Time for a late first update! It has been ages since Clan League began, but still, few games have finished. Only 34 out of 231 have been decided at the time of writing. Yet, even this small sample tells a lot about the state of certain clans. Let’s take a look at the current scores first:
Clearly, two clans are in the lead: Discovery
and French Community
. Both suffered two losses and have the same PC%. Only, due to the French’s penalty at the start of the season, Discovery actually has two more wins. Discovery mostly scored against 101st, while French Community had a more spread out pattern of victories.
Blitz
is getting close, hunting the two leaders from the third place. The gap of 11 PC% might appear enormous but can be closed with a single game at this stage. Now they have positive head-to-heads against every clan but the two leaders, and if they want to promote, they have to start grabbing more points there.
The Good Gamers
and M’Hunters
are way behind Blitz, almost 30 PC%. Both picked up one 3v3 and three 1v1s to get an equal amount of points, but GG has one more loss. Promotion will be a long shot, and the relegation battle is never far away. Both have a long history of excellent performances in B, so it is way too early to count them out of the race.
A little behind them, at the feared 6th place, {101st}
is ready to catch up and remain in B for the nth time. Not a good start for the eagles, but a lot is still possible, although promoting is difficult at this point.
Meanwhile, CORP
is in big problems. Only 1 win in 7 games is not what they had hoped for. Get it together!
Everyone could deduce what I have stated so far from the image. So, time for some in-depth analysis. Despite the low number of finished games, I wanted to understand the current state of affairs better. Is it possible to compute who will promote already? Well, this is quite hard, and polling has not been a good indicator in past Clan Leagues. So, I opted for simulations.
Say, every game is equivalent with a coin flip, a 50/50 chance for both clans to win, and I compute this randomly for every remaining game. Then I get one possible scenario of how Division B will wrap up. If I do this a million times, I have some sensible statistics. Too lazy to worry about tie-breakers, I randomized the order in case clans tied at the end of the season. First, I wanted to assign an ELO rating to every tournament, but there is too little data for that at the moment.
Strangely, this method reveals that Blitz can actually be considered the runner-up behind Discovery at the moment. Why? Most likely because this method likes the extra points that Blitz has at the moment and that Blitz has had more h2h’s already against the other top 3 clans. Meanwhile, on the bottom, four old-timers fight against relegation.
