The Community in more languages
Now the forum welcomes more languages.
You'll see a green translate button in comments and discussions to turn them into English
🔥 Hot right Now! 🌶
🏆 Claim your level milestone badges:
1000 // 2000 // 3000 // 4000 // 5000 // 6000 // 7000
👯♀️Find your Team HERE!
1000 // 2000 // 3000 // 4000 // 5000 // 6000 // 7000
👯♀️Find your Team HERE!
Systematic quirks in the “recommended move” algorithm?
Uncle_Bob
Posts: 162 Level 3
in Discussions
What kinds of systematic quirks do you find in the CCFS algorithm that recommends your next move by highlighting a set of candies to match?
My observation is that algorithm seems to work top-down on the playing grid, and it seems to prioritize making a wrapped candy over making a striped candy over making a fish over matching 3 candies of the same color.
But the algorithm seems to miss some nuanced patterns of candies , and scanning for those helps me me find “better” moves. An example is case today when the algorithm was recommending making a cyan striped candy high in the playing grid, but a much better move was matching 3 yellow candies to “delete the middle” of a column of orange candies, which both exploded one striped candy and simultaneously created another striped Orange candy in its place:
More generally, the recommended move algorithm does not seem to include all of the direct and indirect effects of each possibly move, so I think if we can train ourselves to see those possibilities more clearly and accurately, we can select more effective moves, particularly at the end stages of game when things are complicated and more moves are possible.
Any insights, anyone?
Comments
-
Another systematic quirk in the “suggested move” algorithm is that it typically does not recommend combining a Color Bomb (or Coloring Candy) with fish, striped candies, or wrapped candies. Since the Color Bomb applies the fish, striped candy, or wrapped candy to ALL candies of that color in the playing grid, the total destructive effect is usually much better when using the Color Bomb or Coloring Candy.
An example with the algorithm recommending combining 2 wrapped candies rather than combining the Color Bomb with the yellow wrapped candy is below. Combing 2 wrapped candies gives centralized destruction, but combining the yellow wrapped candy creates and explodes 7 yellow wrapped candies and gives far more widespread destruction.
And an example of the algorithm recommending combine a wrapped and striped candies. BUT, combining the Color Bomb with the yellow striped candy—which then creates and explodes 13 striped candies—gives much more total destruction.
Given this quirk, when you have a Color Bomb or Coloring Candy, you should carefully check if it can be combined with a fish, striped, or wrapped candy for a more effective next move!
Happy crushing!
-
Today the “suggested move” algorithm recommended that I combine the Color Bomb with a red fish, which would just have triggered about 4-5 fish, rather than the Color Bomb + yellow striped candy, which created and then exploded 9 yellow striped candies.
So when you have a Color Bomb or Coloring Candy, check out other possible before going ahead with the suggested move. I often check to see if I could create a fish, striped or wrapped candy and get it to the Color Bomb to increase its destructive effects. Those destructive effects are even more with a Coloring Candy, of course, as you get duplication for at least two colors on the playing grid.
-
@Uncle_Bob You want too much from the game)))). The game does not think about the consequences of the move, it only points you to the standard combination of candies to create a standard booster or a combination of boosters. The game does not think, it is not trained in this.
As for the colored bombs and the examples you gave.
You like to analyze the game deeply. Do your research. Collect good statistics on which specific combinations of candies on the playing field are always offered in priority over other specific combinations. (For example, as you wrote above - "combining 2 wrapped candies, not combining a colored bomb with a yellow candy").
If you find a pattern, it will be interesting to look at it.
I would not be surprised if you find this pattern.
And I would not be surprised if it turns out that the game has a list of candy combinations (the usual list written in the game code, a list without priorities) and the game selects from the list the first available combination of candies that are located on the playing field.
And if this list is written in a different order, then the tips will change)))))))
-
@ElenaVorona I really am just trying to caution players to learn to look more closely at the playing grid for possible better moves than the move suggested by the algorithm, with the goal of players improving their skill at CCFS. Clearly the algorithm has to be kept relatively simple to search the grid and give suggested moved in a very short tine no matter what device CCFS is being played on.
However, I also think you should always wait and see what the algorithm’s suggested move is, before deciding what to do, because sometimes the algorithm can perceive things that you may miss when you scan the grid. An example is when some game pieces are obscured under that orange-yellow film as in the example below:
It is easy to miss seeing the topmost red candy under that film, but the algorithm knew it was there and suggested making a wrapped candy by moving the lower red candy up to make the “L” pattern. This is the type of circumstance where the algorithm may catch a possibility that you do NOT see, so it is always wise to first check what is suggested and the see if you can find something that improves on it. In particular I think the algorithm is just scanning for the “most valuable candy” that can be made in 1 move , and does not do any “look ahead” to see what a sequence of two or three moves could do (which would require a LOT of computation!).
I have indeed kept statistics on some things, but I think that would bore most people.
Happy crushing!
-
This "quirk" can be corrected if we pay attention. 🤔
It bothers me more when I expect a Yeti's wrapped candy to drop on an available cyan candy beside a color bomb/coloring candy when it prefers to land near a fish instead! Unbelievable.
How am I supposed to destroy cups with just one wrapped fish.
And it doesn't seem random either. Since quite a lot of CB/CC got overlooked, I now make sure to isolate/kill every dangerous fish before activating Yeti 😤
-
@Pitty_Kitty I have had exactly the same problem as Yeti is programmed to throw the wrapped candy beside a fish, beside a striped candy, and beside another wrapped candy ALL before Yeti will throw the wrapped candy to be beside a Color Bomb! So just like you, I have to try to destroy all those other options, and then trigger Yeti to throw it beside the Color Bomb. If that’s not possible, I have also tried to work the wrapped candy down to the Color Bomb, or work the Color Bomb down to the wrapped candy, but that is VERY ticklish and difficult work. The problem with correcting for this “quirk” is that we lose turns doing so, and thus risk losing because we run out of turns.
Tifi and Odus, in contrast to Yeti, seemed to be programmed so that they will throw a fish/striped candy right next to the Color Bomb if the correct color candy is present there. Has that also been your experience, Pitty_Kitty? What other quirks have you experienced? It would be good to have a list somewhere for new players to consult and know about.
-
I have noticed like @Pitty_Kitty that the “suggested move” algorithm does NOT handle Color Bombs or Coloring Candies very well. I captured 2 examples of that this morning. The first was a screen where the Suggested Move was to combine the purple Coloring Candy with a Blue fish, which would have created 13 fish. But combining the Coloring Candies with the Green Striped Candy to its right created and exploded about 15 striped candies and caused far more destruction!
In my second example, the Purple Coloring Candy was completely ignored—the suggested move was to combine the Purple Fish + Blue Fish, which would just have created 3 fish. The far, far better move was to combine the Purple Coloring Candy + Purple Striped Candy directly above it, which created and exploded about 20 total striped candies for massive destruction.
So the take-home lesson is that particularly when you get a Color Bomb or Coloring Candy in CCFS, be very vigilant to stop and examine the playing grid for other opportunities to effectively USE that CB or CC to help win that level! In particular, you might want to carefully count how many of fish, striped, or wrapped candies you would produce by combining the CB or CC with a fish/striped/wrapped candy beside it, and what how much total destruction would result from each choice. Happy Crushing!
-
I'm so sorry I've missed this mention 😅 yes, Odus and Tiffi don't cause the same level of damage as Yeti, they have 3 and 2 special candies to drop so they have more chances not to overlook a color bomb/coloring candy like yeti. I don't think I've ever seen them miss on a good opportunity.
About the coloring candy issue. Yes, the game forgets about its existence a lot, I see it regurlarly too 🤔 weird
Another thing I noticed is about Misty. in some occasions I try to have an orange candy near a striped candy or better yet a wrapped candy on the board because I expect to make a striped/wrapped combo so I can cause a big destruction. sometimes it happen. Sometimes Misty drops 4 random and very much less effective fish, and I'm like, Misty, are you here to help me or what? 😅 so random. so disappointing 🤣
-
@Grumpy_Old_Man What would you recommend/suggest to help newer players learn to evaluate the grid and find more effective moves?
I was thinking of something like:
“The algorithm works top-down and stops at the first acceptable match in the playing grid, so you should carefully examine the playing grid from the bottom up to find good moves lower down.”
“The algorithm prioritizes moves that would make a wrapped candy over making a striped candy over making a fish over a simple match of 3 candies of the same color. But you should carefully judge whether a lower-priority move would actually advance you toward your goal better than the algorithm’s suggested move.”
If you have any ideas (or better ways of wording the above guidance!), I’d appreciate it.
