Discussion Forum

When reading your answer one question comes to my mind: Against how many decks would you test a newly designed deck, and how many games would you play, before you would call the deck "tested"?
Personally I would have chosen 3-4 solid decks as testing opponents and 5-10 rounds (half of them with, half of them without sideboarding). 1-2 Aggro decks, 1 control and 1 combo deck.

Burn really isn't one of my favourites to play...neither as burn player, nor against it. I'm not saying it's stupid to play, but I'm missing the interaction between players and, simply said, the fun.The same goes for mono black or mono blue control, because for me it's simply frustrating and boring if you want to play a creature/ spell, and can't because it gets countered or removed in an instant.
I did choose mono black (MBC) control as testing-opponent, because I don't have a pauper red deck win or burn built at the moment and I wanted some sort of control-deck in the tests. But my experience tells me that it will be very hard to beat a deck with 16 removals against creatures in a very creature-based environment (tough I'm really interested how my Dimir Mill will compete against MBC).
As for [[Pestilence]], I'm using this card in my MBC built, and it can be devastating when combined with [[Eternal Thirst]]. At first I was quite sceptical about [[Cuombajj Witches]], but combine it with [[Eternal Thirst]] and you got yourself a life-gaining removal for 1/1's (not as powerful as [[Goblin Sharpshooter]], but still worth the mana) that will gain you life every time you kill something AND get a +1/+1 counter. Throw in [[Pestilence]] and things start to get really ugly.

Tron and Walls Combo are decks that I don't have any experience with, simply would have to copy a good deck-list from the internet and see what I don't like personally and change that.

I'd try to play against at least 1 of every major "archetype"- something to represent control, something blitzy, something comboey maybe, ramp and big mana decks. Though if you're trying to be competitive- you probably should just be copying known winning meta lists and see if your decks stack up. Mono black is a really popular and powerful deck- I'd say if you're trying to be competitive and your not keeping it in mind...you're probably gonna get wrecked by it if you ever come across it haha.

I actually just played against one the other day- and I IMMEDIATELY revised my list. I had no game against it and it really showed the weaknesses of my deck. You should at least have some kind of game plan/side strategy for it. Though I guess it should also be noted- you're not going to have a good matchup against EVERY deck- but you should probably have a plan against the popular ones at least... at least if you're going for competitive build- if you're not though then just have fun and go for it haha. I don't like playing burn either for the record haha. But there's a reason it pretty much exists in every format.

It may be hard to beat a "Removal" deck if you're a creature deck- but if thats what the weakness is- you may need to rethink the list somewhat- get rid of 3 and less toughness creatures to get em out of bolt range/ or have some type of recursion etc. There are ways to make it work- you just need to be able to out creature your opponents removal. Maybe that means play a creature that puts another token into play etc so they need two cards to deal with one of your cards etc.

hipponox has deleted this comment.

hipponox has deleted this comment.

hipponox has deleted this comment.

hipponox has deleted this comment.

hipponox has deleted this comment.

hipponox has deleted this comment.

hipponox has deleted this comment.

From the decklists I've looked at, and the articles I read, I feel that I will be more at home with a walls combo deck, where Fangreen Tron will be the next comparable Tron variant.
hipponox also mentioned elves, and this is a deck I have no problems testing against, so I could give them a try.

Burn/RDW...*sigh*...I know that this is an archetype that is really played often, but I feel that it's a pain in the ass to play against, because, like my MBC variant with tons of remval, i't simply no fun when you're creatures are either removed in an instant, or ignored, because your opponent knows you can't hurt him enough and focuses on your life.

I also asked hipponox this question, so I'm simply copy & pasting it here:
When reading your answer one question comes to my mind: Against how many decks would you test a newly designed deck, and how many games would you play, before you would call the deck "tested"?
Personally I would have chosen 3-4 solid decks as testing opponents and 5-10 rounds (half of them with, half of them without sideboarding). 1-2 Aggro decks, 1 control and 1 combo deck.
I fully understand that the more decks you test against the better the results will be. The thing is none of our small playgroup really wants to help me on this, so I have to do this all alone. Testing a single deck will be no problem, but I have 50+ decks that I want to get a feel for how they behave and streamline them. With 3 decks to test against and 5-10 games, that will be 15-30 games per deck and 750-1500 games for all decks!

Haha, that latter question is a statisticians dream! Though things like p-values are hard to come by in the field of Mtg deck testing xD And given the game's complexity you could always argue that a deck has not yet been tested. It seems to me that you are undertaking a task that is much too monstrous in magnitude to be a reasonable way of spending time. But here's my two cents;

I think that to get a feel for a matchup you should have something like a checklist of all the situations that can be reasonably expected to arise in that matchup a statistically significant number of times. E.g. testing Stompy vs MBC;

Games with an early Young Wolf (as that is the bane of control decks)
- When MBC has Chainer's Edicts
- When it does not

Games with an imbalanced pump-spell to creature ratio drawn by Stompy
- When MBC draws mostly removal
- When MBC draws mostly creatures

Games in which Stompy draws the nuts
- When MBC also draws the nuts
- When it does not

etc.
And then all of these when MBC goes first & when Stompy goes first, and when one mulls to 6, and when the other one does. Then when they both do, etc.

You could drive yourself insane doing this tbh. And again, I think a testing project of this magnitude is not an optimal way of spending time for the simple reason that your testing will either be shallow and therefore inconclusive, or so deep and so exhaustive that we'll have sent some poor soul to Mars by the time you're done.

The 'easy' way of doing this might be this: You ask yourself 'What is this deck's gameplan?'
Then you test for situations where one or both decks have their gameplan go well (e.g., Slivers draws a bunch of lords, TE gets a TE on turn 1, Delver gets its namesake flipped on turn 2, etc.), and for situations where one or both decks have their gameplan go poorly (no TE for TE, burn floods a bit, Delver takes a while to learn to fly, no Gary for MBC, etc.). You repeat until you've tested this for going 1st/2nd as well as for mulling to 6/5 for both decks.

OK, I really had a good laugh on your answer, especially on "...or so deep and so exhaustive that we'll have sent some poor soul to Mars by the time you're done." :)

"You could drive yourself insane doing this" ... *smile* Overall this sounds like a call for wickeddarkman as I know he is heavily into statistics and analysis. (No pun intended, I really appreciate your work here wickeddarkman)

At the moment I have tested most of the decks against my mono green stompy, so they have seen at least some action...and thus I have seen their weak-points and also was able to repair some of those. Some of those decks will never be safely playable against other decks (e.g. dimir mill is simply to darn slow now matter how you tune the deck). So in the end I guess I will take a second beatdown deck with a go-wide strategy and/ or something midrange-ish like boros/ mono white bounce and play some games against the decks to be tested. This should at least give me a feel for how the decks will behave against different opponets. The polishing of the deck then will have do be done via sideboard and/ or if you find someone who likes the deck, is willing to play it (no matter if IRL or e.g. via Cockatrice) and write down the results and sense of play for me.

"It seems to me that you are undertaking a task that is much too monstrous in magnitude to be a reasonable way of spending time" ... (sadly) this is exactly the form of project that I will start/ pick up...ALWAYS. Designing your own variant where you combine Heroquest+AD&D+Zombicide+ and so on, with a complete world (/continent). Making your own eating AND gaming-table. Designing a model-ropeway to be printed on the 3D printer. Just to name a few of those nuts project.

MUKTOL:
I'd say I'm more into evolution than statistics.
At the moment I'm taking a "small" break from my main project, my shrieker deck, to speed test my paperstrip process. I'm doing it to see if I'm capeable of beating a metagame from scratch within a month.

Search for the decktag: ingest
And click on version three of the deck.

I'm currently running 10 games against each of my usual 64 testdecks.
I know that in the past I've run it with as low as 5 games.

This time I'm not keeping track of anything but individual points for the cards in order to cut down on the gaming time.

I think the longest time it's taken so far to play 10 games is three hours.

That would be 18 minutes per game.

Even if it would only take 5-7 minutes per game (which can be realistic as I'm using Cockatrice to play against myself, so the waiting-time is nearly zero) that would still mean 60-180 hours of pure testing time for all the variations, not included the time needed to improve the decks.
So I will have to come up with a different idea. Maybe I will only test the decks I think REALLY are fun/ worth to play, or limit the number of games or...IDK.

My own system is based on physical play, which means I have total control of what gets registered.
I have had many problems and many restarts, but overall things are shaping up.

Since you play against yourself on cockatrice you might be able to simulate some of the processes that I use.
During runs I have "blank" cards that can be played as anything when I have them in hand.

You should be able to play some cards that have no use in the deck but getting you to write down cards you wanted in certain situations.

When you play against enough decks there will emerge a clear pattern of what you could change your blank cards into.

If it's possible to play with alternate art on cockatrice you could replicate almost all of my system.