Well, (in my first post to BGN) I wanted to point out that it is (probably not well-) known that Candyland is very nearly a Markov chain (and it would be if the deck is shuffled after every turn). 

I rarely recommend Wikipedia to anyone (the compendium of knowledge, some of which is occasionally true).  I only ever do so for a brief overview for any topic and the (usually more important) external links an article has. 
The external links to the Candyland article give you various mathematical analysis as to the average number of turns in a game, which is much more interesting than the game itself. 

I think in this case, regardless of quantum interactions, Candyland isn’t a very interesting game to play, and that’s a reason why it’s only played by the very young (or the very drunk).  I *would* agree that good parents will find something else to teach their children to play.

After playing Advanced Pirate Candyland (Cartagena), I just can’t see myself making it through a game of “regular” Candyland without instituting some heavy-handed new house rules.

My one significant memory of Candyland as a child is that it made me hungry to play and that it was the easiest game to cheat one (not that I recall doing it often, but you could stack the deck with those jump-ahead cards without too much work...)

I’m actually considering something here that is less of a Markov chain than Candyland played normally, due to the lack of shuffling, since the card-counting aspect is more applicable here.

If there were an infinite number of decks shuffled together at the onset, I suppose it would resolve into a Markov chain.

I’m not sure I see the problem here.  Any game of pure luck (like Candyland, or, say, LCR) has a predetermined winner if you know all the random events ahead of time.  For LCR, a non-player (or a computer) could roll the die a jillion times and create a deck of cards showing all the results in order.  Something like this is often done for Settlers in tournaments.  So there’s no real paradox, just a little confusion over the multiple meanings of the word “random”.  I’m relatively certain there’s no need to turn to quantum mechanics!

And Eric, please don’t tell me you believe that there’s a difference in the player experience if the Candyland cards are in a deck or if they’re all tossed into a bag and chosen randomly!  I’m assuming you don’t believe in THAT bit of superstition!

The question here is whether the result is randomized due to the shuffling alone, and thereafter predetermined, or whether each individual card flip is a completely random event.

This does create a difference between a deck-stack (wherein one can take their entire game into their hand at the beginning and know how many turns it will take to cross the finish line) and a bag-draw (which is equivalent to a die-roll).

I’m also pretty sure there is a distinct difference between a potential probabilistic result and one whose results are already known.

Sure, if you choose to stack the deck before beginning play, then a deck of cards is different from drawing cards from a bag.  But if you’re playing by the rules and randomly shuffling the cards at the start, the two are completely equivalent as far as the players are concerned.

And as far as creating the dice rolls ahead of time, that’s why I specified they be done by a non-player or computer.  The players won’t know what the rolls will be, just like they don’t know what the next Candyland card will be.

There are no player choices in Candyland. That makes it a deterministic operation and not a game.

Note that this definition puts some other “games” into the “not a game” category: Snakes and Ladders and LCR would be the first two that come to mind.

But note that not all “Roll and Move” games are deterministic: Parcheesi involves player choice over which of the player’s pieces to move the amount shown on the dice. This decision becomes more complex with more pieces, like Backgammon.

Even if you had only two pieces in Candyland and choose which of the two to use the color card to move, Candyland would then involve at least some player decision and would then become a game. Keep adding more pieces and more decisions and you get “That’s Life” http://bgg.cc/boardgame/17240 , which people actually enjoy playing.

I’m amazed by the accusation that only bad parents play Candyland with their kids. Do you really not get the point behind Candyland? It’s not about strategy. It’s about learning colors… learning to take turns… learning to follow a path… learning the difference between one and two… learning how to follow a sequence of actions. I tried playing Agricola with my two year old, but that didn’t go so well… so Candyland it is. And for those that insist that “good parent don’t play Candyland with their kids” it would be nice to hear your alternatives.

I think that discussions of the probabilities underlying Candyland, and the consequence of incorporating different methods of randomization is an interesting discussion.  It is, sadly, the one conversation regarding Candyland to which I can contribute least, but gain from most.

That said, I find myself in agreement with Alan Reeve with regard to how “good” Candyland is as a toy/game.  As he writes, it is about learning certain concepts and norms.  In this regard Candyland is a perfect tool.  Candyland also has the benefit of teaching players how to be good losers by separating losing at a particular game from any potential “flaw” within the loser.

If you lose at Chess, it is likely because you are an “inferior” player.  If you lose at Candyland, it is because of chance.  In fact, you were destined to lose, but didn’t know it.  This can be a wonderfully useful tool.

Do you want to play hour and hours of Candyland once you learn the norms and concepts?  The likely answer is no. 

But will learning those norms help you play later games?  Likely. 

Will playing Candyland engender a life long love of game play?  It did for me.

Alan, my son is now only seven weeks old, so I’ll have get back to you in a few years. In general, many titles in the HABA line seem like nice alternatives to Candyland as they require the players to make choices, while also teaching counting, colors, turn-taking, and so forth. I’m looking forward to seeing which ones make the cut in a few years, although I know that Trotofants is already in the win column for sheer goofiness.

Eric

I’ll come to the defense of Candyland, too. Kids can’t easily learn how to play a game with choices until they’ve got the very basics of how a game works. Candyland lets them focus on the whole taking turns and winning/losing things with little distraction. A month of two of Candyland (until they get bored) is a perfect introduction to board games, in my experience.

God does not play Candyland with the universe.