Conquer Club

I found only 1 topic http://www.conquerclub.com/forum/viewtopic.php?f=57&t=111357&p=2492787&hilit=+spoil#p2492787 that touched on this subject so forgive me if it has been discussed before.

Concise Description:

From my observation it appears a cashed card has an equal percentage chance of being put back into play and handed out as a "New" spoil card. You would think a totally unused card from the deck should pop up first but it doesn't appear to work that way. IMO that's a bad algorithm .

Using game 7351029 as a reference
In this case the number of players is 5 with 80 territories on the map

• 5x5 < 80, duplicate territories should not be handed out
  
  Round 10
• 2010-08-05 14:57:03 - Nikita42 cashed in a group of F Train: Sunset Park Station, A Train: Hollis Station, and Jackson Heights nuking each region
  
  Round 11
  You can't exactly see the following but boken received the same spoil card F Train: Sunset Park Station
• On 2010-08-07 13:49:17 - boken ended the turn and got spoils
  
  Round 12
• On 2010-08-08 19:34:45 - boken cashed in a group of Bedford Park, Gramercy, and F Train: Sunset Park Station nuking each region
  
  In a face to face game I would never have received F Train: Sunset Park Station that soon. We would not return a set of cashed cards to the deck and then reshuffle the deck. We would:
• 1) Put the cashed cards at the bottom of the deck and let them resurface
  OR
• 2) Put them on the side in a used pile and when all unused cards have been dealt reshuffle the used pile.

Specifics/Details:

A few suggestions to make the distribution of available spoil cards more uniform and fairer all around

• 3) Implement algorithms like item numbers (1) or (2) above
• 4) Create a new/different algorithm
  
  Different Algorithm
• 5) In the beginning all cards in the deck have a weighted value of 1
• 6) After the end of any turn all remaining cards in the available deck increment the weighted value by 1
• 7) Whenever a card is returned to the deck the weighted value starts at 0
• 8) Handing out a card would be the same except
• 9) If a card has a weighted value of N then that card would have N chances to be chosen
• 10) If a card has a weighted value of 0 then there is 0% chance it will be chosen
• 11) In this way weighted cards have a better chance to be dealt out and newer cards entering the deck have less of a chance

How this will benefit the site and/or other comments:

• Better distribution of spoil cards so the same card is not handed out as soon
• Provide the new algorithms as a game option and get feedback from the community

I am a software developer, I would be more than happy to volunteer my time to review the current code and/or implement the algorithms.

Yes, this has been discussed before, but there are situations where you may have more cards than territories. Doodle Earth only has 18 territories, and if you got about 5 people playing, and they all have 5 cards, there are bound to be repeats. We didn't find this to major hindrance in any way, so we left it as it is.

And by the way, we already have a Web Developer, but if you would like to apply, talk to AndyDufresne.

This is not a Doodle Earth map but a NYC map with 80 territories. There should be no duplicates; it defeats the purpose of having a clean card deck in a large map.

It is a hindrance in a Nuclear game when the same territory keeps popping up. Percentage wise if I own the card or if the card has been played I should consider this territory relatively safe, but this is obviously not the case.

Based on your comment, someone took the easy way out and decided to just make every game have duplicate cards. I would like to see this changed

You need support for you're suggestion then.

What does one usually do to get support for a suggestion?

More evidence: 5 player Classic game (7330599) 42 territories, currently running.

As of round 24
After 26 cashes = 78 cards

Number of territories played 0 times = 04 (0 cards)
Number of territories played 1 times = 11 (11 cards)
Number of territories played 2 times = 15 (30 cards)
Number of territories played 3 times = 11 (33 cards)
Number of territories played 4 times = 01 (4 cards)

And it just so happens I will be playing on my next turn a territory (Jakarta) for the 5th time when there are still 4 territories that have not been played. Statistically it makes no sense to have so many territories played 3 and 4 times. It is showing if a card is played it is more likely to go back into play.

After 26 cashes
3 Anchorage
1 Astana
2 Bangkok
2 Beijing
3 Berlin
2 Bogota
1 Buenos Aires
2 Cairo
1 Cape Town
3 Chicago
2 Dakar
3 Delhi
2 Dubai
1 Edmonton
2 Havana
0 Hong Kong
1 Istanbul
4 Jakarta
1 Johannesburg
3 Lagos
2 Lima
3 London
2 Los Angeles
0 Madrid
1 Magadan
3 Manila
2 Mexico City
0 Montreal
2 Moscow
1 Mumbai
2 Nairobi
1 New York
0 Novosibirsk
3 Perth
3 Port Moresby
1 Reykjavik
2 Sao Paulo
3 Stockholm
1 Sydney
2 Tokyo
2 Vancouver
3 Yakutsk

It all depends if people like the suggestion.

If this was a deck of cards, what happens is that a card that is used isn't placed on the bottom of the deck, but instead is shuffled through. This means that it is indeed possible that some cards get used 5 times when other cards haven't been used yet.

I don't believe this is a problem. It may make the game less predictable and more luck dependent, but it doesn't actually make it unfair, because the algorithm is the same for everyone.

Simply putting the cards at the bottom of the deck isn't a solution, because with the lenght of some games on this site, they would actually become predictable. In a nuclear spoils game, the knowledge that your opponent holds the spoils for territories X, Y and Z is far more damaging then the fact that the same territory got nuked for the 5th time this game.

Putting them aside in a new deck and shuffling them when the current deck is empty is also unfair. The cards which were drawn last in the first run through the deck, would only get shuffled in with other cards that already have been drawn two times. Therefore, even with this solution implemented, you would still see an uneven distribution of the amount of times certain spoils get turned in.

Overall, I see no reason to change the current algortihm.

MichelSableheart wrote:If this was a deck of cards, what happens is that a card that is used isn't placed on the bottom of the deck, but instead is shuffled through. This means that it is indeed possible that some cards get used 5 times when other cards haven't been used yet.

What other game do you know that reshuffles the deck after each turn/cash? Is that how you play Risk in a face to face game? If no why not?

MichelSableheart wrote:I don't believe this is a problem. It may make the game less predictable and more luck dependent, but it doesn't actually make it unfair, because the algorithm is the same for everyone.

It seems you are implying games should remove skill as a factor and add more randomness. Look at all the discussions about the "Crazy" dice algorithm.

MichelSableheart wrote:Simply putting the cards at the bottom of the deck isn't a solution, because with the lenght of some games on this site, they would actually become predictable. In a nuclear spoils game, the knowledge that your opponent holds the spoils for territories X, Y and Z is far more damaging then the fact that the same territory got nuked for the 5th time this game.

Are the games supposed to be more about luck or more about skill? Having knowledge of when a card might or might not return implies you are using skill to make a correct strategic/tactical decision.

MichelSableheart wrote:Putting them aside in a new deck and shuffling them when the current deck is empty is also unfair. The cards which were drawn last in the first run through the deck, would only get shuffled in with other cards that already have been drawn two times. Therefore, even with this solution implemented, you would still see an uneven distribution of the amount of times certain spoils get turned in.

That makes no sense. Used cards would never be played a second time until the unused deck is exhausted.

I also mentioned a weighted algorithm can be used to make it more likely (percentage wise) an older card in the deck is chosen before a newer card.