Do you ever feel like
maybe the shuffling system in the game isn't truly random? It's not something you could really notice with using an ordinary deck, but it does seem like certain cards are often close together in games (as in two or three cards away). So, just to settle the matter, why don't we test it out? It shouldn't be too difficult to do. We need a standard deck and a way to compare the drawing orders between games. Here's the deck:
6qq 6rj 6rk 6rl 6rm 6rn 6ts 6u5 710 71a 744 74c 778 77f 7ac 7al 7dg 7dn 7gk 7gs 7jo 7k1 7ms 7n5 7q0 7q8 7t4 7tb 808 80eWith this, you'll practically never have to discard, which kinda slows you down. If you use it against AI1, you should have enough healing power to last you through the entire deck. There are two cards from every element and six other cards, and each card type is the same, save the shards (weapons/towers only).
We can assign a number to each card. I shortened the names in the list just to make writing them faster.
Card order in deck (you may have to adjust the numbers if you copy this into word so that they line up).
Quantum 1
Dirk 2
Sword 3
Gavel 4
SoD 5
SoG 6
Amethyst 7
Discord 8
Bone 9
Arsenic 10
Gravity 11
Titan 12
Stone 13
Pulverizer 14
Emerald 15
Staff 16
Burning 17
Fahrenheit 18
Sapphire 19
Poseidon 20
Light 21
Glory 22
Wind 23
Eye 24
Time 25
Eternity 26
Obsidian 27
Vampire 28
Aether 29
Electrocuter 30
Once you have copied the order of cards from a game, you can throw in the numbers that will be used for data (which I will cover in a minute). Here is an example game (I didn't finish it because it takes
forever... this would be a lot easier if someone would write a program).
The "One down" columns mean one card below that card. The first number is in reference to the deck list, the second is in reference to the drawing order, the third is just the two added together (not really important but I put it in anyway). So if we look at the One Down column for the first card I drew (light tower), to get the first number I looked at the number in the "deck order" column (21), then found the card below it from the deck order list (number 22, the morning glory), then looked at how far apart they were in the drawing order (13). The second number was produced by looking at the next card drawn after the light tower (an obsidian tower), then finding the difference in the deck order (the number is positive if the card lower on the drawing order list is higher, negative if it's lower... I'm not sure if it matters if we make a number negative or positive, but I decided to differentiate between the two anyway, and I think it may be interesting to compare the total number of positives to the total number of negatives). I'm assuming from all of that information yall can figure out everything for the other columns, and the "total" column is just the sum of the other columns.
Here is what I think the final "deck order" list would look like with information from games using the drawing order "tables."
The columns are basically the same as in the drawing order table. The only difference is that they are averaged.
Like I said, this would be easier if someone wrote a program for it. I also think it would be possible to set something up in excel/spreadsheet, but I haven't used that for anything complicated in years.
Statistics aren't really my forte, which is kind of what this is. Patterns are though, so if there are any types of patterns that can be gathered from the information, I'll probably find it. If there's any way to improve what I've started, suggest it. I highly doubt that the information I'm trying to gather is the best way to observe randomness (especially since I only look at cards up to three draws away). It's just a start.
