Welcome to the Candy Crush Saga Community
Do you have a question or need help with your game? Ask the Community.
๐ถ๏ธ๐ฌ Hot & Sweet right now in Candy Crush Saga!
What's happening in Candy? ๐ฌ
๏ปฟ โ Candy Holiday Season โ - More info HERE
๏ปฟ ๐ Winter Cup - Win a yearโs supply of Gold Bars (100 every month) - More info HERE
๏ปฟ โ Candy Holiday Season โ - More info HERE
๏ปฟ ๐ Winter Cup - Win a yearโs supply of Gold Bars (100 every month) - More info HERE
New Year - New device? ๐ฑ Check out how to sync it! Android ๐ค | Apple ๐
I think the community is broken, what I hope to learn from leaderboards
christinewupp
Posts: 10,033 
The only way I seem to be able to communicate is by editing my existing post. I got a fright at first as all our posts had disappeared. I thought my account got suspended but now I think that it is that the community is down as they still allow me to edit my post. All I see is the original posts and the original questions in Support but none of the answers. Hope it gets fixed soon!
This is a change from last year: We can now see every player's player card and it's full of information, which seems to be more reliable now. I have two games left playing in the Winter Cup and just completed two games in the Gold Club event, which seems to be for those players who failed to make it in the Winter Cup. That's a total of 150 gold club players and 90 Winter Cup players. So much data to crunch through!
The questions I'd like to find out about are:
- Are some players at an advantage over others because they are playing at better level ranges?
- Is it safe and advisable to join a leaderboard late?
- Are players grouped completely randomly?
- How many new levels do we have to play to win?
- What is a plausible and achievable score?
Anyone who can help answer some of these questions and share an end of round leaderboard would be very helpful. I have noted down the starting and finishing levels of each player in my groups, so I can work out how many levels they played and from this work out the average score per level. I'm hoping this may help to answer Question no. 1.
Of course it's all preparation for All Stars, both to dispel the myths and to try find the best strategy to compete.
There may be some surprises.
5
Comments
-
Is it safe and advisable to join a leaderboard late?
Here's my take on this.
Regardless of the possible changes made to how leaderboard events are run, joining late, especially very late, with only hours left on the clock, is not safe. If you join very late, it's highly likely that you'll be grouped with at least one astute and fast player who deliberately uses this strategy. That can be a dangerous encounter, and you might not have enough time left on the clock to deal with the situation. Joining a bit late (one day late in the 3-day round) is a safer approach and generally beneficial as it at the very least allows to avoid being grouped with very active players who crush 24/7.
Another pitfall is that in the 'unlimited leaderboard' round, all players in the group are supposed to start at nearly the same time. This is only true when there are many other players starting the round at the same time as you. When you join too late, particularly in the later stages when there are few participants left in the tournament, you get into a situation where the group fills very slowly. As a result, it's possible to be grouped with players who started the round hours earlier (based on my own experience in All Stars).
I don't think joining late is advisable for everyone either, as it largely depends on the preferred playing style. This is for those who would rather play hard and fast for a shorter period of time than play at a more comfortable pace, but play more, and for a longer time.
In my opinion, the most important aspect of the late entry strategy is deciding when to start in a specific round of a specific event, taking into account the round length.
4 -
Thank you for that precious and well informed advice @MannyFae . I totally agree with everything you state. My own experience in the 2nd knockouts that just finished is this:
The group I joined later (I remember you once advised that Thursday evenings might be a good time, so that's what I did, roughly halfway through the event) I managed to easily win and a score of 409,029 would have been enough to get to the final.
In the game I started within the first few hours I did not actively play, but watched. Here is that final leaderboard, full of enthusiasts:
In this particular event three players got to proceed, but in All Stars of course, at this stage, only the top player will go through, so there is always a risk of failure if you are up against just one other serious top player (That Panda played 485 levels during the 2nd knockout round, and 756 levels in the past seven days)
How did you get on and do you happen to have any data on player levels? I have gathered some and will share in my next post.
2 -
This is not exactly answering any of my questions yet, but as I also had the Gold Club data I decided to use both sets to try to see what levels players are at. Both my games that I failed to qualify for the 2nd knockout round ended up in the Gold Club, so a combination of both events might be fairly representative of overall numbers of active players. I say "might" because this would assume that King assign players at random to these events, which we know is certainly not true with regards to Windows players at the top of the game.
In the 2nd knockouts these are my figures (n=88, two games excluding myself)
Levels 1 to 1000: 20%
Levels 1000-4000: 26%
Levels 4000-10,000: 42.5%
Levels 10,000 + 11.5%
In the Gold Club leaderboard these are my figures: (n=148 in two games, excluding myself)
Levels 1 to 1000: 28%
Levels 1000-4000: 27%
Levels 4000 - 10,000: 31%
Levels 10,000 + : 14%
I have more broken down data, but last time I tried to post too many figures the community bots would not allow my post, so I have become wary of posting too many numbers all at once.
There seems to be an overrepresentation of the middle ranges in the 2nd knockout leaderboards compared to the gold club. I think this may genuinely be a result of those players generally doing better in the competition, but I need to analyse it some more before coming to that conclusion.
2 -
So in summary (as the bots seem to allow my posts) in the combined events of both gold club and 2nd knockouts these are the starting levels of all 236 players on my four leaderboards, which may be representative of the active mobile players King have assigned to either event this past week:
level 1-1000: 60
level 1000-2000: 20
level 2000-3000: 23
level 3000-4000: 20
level 4000-5000: 15
level 5000-6000: 24
level 6000-7000: 18
level 7000-8000: 10
level 8000-9000: 10
level 9000-10,000: 6
level 10,000-11,000: 5
level 11,000-12,000: 3
level 12,000-13,000: 4
level 13,000-14,000: 1
level 14,000-15,000: 4
level 15,000-16,000: 2
level 16,000-17,000: 7
level 17,000+ : 4
2 -
This is my board. I joined with 36 hours left on the clock. I was 'lucky' to be grouped with two (!) opponents who kept a low profile until the last 10-12 hours when they both attacked. It wasn't really necessary to push for victory, and settling for 3rd place with a score of around 690k would certainly be a reasonable thing to do. But it was great fun playing against them.
I took screenshots of all players cards both at the beginning and at the end of the round, and of all the final scores as well.
2 -
Interesting. It looks like I got a much luckier group. None of the others attacked late. The eventual second placed player was the only one who was playing during the night (night in Europe that is, they were from Mexico) and for some reason stopped just short of catching me. I then played one episode in the final hour, just to be quite sure but no one else was playing this morning at all.
I joined the event with about 38 hours left on the clock, at about 5pm UK time. As it is a test game I will join the final only to observe, not compete. Good luck in the final @MannyFae !
2 -
Do I need to process the data related to the start or the end of the round?
0 -
The level I used was the level they started at @MannyFae
0 -
Here's my data for 2nd Knockouts:
level 1-1000: 8
level 1000-2000: 9
level 2000-3000: 2
level 3000-4000: 4
level 4000-5000: 2
level 5000-6000: 4
level 6000-7000: 4
level 7000-8000: 6
level 8000-9000: 1
level 9000-10,000: 0
level 10,000-11,000: 0
level 11,000-12,000: 2
level 12,000-13,000: 0
level 13,000-14,000: 2
level 14,000-15,000: 0
level 15,000-16,000: 0
level 16,000-17,000: 0
level 17,000+ : 0
Here's a diagram of my data as well. Too little data for proper analysis, but it seems like there might be a bump around the middle ranges.
There seems to be an overrepresentation of the middle ranges in the 2nd knockout leaderboards compared to the gold club. I think this may genuinely be a result of those players generally doing better in the competition, but I need to analyse it some more before coming to that conclusion.
II tend to think this is true. I have a theory that in a multi-round tournament, the number of lower-range players decreases as the tournament progresses through the stages, and I expect there to be fewer of them in the final round tomorrow. I will probably only join to observe. The Winter Cup is a hard-fought tournament due to the big prize, and I don't feel like playing hard for two days. I had a final of sorts in the 2nd Knockouts ๐
2 -
Thank you so much for that data and the great graph. I love it!
I am at the moment looking at "levels played" vs "points gained" to work out an average score per level for each of the level ranges. This is a laborious task and you don't have to join in with this, but it would be interesting to see if you find the same trend I'm noting: It's definitely looking like all the players in the lower, recently rewritten and doctored ranges are scoring very similar average per level scores across the board and it's a lot lower than I expected and remember from last year's count. It also looks strangely beneficial to be playing in the 6000-9000 range. Those levels seem to still yield average level scores 2000 points higher than those of the early levels. I grant that this may be due to players turning to older go-to levels to top up their scores, which would mess up the calculations, but this seems to be across the board for all players, and I can't imagine they are all playing lots of older levels during a competition like this. I wonder if there are still more and better cascades in those levels King have not yet reviewed.
1
