairly constant for the first 300 to 400 days, flattening out slightly until around 1000 to 1100 days, at which point it clearly steepens. Cleary the second round initial days were dominated by mass timeouts. but does seem to be also flattening some now. I could theorize the final “acceleration” could be due to endgame generally played faster, and final games having some psychological pressure (the needs of the many outweighing the needs of the few or one)
1 Like
For those of you who are fans of powers of two, there are now fewer than 256 games remaining to complete in round 2… so for half life estimate, still 800 days to go
2 Likes
19 days later (data of 12/1/2025 from Purble):
- 253 games in progress, 274 players involved, 64 groups
- 20 completed games
- 17 players completed all their games, 1 of them with 9 wins
- 3 groups completed all their games
- 1 disqualified player, 1 resigned
7 Likes
After the Maya calendar, the Gregorian calendar, I propose we get the Lys calendar.
6 Likes
Hopefully will end (and win 
) my last ongoing game before those 800 days. Already impressed by being in the last 5%.
3 Likes
Thanks @Lys - because I’d be interested to see how fast games get finished in this 2nd round compared to the 1st round. … And if it’s not asking too much, both a “games completed (in any manner” vs. time and a breakdown comparison of how many finished by 1) “full play, passing, and counting”, 2) by a player resigning, and 3) by “time out” - that would be interesting to me.
3 Likes
14 days later (data of 12/15/2025 from Purble):
- 241 games in progress, 261 players involved, 62 groups
- 12 completed games
- 13 players completed all their games, 1 of them with 9 wins
- 2 groups completed all their games
- 1 disqualified player, 1 resigned
@Sighris I’m looking into your requests
5 Likes
How do I know if I was disqualified from the tournament please?
By going to the Tournament page and searching for your nickname.
Happy new year. we were down to 220 active games on the 1st.
4 Likes
22 days later (data of 1/6/2026 from Purble):
- 212 games in progress, 238 players involved, 58 groups
- 29 completed games
- 23 players completed all their games, 3 of them with 9 wins
- 4 groups completed all their games
- 4 disqualified players, 2 resigned
Now about @Sighris topic, I tried to brew some meaningful charts.
The two rounds have a different overall number of games, so I decided to use percentages in order to have meaningful comparisons: the number presented in the following charts are all percentages of the total number of games in each round.
First round had 10038 games.
Second round has 6246.
The first chart shows completed games by their duration (days).
The first round had a spike on day 3. It was a wave if timeouts, affecting the 5% of the total games.
Second round had a huge wave of timeouts around day 40, with a peak on day 42 (!) of about 29% of the total.
These spikes are so big that they squeeze all the rest near the x axis.
Next chart splits the above data respect to outcome into three groups: games ended by resignation, by timeout and by scoring.
We can see a spike of resignations in the first and second day.
About timeouts, there’s a significant difference, since round 1 had a bunch of them on day 3 and then a someway regular tail from day 40. Second round skipped day 3 and jumped right to day 40, with a much higher rate of players timing out.
I have some zoom-in of the above chart.
Resignation:
Timeout:
Score:
It’s quite interesting to look at the running sum of completed games.
Next chart shows all completed games adding up to 100% against days.
After 250 days round 2 has 90% of finished games. Round 1 was slightly lower.
A zoom-in on the first 365 days shows that the second round was having a lower rate of completing games until day 40. Then the timeout tsunami happened and brought the completed games ratio from 10% up to 58%.
It seems to me that the second round is still progressing slower than the first one. I wonder if it’ll take more time to finish.
The last chart shows the above data splitted by outcome, as previously seen.
Of course the final ratios in the two rounds will be very different, because of the huge number of timeouts in the second one.
9 Likes
From the zoomed in graphs it seems like the last ten games took quite a bit of the time of the tail. I wonder if we have any reason to think the second round will be much different, and there isn’t a lot of data anyway to have much faith in a statistical analysis. At least for round 2 I don’t think there are any games still running with fewer than 50 moves made . (the last one I knew of that was under 50 moves ended by resignation ). I suspect round two will also require patience as the last handful of games move to completion.
1 Like
For those of you who like multiples of 10, today we have reached only 200 active games remaining in round 2.
3 Likes
Two weeks later (data of 1/20/2026 from Purble):
- 198 games in progress, 224 players involved, 55 groups
- 14 completed games
- 14 players completed all their games
- 3 groups completed all their games
- 2 disqualified players
Ten more days have passed since the data was collected, so these stats aren’t up to date, but I think it’s good to keep the series going.
6 Likes
