Time taken by players to answer 15 questions in Division 2

26.48
58.12
61.26
61.3
66.48
70.0
73.47
78.4
92.49
117.76
122.97

At the time that these results were taken there was a total of 4 divisions.
The person leading division 2 was also the leader of all 4 divisions.

There must a flaw in the the algorithm.

The algorithm uses players' average scores to determine divisions.

The person you mention is well known for playing a very unusual strategy - they play extremely fast without any regard for accuracy. This results in them getting almost guaranteed wins if they hit 15/15, but their average is low due to many 14s, 13s and lower scores.

Yes but this makes a mockery of divisions because slower players cannot get any wins.

I have the solution division by the tier in achievement word wizard :
Tier7(15/15 in 30 sec) division 5
Tier6(15/15 in 40 sec) division 4
Tier5(15/15 in 50 sec) division 3
Tier2(15/15 in 65 sec) division 2
Tier 1 or tier 0 division 1
With my solution division determined by your achievement no more
problem with the player who is forever in division 5

Gme24, it's not really a mockery. If a player plays it in under 30 seconds but only gets 15/15 one out of every 15 times, then others will win 14 of every 15 times. If you had 58.12 seconds, what's the difference between losing to someone who finished in 26.48 and losing to someone who finished in 58.11?

Philvv, the disadvantage of your idea is that it relies on only one prior play. I've had one or two times where I was 15/15 in under 40 seconds. On the other hand, most of the time that I get 15/15 (which is less than 50% of my plays), I'm somewhere in the 60-90 second range. I usually float back and forth between division 2 and 3. Basing my division solely on my best performance would mean that I'd probably win about once every 200 plays or so.

The issue is known, by the way, and we have discussed possible fixes internally in the past. However they would either be very complex, making the game and scoring hard to understand, or have side effects that would be worse than what is happening now.

I agree with gme24. The divisions are flawed because they're based on mean average. But, there's probably not an easier system to implement. The median score would get my vote if it's possible to sort the divisions. Someone scoring 15/15 in 30 on WW 7/10 times and scoring 15/15 in 30 seconds 3/10 times puts their median score at 1470 and definitely the top division but their average is 1440 which would likely sort them elsewhere. Or possibly a win % sorting system could be used. Someone winning their division 50+% of the time should not be in that division unless they're already in the top division of course.