*Author

Differ

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4643#msg4643
« Reply #24 on: December 15, 2009, 10:09:53 pm »

Clearly 6.71 is different from 5.685. Given that my simulation is achieving rather stable results (+- 0.01), I'm assuming that one of us has a bug.

Here's my new code: http://pastebin.com/d49e77a4

Differ

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4644#msg4644
« Reply #25 on: December 15, 2009, 10:09:53 pm »

Found a bug. New stats are that fahrenheit is awesome, and you should run 3 of them. Sorry for sucking at programming.

Was kind of wondering how the deck won in 4 turns.

joob

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4645#msg4645
« Reply #26 on: December 15, 2009, 10:09:53 pm »

I'm pretty sure that would get better results than mono fire. Neither of them can win in 4 turns but I think earth deck might have a better chance of winning in 5.
I kept track of turns taken to win to compare my monofire ( with 2 fahrenheits) with monoearth (5 shriekers and 2 long swords, though I'm not sure about the swords). I've only done 25 games with each, but so far fire wins in 6.28 turns on average and earth in 5.74 on average. Fire won only 2 games in 5 turns, and earth was about half and half between 5 and 6, which would point to earth being a bit better.

However, if you consider actual time spent instead of turns, clicking evolve on each graboid will take time that otherwise wouldn't be taken. Not sure how much extra time that would come out to, but I guess it could add up.

Also, going second has the distinct advantage of 1 extra draw, and all but one of the 5 turn wins were games I went second, so the data could be off if I just got more second turn games using earth.

Even with those questions, I still feel earth is a bit better, but maybe it's just because I find it more fun. :P

JTWood

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4646#msg4646
« Reply #27 on: December 15, 2009, 10:09:53 pm »

lol  ;D

lomus

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4647#msg4647
« Reply #28 on: December 15, 2009, 10:09:53 pm »

Heh, Scaredgirl :)

run:

24 towers
6 dragons

Average: 7.15 turns to won :)

lomus

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4648#msg4648
« Reply #29 on: December 15, 2009, 10:09:53 pm »

I can try earth deck.
Only need to change simulation program.

Maybe tomorrow I will post results.

lomus

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4649#msg4649
« Reply #30 on: December 15, 2009, 10:09:53 pm »

Well, because I programmer as well, i wrote my own test of "brick wall" simulation.
Of cause, it could be buggy, but, here are results:

Fastest 30-card deck (according to my simmulations):

16 Towers
6 eaters
6 dragons
2 fahr

This deck won on average of 6.71 turns against totally passive opponent.

Most interesting fact, is that eaters actually didn't speed this deck up a lot.

22 Towers
6 dragons
2 fahr

Have average of 6.81 turns to win. (Difference is only 0.1 turns!!) So, I suggest not to upgrade eaters :)

Scaredgirl

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4650#msg4650
« Reply #31 on: December 15, 2009, 10:09:53 pm »

Computers don't know s***. I still think Fahrenheit sucks. :)

Scaredgirl

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4651#msg4651
« Reply #32 on: December 15, 2009, 10:09:53 pm »

that would really suck to do all that work just to come up with basically the same assumptions i made in my original post with no math at all lol.
Differ did a good job with his calculations. I find computer simulations like that very interested.

If I read that correctly, it means that Fahrenheit sucks, something I noticed myself. I played with and without Fahrenheit and it seemed to just slow me down. Problem is that Fahrenheit doesn't do that much damage when you spend all your quantum on Dragons.

What I really don't understand is why you have that shield because ou don't need that to beat level 3 and having it in your deck will slow you down for sure.

Scaredgirl

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4652#msg4652
« Reply #33 on: December 15, 2009, 10:09:53 pm »

Heh, Scaredgirl :)

run:

24 towers
6 dragons

Average: 7.15 turns to won :)
Like I said, computers don't know s***. :)

Have you tried simulating:

Mark of Time
18 x Stone Tower
6 x Graboid
6 x Shrieker

OR

Mark of Time
19 x Stone Tower
6 x Graboid
5 x Shrieker

Or is that too much coding because of all the evolve and stuff?

I'm pretty sure that would get better results than mono fire. Neither of them can win in 4 turns but I think earth deck might have a better chance of winning in 5.

Scaredgirl

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4653#msg4653
« Reply #34 on: December 15, 2009, 10:09:53 pm »

I can try earth deck.
Only need to change simulation program.

Maybe tomorrow I will post results.
Cool. I'm interested in seeing the results.

Chival

  • Guest
Mono Fire Speed Deck for grinding AI3 https://elementscommunity.org/forum/index.php?topic=202.msg4890#msg4890
« Reply #35 on: December 15, 2009, 10:09:54 pm »

you guys also have to realize these are just simulations that do nothing but damage per turn. If you replace the 2/1s with towers i guarantee the impact would be greater than the simulation thinks. The ai loves to waste their damage cards on my 2/1s before i get my dragon out. thats a huge thing.

 

anything
blarg: