I've been checking out the forums a lot the past few days on Karakuri and Dark World. Seems like the trend for Karakuri now is to run 2 Kuick and 1 Soldier, and also cutting out the lone Strategist. I can agree with cutting the Strategist; it never really served too much purpose other than being another tuner and a draw trick with Bureido. A hand of Komachi+Strategist kinda sucks, and it's happened a lot. I'm still not entirely sure on the lower count of 4-star Karakuri, but I can admit that 3 Kuick led to some dead-draws. I may actually include a Haipa as well, as it runs over T-King, you can grab it with Soldier, is another 4-star, and lets you easily draw with Bureido. I haven't tested any of these things, just stating the trend I've been seeing on the boards.
As for DW, DGz DW folks are still in love with Claudio's turbo build, even going to the extent of saying it's the best deck this format. Maybe it is, but then again, most people say the same thing with what ever deck they happen to see success with or looks really great on paper. I kinda like Marcus Carisse's build as well, so I may tweak the build I have to find a happy middle between the two. I know I was playing the deck completely wrong in my testing with Danny, so I'm willing to take a look at the deck again. The one thing I don't like about either deck game 1 is that it pretty much rolls over to a turn-1 Laggia + D-Fissure, or hell even D-Fissure itself, since the decks don't main any copies of MST. Neither deck can go off if the DWs don't hit the grave. Sure there's games 2 and 3 when you can side the S/T removal in, but it seems unwise to play something where you're willingly OK with auto-losing game 1 against a particular (and in a larger tourney, popular) match-up if they open with a 3-of card. Even if no one's playing DinoRabbit besides the team, in my randy locals, rogue stuff like Macro pops up from time to time as well.
So for the moment I've decided to switch back to TGU Agents for my main deck so that I can utilize the Tour Guides I have. I ran some numbers to compare it with Karakuri, at least my particular builds, and each criteria that I considered and compared was in Agents' favor or a tie depending on how you looked at it. My categories were "first-turn plays", "deck-thinning cards", and "monster removal" since I thought those were some of the more important aspects of any deck, and those are also easy to calculate. I did these with the intent of not being biased towards either deck.
For first-turn plays, I assumed that I was going first and the match-up was a non-factor. If it was a reasonable and potential play, I included it. For example one may say that setting Sangan and passing is a bad play, but it is still a reasonable play nonetheless and is only hindered by TKing, which only the team runs maindeck anyway locally. I also didn't include trap-based plays since basically the same traps get played in both decks, and most people don't say "oh yea my ideal first turn play is to set Warning and say go."
It was in Agents' favor, 12 to 10.
Agents have: 3 Venus, 3 Earth, 2 TKing, 3 TGU, and 1 Sangan (assuming the play is to set Sangan).
Karakuri have: 3 Merchant, 2 Neutron, 2 TKing, 2 Duality, and 1 Soldier (set).
In this category, I gave a point for every card that either thinned the deck, digged for cards, or gave you an immediate +1 upon successful resolution and assumed its effect was not negated in some way. I also assumed that there would be legal targets for Stratos-like cards. I also didn't include extraneous things like Synchros, but in retrospect it could and probably should be, since Burei and Bureido are important aspects to Karakuri.
Excluding Synchros, Agents' favor 13 to 11. Including, a tie (1 point each for Burei and Bureido)
Agents: 3 Venus, 3 Earth, 3 TGU, 1 Sangan, 3 Maxx C
Karakuri: 3 Merchant, 2 Neutron, 1 Soldier, 3 Maxx C, 2 Duality, (Bureido), (Burei)
Even though most builds run 2 of each, I'm only giving one point for each since they require some kind of 2-card combo. I wanted the comparisons to be done as single, stand-alone plays.
Here, I didn't include Synchro-based removal like Orient, Black Rose, Scrap, or Trish. I also assumed that all cards would not be negated or destroyed.
Agents favor 13 to 9.
Agents: 3 Hyperion, 1 BLS, 1 Sorcerer, 1 Hole, 1 Torrential, 1 Force, 2 Warning, 2 D-Prison, 1 Bottomless
Karakuri: 1 Hole, 1 Smashing, 1 Torrential, 1 Force, 2 Warning, 1 Prison, 1 Bottomless, 1 Judgment.
Honorable mention to Fiendish Chain for Karakuri as it is pseudo-removal, but I didn't feel that it was quite enough to be counted here.
Of course all of these things depend on build, but I tried to do it in an "in-general" fashion as well as considering my builds. I don't buy crap like "well you could max Neutron, add 3 Cash Cache, 2 Smashing, 3 Fissure, splash the TGU engine and some lights and add BLS to increase the counts for Karakuri." Those would lead to false and skewed calculations. 3rd Neutron and 2 Smashing is reasonable, sure, but it's not what I'm running and I ain't seen the TGU engine and the rest in YCS-topping Karakuri yet.
Moving on, I've always been curious to see the exact number differences between 40 and 41-card decks, so I set to find that out after reading Pat Hoban and Claudio talk about the subject on DGz. I took stats and all but was never really exposed to finding out how I could specifically find and calculate the differences. It was actually pretty easy via Google search, and I found this site that has a handy hypergeometric distribution calculator: http://stattrek.com/Tables/Hypergeometric.aspx.
Below is an Excel shot of calculations: