Good luck

whocares2

King destroyed all the fun in playing this game. I enjoy a challenge, but repetititively being given 12 moves to get 400+ blockers is ridiculous. And then when you think you have used enough boosters to win, the game shifts last minute which requires using gold bars, which now requires you to use 25 gold bars instead of 10. Quickly drains the supply of gold bars and boosters you accumulated. Sweet Cinema only gives you 3 or 4 moves to win boosters before it shuts down. I am done. Good luck everyone.

[Deleted User]

I believe I can explain what you are seeing.

The game is not actually changing the board to force you to use gold to win. You have already lost the round and the game is trying to make it seem like 1-2 more moves will allow you to win. It is trying to get you to use gold but it is not "cheating" to do so.

To go into more detail. The game uses an algorithm to make the levels and make a winning set of moves. This is what you are playing against.

Basically each level has a set number of mistakes you can make before the board is considered lost. When you make a wrong move the game tries to adjust the board to allow you to win.

Once you lose the game adjusts the board to make it look like you only needed another move or two.

The algorithm does NOT build the wining strategy on what is on the board, it uses drops of new candies as well as part of the wining strategy. That is why you can make what seems like a less ideal move and have a massive chain. No human would be able to guess that if they moved this candy, you would have a 5-10 move cascade from candies that do not even exist yet. I have had on the last move a 1/4 of the board clear this way. It actually made complete sense as all the other moves seemed to be only 1 or 2 changes even when you would think it would be more.

This also I suspect is why sometimes special candies just go boom and sometimes don't. It is the algorithm adjusting in real time.

This does lead to another theory. You are not exactly trying to find a wining path, you are just trying to match the algorithm enough times and the code behind the scenes makes sure you win or lose based on that.

I forgot to add that it is actually quite annoying that is does the "few more moves" stuff. It preys on the need to complete tasks and the "well it is just x gold". Though to be fair it is basically how this kind of stuff works, think the booster wheel and getting super close to the jackpot.

christinewupp

https://community.king.com/en/candy-crush-saga/discussion/comment/2921854#Comment_2921854

Nice explanation @Pyrolight Would you agree with me that the algorithm includes help to pass the level if the player uses a certain booster? I sometimes find that if I use a booster such as a UFO with a few moves to spare even if it does not clear the board, the candies that then drop are always the right ones to then win the level.

[Deleted User]

Yes I completely forgot about that.

Boosters that are usable while playing a level seem to increase the odds of a lucky board/give more chances. UFOs are glorious.

I think the same effect exists with Sam's Streak but far more random, but could a result of what I mention below about color bombs.

Boosters used before the game do not seem to have quite the same effect. I swear color bombs are just a worst thing to use on anything but a normal/large selection of starting candies type level. They seem to almost always mess up the algorithm. Stripped and wrapped seem to be a more mixed bag. I have almost completely stopped using color bombs for super/nightmarish/legendary levels.

Another thing I noticed recently, mostly with regards to the start of a level, is that some moves produce a specific candy result every time, but if you do the mirror the result is different. This would be where the candies would have to drop onto the map. I forget what level I tested it on recently but I was almost always able to produce the exact same result. Maybe one time in seven or eight it would be different.

On a side note I actually has the same "setup" happen twice and when I tried something different the results were dramatic. So in theory you actually can loop over the same algorithm result eventually.