Level 3178 mechanics: Don't play boosters! (and other stats)

christinewupp

I had the pleasure to start level 3178 in my latest account today. This level is rated "hard". I was on my top win streak and saw I had two side by side colour bombs on my starting board and they self destructed in front of my eyes! I have not experienced this before. Since then I have extensively played and replayed this level as it has the easiest mechanics of any level I have ever come across. It is a perfect level for testing out certain theories…

This is the level. 10 moves only. The candies drop down onto the starting board, which is weird and unusual. There is a colour bomb and a wrapped green candy, also weird and unusual.

There are only two outcomes to this level: If the colour bomb sits next to the wrap you will definitely pass. If they do not sit next to one another you will definitely fail.

So simple, such a beauty. @Pyrolight Have you come across this one yet in any of your games? This is definitely one to test out certain theories on!

christinewupp

I ran some tests playing and replaying this level.

On my iPad scrolling back after reloading the game I passed 50% of times with no boosters. Once I had lost my 5 lives the fishing boat appeared with timed wrap/striped boosters. After this I failed on the next six attempts, because the boosters interfered with the drop of the candies and exploded the wrap or CB, or because the colour bomb did not sit next to the wrap. My pass rate with the timed booster was 20%, and I ended with another chunk of five consecutive fails as I used up my last lives from friends.

On the iPhone I scrolled back and found that without boosters I passed 80% of the time. On the iPhone I had timed colour bombs from the fishing boat. With the timed boosters my pass rate dropped to 40% initially, but then I had 13 consecutive fails (100%) until I ran out of lives. I have often wondered if there is a "the more you fail the more keep failing" part of the algorithm.

I know this is an old level but King probably have kept it in the game because the outcome is so easy for them to control. Too easy in fact: I will certainly come back to this level and test it again in the future. Does anyone else feel bored enough to scroll back and have a go at this one?

Pyroyeti

https://community.king.com/en/candy-crush-saga/discussion/462250/set-up-to-lose-your-win-streak-how-boosters-will-make-you-fail

No not got to 3K yet.

I did once see a initial cascade due to special candies triggering immediately. It cleared half the level and it was just the once so assumed it was just a bug.

Also ya some have a specific first turn/two turns mechanic and boosters will completely break those. Also I found if you don't trigger the specific mechanic the level is far harder.

MannyFae

The starting board before the candies settle looks like this:

The top three rows form a wrapped candy, while the fourth row forms a color bomb. Where the game decides to place them (which can be anywhere in their respective creation area) determines the outcome. For random placement, the probability can be accurately calculated I think.

koimoko

Better not using a booster

christinewupp

https://community.king.com/en/candy-crush-saga/discussion/comment/3125890#Comment_3125890

Now that you show this screenshot I can see how the level works, thanks. It happened so quickly I missed the initial board.

I did my homework last night and played this level 100 times on my iPhone 6 without boosters (I had to change the time setting to get rid of the boosters when the fishing boat showed up)

The pass-rate was 50%.

Colour bombs formed in columns 1 or 5 (so on the edges) 43%

Colour bombs formed in columns 2 or 4 29%

Colour bombs formed in column. 3 (the middle). 28%

Wrapped candies formed in the middle, (which seems the most likely place judging from the way they form?) 72% and in either columns 2 or 4 28%.

I missed some data because of automatic explosions I could not follow. (If more green candies drop from the top this can happen)

Do you know what the algorithms are for the formation of colour bombs and wrapped candies in general @MannyFae ? Should we expect colour bombs to form 20% of the time in any of the five spaces? If that is so then what I found is not random. Or would they normally form more likely in the middle? Again, then what I found is not random.

Wraps presumably would normally form in the middle, so their formation might be random? (75% middle, 25% either side?) I would assume that the CB would form first. Quite often the wrap sits on top of it in column 3.

If I get the time I will try again today and see what I get.

Of course none of this matters with regard to the original header of my thread: Do not play boosters for this level because you are more likely to fail using boosters.

I will change the header of the thread so as to not go off topic with my calculations and tests.

MannyFae

https://community.king.com/en/candy-crush-saga/discussion/comment/3126270#Comment_3126270

Wow, thanks. 100 attempts is a large enough number to draw conclusions. I've always noticed that if a color bomb is created not directly by making a match, but by a candy falling into place (i.e. cascades), a color bomb can be created in any of the five spaces. I don't know if the probability is supposed to be 20% for either space on other levels. For this particular level it's clearly not. It's not necessarily intentional manipulation, it's possible that the spawning probabilities are just set to specific numbers (hypothesis: it is 20/15/30/15/20 for spaces 1-5).

I believe that the probability of wrapped candy placement is also skewed. I'm not sure if the wrapped candy can be formed in 1st and 2nd row (I didn't notice this happening), but even if yes, for random placement, the probability of it forming in the middle column would be 60%. A color bomb and a wrapped candy start forming at the same time, but it hardly matters.

It would be interesting to work out if the color bomb placement algorithm is the same for all levels, or different. There are more levels like this one, but unfortunately they are more complex than this one so not good for testing.

The pass-rate was 50%.

'Flip a coin' level, literally 😄