[OUTDATED - v1.27] turns to win study 3.0 (version 1.27)

[willng3]

Okay, I've just updated the upgraded list which took longer than I expected.  Special thanks to everyone assisting with this project; you are the reason that we can give newer players the information they need.

Testers:  If you're confused about why there's a slight difference in the formatting for the stats you posted and what I've posted, don't fret, I just nitpick at formatting inconsistencies when posting is all.  hrmmm and I are likely going to talk things over to see if there are details we can add/change to the current layout, but for now, enjoy

Also, if you find an error in the current lists (wrong mark, broken deck links, etc.) then feel free to let us know.

[pikachufan2164]

Commencing testing for PvP Graboid Rainbow (http://elementscommunity.org/forum/index.php/topic,7528.0.html) (the classic Antagon Novagrabby).

Stat collection will be slow, as I have War and an intense summer course to deal with, but hopefully things will free up and I'll find some spare time to test over the next month or so.

[Jappert]

Testing:

Poison and Druids on Drugs (http://elementscommunity.org/forum/index.php/topic,13621.0.html)
by geekz_always_win

[willng3]

AI setting on high speed.

Fire Lava Destroyer Rush Ver. 2 (http://elementscommunity.org/forum/index.php/topic,11389.0.html)

Games: 200

Avg ttw:   5.948

Avg. sec/game: 56.950

Wins: 194

Losses: 6

Ems: 18

posted by - willng3

Code: [Select]

5f2 5f2 5f2 6rl 7dg 7dg 7dg 7dg 7dg 7dg 7dg 7dp 7dp 7dp 7dp 7dp 7dp 7dq 7dq 7dq 7dq 7dq 7dq 7ds 7ds 7ds 809 809 809 809Games:  200

avg. Sec/Game:   56.950
Sec/Turn:   9.343724364
Clicks/Game:   15.580
Clicks/Turn:   2.556193601
Score/Min:   19.0166813
Electrum/Min:   19.49078139
   
Time overall (min.):   189.833
Wins:   194
Ems:   18
Losses:   6
Score:   3610
Electrum:   3700
Avg ttw:   5.948

ttw 3:   0
ttw 4:   24
ttw 5:   74
ttw 6:   47
ttw 7:   21
ttw 8:   11
ttw 9:   10
ttw 10:   4
ttw 11:   0
ttw 12:   2
ttw 13:   0
ttw 14:   0
ttw 15:   0
ttw 16:   0
ttw 17:   1

Surprise, I didn't forget about it after all!  Actually this would have been completed several weeks ago, but my computer decided to screw up on me 4 times after I had done 100 games so I sort of put it off out of frustration.  Anyways...

Not too surprised by the slight decrease in speed as the AI is smarter compared to its wimpy 1.26 self.  The 17TTW was a huge disappointment however; bad hands combined with the Pandemonium + Voodoo Elder makes for terribly long games.  I was also disappointed to not see a 3TTW appear here; it had appeared in at least 2 of my previous attempts.  But oh well, you expect Cremation rushes to be at least somewhat inconsistent by now.
By popular demand...I have selected Lava Rush vNG 2.0 (http://elementscommunity.org/forum/index.php/topic,6960.0.html) as my next deck to test.  With T500 drawing ever closer, I figure it's a good idea to have its stats immortalized forever more.  Assuming my computer likes me this time, I hope to get this tested within a week at most.

[Jappert]

Testing:

Poison and Druids on Drugs (http://elementscommunity.org/forum/index.php/topic,13621.0.html)
by geekz_always_win

I am very (VERY) slowed down by not having acces to the internet anywhere but during work. Just want you guys to know I haven't forgotten about this project, I just didn't get the chance to pay any attention to it.

[willng3]

AI setting on high speed.

Lava Rush vNG 2.0 (http://elementscommunity.org/forum/index.php/topic,6960.0.html)

Games: 200

Avg ttw:   6.264

Avg. sec/game: 58.385

Wins: 197

Losses: 3

Ems: 20

posted by - willng3

Code: [Select]

5f2 5fc 5fc 5fc 77e 77e 77e 77e 77e 7dg 7dg 7dg 7dg 7dg 7dm 7dm 7dp 7dp 7dp 7dp 7dp 7dp 7dq 7dq 7dq 7dq 7dq 7ds 7ds 7dsGames:  200

avg. Sec/Game:   58.385
Sec/Turn:   9.194488189
Clicks/Game:   15.325
Clicks/Turn:   2.413385827
Score/Min:   19.08366875
Electrum/Min:   19.59749936
   
Time overall (min.):   194.617
Wins:   217
Ems:   20
Losses:   3
Score:   3714
Electrum:   3814
Avg ttw:   6.264

ttw 3:   0
ttw 4:   12
ttw 5:   68
ttw 6:   50
ttw 7:   29
ttw 8:   21
ttw 9:   8
ttw 10:   1
ttw 11:   4
ttw 12:   3
ttw 13:   1
ttw 14:   0
ttw 15:   0
ttw 16:   0

I'm glad to see that the results here are pretty darn close to what DesertKnight posted on the vNG thread already.  The results here show exactly why vNG is referred to for T50 grinding purposes; 3 losses from a Cremation rush of this speed is simply unreal.  However I did get the feeling that the speed could be improved somewhere; there were countless games were I was exactly 1 or 2HP from finishing off the AI, which would have yielded a 4TTW as a result.  However, running back over the duel and seeing where I could have improved my damage output, I had to conclude that there was nothing else I could do that would have given me that extra 1 or 2 damage I needed.  I suppose an explanation for this could be that the change from a to an mark didn't allow my Destroyers to grow one final time due to quanta shortage or possibly the BLights were better for damage than Gnomes; as long as we're comparing this to V2 that is.  I honestly can't say conclusively, but for now we have these excellent stats to remember forever more.
Next to be tested is .........And Slightly Faster! (http://elementscommunity.org/forum/index.php/topic,13380.0.html)

[Gorthos]

We should take these figures with a grain of salt.  200 games is not enough to get a precise estimate of a deck's speed.

willng3, thanks for providing a detailed breakdown of your TTW results.  With a little statisics, that allows us to estimate how close the TTW average from those 200 games is likely to be to the TRUE average if you played thousands of games with the deck.

ttw 3:   0
ttw 4:   24
ttw 5:   74
ttw 6:   47
ttw 7:   21
ttw 8:   11
ttw 9:   10
ttw 10:   4
ttw 11:   0
ttw 12:   2
ttw 13:   0
ttw 14:   0
ttw 15:   0
ttw 16:   0
ttw 17:   1

A little calculation tells us that the standard deviation of a given game with this deck is 3.01 turns.  From this we can calculate the standard deviation of the average, by dividing by the square root of the number of games.  The standard deviation of the average TTW is 0.216.

What does that mean?  It means that we only have 68% confidence that the TRUE TTW of this deck lies within 0.216 turns of 5.94.  If we said that the true TTW of the deck is between 5.724 and 6.156 turns, we would have a 32% chance of being wrong.

That's almost half a turn - not very precise at all!

In conclusion, if the calculated TTW of two decks lies within about half a turn of each other, we can't be sure which one is faster - the TTW for either of them could easily be a random fluke.

[Camoninja]

But it does not matter if it deviates within a quarter of a turn, you know most of the time it will take 6 turns to win, maybe 5. I really don't think that makes as much difference as your clicking speed.

[Gorthos]

Yeah - and that's basically all we can do here, separate the 5-6 turn decks from the ones that are slower.  We can't answer the question "Which is the fastest?"

[badivan1]

While we may not determine which deck is truly the fastest, we can at least observe the "explosiveness" of decks (i.e. the chance of drawing a very good hand) and, more importantly, the efficiency and/or effectiveness of Speed EM decks.  While one may grind AI3 with, for example, a Cremation deck, one could use an Speed EM deck, which trades some speed (usually 1 to 2 turns) for a lot of EM's, and effectively grinds AI3 more or less twice as fast than with the fastest deck against AI3.

[Gocubbies1212]

I found a deck that wasnt in the list that should be tested.  it's supposed to be faster than da fastest speedbow eva!  I just don't have the resources to make the deck though :/

Link here. (http://elementscommunity.org/forum/index.php/topic,13380.msg179935#msg179935)

deck:

Code: [Select]

6rk 6u1 6u1 6u3 6u3 6u3 6u3 6u3 6u3 6ve 6ve 6ve 6ve 6ve 6ve 713 747 747 77g 7ae 7ae 7dq 7dq 7dq 7dq 7gm 7jr 7qb 7qb 80g

[willng3]

I found a deck that wasnt in the list that should be tested.  it's supposed to be faster than da fastest speedbow eva!  I just don't have the resources to make the deck though :/

Link here. (http://elementscommunity.org/forum/index.php/topic,13380.msg179935#msg179935)

deck:

Code: [Select]

6rk 6u1 6u1 6u3 6u3 6u3 6u3 6u3 6u3 6ve 6ve 6ve 6ve 6ve 6ve 713 747 747 77g 7ae 7ae 7dq 7dq 7dq 7dq 7gm 7jr 7qb 7qb 80g

Yep this one's on the list at the moment; I'm in the process of testing it as we speak.

Next to be tested is .........And Slightly Faster! (http://elementscommunity.org/forum/index.php/topic,13380.0.html)