Community

Oct 10, 2011 -- 1:10PM, SheerMadness10 wrote:

The wolf run ramp deck needs a creature to win, right? Just add more 1 for 1 kill spells. This isn't rocket science.

You should actually test match ups before you declare "simple solutions" like that.  I'm not sure how much you test anything though.  All of your decklists you post on here were thrown together in 5 minutes. 

Testing isn't a truly objective way to build and tune a deck. There are too many variables involved that can skew your result set. I'm not discrediting it, but if it is the only way you evaluate a deck you are doing your list a disservice. Conversely this means denouncing an untested list is foolish.

Finally, Kessig Wolf Run is the land that gives +X and Trample to a target creature, right? And its the one that Primeval Titans them out to make it effective? Right?

K, make sure you have access to ample removal. To be more expansive if you can spare slots for ghost quarters do it; sideboard may be sufficient.

Oct 10, 2011 -- 1:10PM, SheerMadness10 wrote:

Sorin Markov seems way too slow vs RDW, where you actually need the life gain.  It's +2 ability doesn't do much vs pod, flare, or ramp.  It really seems like it's only worth running if you get a vengeance to go along with it, and survive long enough to get them both off.  That's my opinion on Markov anyway. 

This is pretty accurate, but its not tough to hold RDW down early game in black. The problem is you're playing 1:1 with them and when you both go hellbent a Shrine puts them into an inevitable spot for victory. You need a Sorin to take the game away from them. The lifegain into the mindslaver pointing their shrine back at their own face is pretty strong.... generally forcing them to crack it asap to kill Sorin or ineffectually at your face.

Oct 10, 2011 -- 6:40PM, SheerMadness10 wrote:

Testing is really the only way to make a viable deck.

I think you're not really saying what you intend here.

Let's go over a healthy life cycle of deck construction.

First, we generally have a decision tree:
1. A builder has found a certain concept they wish to pursue in construction.
2. A builder has a card they love and want to focus in on using it.

Generalizing still let's assume both are being initiated with the ultimate goal of high level competition, say an REL-32.

Next the builder will reach another decision:
1. Focus the concept; I believe it can do something so well I do not need to acknowledge the meta.
2. Survey the meta and consider my interactions. Are they meaningful in my first draft? Repeat per Tiered deck.

From here the builder should go through a series of iterations and evaluate each card in their deck. Is a given card useful in every matchup? Perhaps only 66%? 33%? Does it deserve a mainboard slot? Should it be pushed to sideboard? Should it be cut entirely?

After the builder has thoroughly considered each card should testing begin. At this time it is extremely appropriate to request feedback and discussion because testing is not as reliable as theoretical deck building. The builder should consider all feedback (removing unnecessary flak from commentary as needed to create a core statement about their deck).

From here testing should begin next. A builder should consider the following information about statistical analysis before adjusting their deck:
1. To create a relevant statistical solution you must work off a sample size of at least 200.
2. To create a perfect scientific experiment you must have 1 variable and everything else must remain constant.

That's a really simplified list of things for why 'testing' should be considered helpful, but not the end-all to deck construction. It should be your finale, not the quick 2nd step. Any magic player who isn't a 'professional' will not have time to build a sample size of 200 games against every tiered deck to begin drawing reliable conclusions from. Most any magic player will also tell you that there is always more than 1 variable in any game of magic.

Oct 10, 2011 -- 6:40PM, SheerMadness10 wrote:

There are extremely rare exceptions but if you just throw together some random deck with 0 testing and take it to a big tourney you'll get owned 99% of the time. 

Completely false. Did you know 70% of statistics are made up on the spot, just like this one? If you don't want me to make inflammatory statements perhaps you should consider your own advice?

Oct 10, 2011 -- 6:40PM, SheerMadness10 wrote:

ANY kind of testing is far better than none, and it really doesn't seem like you test your decks at all before posting. 

I think I've illustrated that actually this isn't correct at all. Bizarre.

And to be fair, everyone who has been around for a while knows I do not have time to do a great deal of testing but I adhere to my life cycle of deck construction before posting, I tune per suggested feedback, and I ask questions based on testing of people who have worked on my lists so I can make the best adjustments possible.

It's not important, but I've got proof of my solid deck construction skills... particularly as it relates to mono black. I just don't carry it around on my sig because I don't feel compelled to substantiate myself to an anonymous person on the internet.

Finally, if you want to argue about statistical relevance as it relates to deck construction and testing please post a thread in the Standard General forum. It will be extremely off-topic to continue any discussion on it here. Plus if you take it out to Standard General its not only very topical but you'll get responses from people who you respect more than myself (and you'll find they will agree with my construction steps and the relevance of testing + sample sizes).