Playtesting: Learning a new deck

[Parcher]

Another thread got hijacked by this topic, so I felt it needed to have it's own.

When first learning a completely new deck, that for whatever reason you are reasonably sure you would like to devote enough time and energy to seriously consider taking to a tournament, how do you go about it?

I'll start off with some of the previously presented ideas:

Godder: For what it's worth, we've found takebacks while testing to be a key part of improving my team's performance. It's critical to know the right thing to do in common situations, and also to avoid silly mistakes, and making bad plays just reinforces making bad plays. If we spot a bad play while testing, we stop, examine the play and the board, and then take it back and continue with the right play. If you're continually spotting the right play immediately after making the wrong play, I suggest testing more, and playing slowly to force yourself to think about what you're doing

andrewpate: I think that if losing games due to stupid mistakes in playtesting makes you upset and causes you to play better, you are not approaching your testing correctly, at least in the early stages (developing a deck or learning to play a new netdeck).  At that in point testing, you are more concerned with learning what the deck does and how it works than with winning; it does not upset me if I take a new deck off the PT coverage and lose 20 games in a row with it; I simply look back over the list, look at some coverage, and attempt to determine why my approach to using it is incorrect.  If I continue to lose, I may look for a different deck.  But if I lose due to mistakes, I might get the wrong idea about things, and those mistakes can be ironed out in the tuning stages later on.  Also, it can be assumed that a person will make few mistakes with a deck by the time he or she gets to a tournament with it if he or she is someone worth taking seriously, and that means that if you lose several games due to repeated mistakes you make playing Stax, your playtest partners will not have a good idea about how to play against Stax.  They will get used to winning because you play lock parts in the wrong order or forget to ramp up your Smokestack.  My team always allows takebacks during early testing, and we have gotten good results from this.  You simply have to maintain the maturity level to internalize your errors and commit to eliminating them even when you take them back.

freakish777: To make it more clear, Flores states that while developing decks (particularly with the use of proxies) you should allow take-backs to get a clearer view of what the deck you're tweaking/building is capable of (I imagine to increase the speed of making the deck the best it can be in the shortest amount of time, and not short-changing yourself in the short run by saying "Oh it doesn't win" when in fact it's the player's fault for various misplays.

In contrast in the long run for one's own personal technical playskill, I feel (and have had reinforced) that not allowing takebacks negatively reinforces mistakes (ie, "I won't do that again") and that allowing them positively reinforces them (although not to the same degree, more on a subliminal level).  Practice makes Permanent.  If you allow yourself to make mistakes in practice, you will subconsciously allow yourself to make mistakes when it's on the line.  Your team may have hit a short run increase in skill (or in deckbuilding), but I would be interested to see the results with the same people over a longer period of time testing without mistakes.  As a side note I also think it'd be interesting to see just how much the deck you're playing vs. playskill matters (assuming you're talking about becoming better deckbuilders) against an opponent with better playskill.

Anusien: One of the things I love to do in testing is at a critical juncture, ask my opponent for advice on the correct play.  I'll be playing versus Josh, and I'll show him my hand and say, "Help me make the correct play here."  I think while it's useful to remember that opponents make mistakes, when you want to test a matchup you want perfect play.

As for myself, I agree with all of these to some degree. In Anusien's example, I know doing things this way has helped immensely. We have played test games where you honestly ask your opponent "Cast Meddling Mage. What should I name at this point?" Or "Do I try and go off in response?" Having someone who will give you the best answer in these situations can be invauable. Especially when they are familiar with a deck that you are not. You just have to check your egos.

I think that not allowing mistakes is correct, but I also believe in takebacks, as long as the game has not advanced from there. Forcing someone to commit to an incorrect block, or Fetch will make them play closer attention to the game state. Not allowing someone to re-think their Gifts Pile, or sequence thier accelleration spells in a Storm deck actually does them a disservice. When new with a deck, you need to figure out what the best plays are. It's not a mistake if you have never faced the situation, it's ignorance. Not only should you be able to backtrack, your partner should assist you.

This can, of course only be continued for a short time. If after this you are still unable to make, or even see the correct plays, then that particular deck may not be your friend.

Please post your ideas, theories, and experiences. There is so little room for players to make sweeping deck changes in the Eternal formats, I'm certain that everyone could benefit.

[Anusien]

One thing that's worth pointing out is that most of us have been playing for a long enough time to have played all the pure archetypes.  For example, I can go to T2 and instantly pick up Flores's Jushi Blue because I've played Mono-Blue Control back with Ophidians.  The relevant threats are different, and your tools aren't the same as Capsize or Thieving Magpie, but the basic principle of counter, counter, counter, fattie is the same.  I've played Long in Vintage, so when I come to Extended and play Heartbeat or Ritual Desire, I can see all the plays.  Sure, you have to learn Sins of the Past tricks and whatnot, but I know how to do Desire math, how to play past counters, when to go off.

This is especially where being an Eternal player is useful.  I played Rift/Slide in T2, and worked on it for a while in Legacy, so when it comes time to test Astral Slide decks or Solitary Confinement decks, I have more relevant experience than any of my PTQ opponents.  I agree with whoever posted in the thread that it doesn't take me much to learn a new deck.  I try not to pick it up the day of the tournament, but Reb or Josh could hand me Ur Snow and I could probably play it pretty well at a FNM after just goldfishing it a few times.

One relevant thing is to work out mana ratios.  For example, when playing T2 Heartbeat in CHK-RAV, suppose you fan open this hand on the draw:
Forest, Sakura-Tribe Elder, Heartbeat of Spring, Muddle the Mixture, Kodama's Reach, Sensei's Divining Top, Remand

You should know instantly whether this hand is keepable.  Same thing if the Forest is an Island.  (On the draw in game 1, generally the answer was Yes, No because of the land to mana ratios).

[andrewpate]

While this is true, not all decks fall into pre-determined archetypes.  Take, for example, a new combo deck such as Sunny Side Up.  Knowing the likelihood of going off with that deck from a given position and with a given hand is tough without several goldfish games under your belt.  Sure, I took to TEPS quickly thanks to my Vintage experience, but that's not true with all decks.  For example, if you had Vintage Gifts experience and tried to play Extended Gifts Rock treating Gifts as the same card, you'd play it far too early a lot of the time (for those who don't play Extended, Gifts Rock almost never wants to play Gifts early, even if it is sure to resolve).

That said, I agree with most of what you said, and I'm not saying I can't play a deck reasonably well within a few games most of the time.  But I'm also saying that if you sack a couple of Skycloud Eggs, cast Second Sunrise, sack the Eggs again, and say, "Wait, I should have casted Reshape for Lotus Bloom before that first Sunrise!" you should rewind and do it most of the time, both to assist the person playing against you and trying to learn how to beat Sunny Side Up when it's played well AND to make sure you're taking note of your specific errors and seeing how you can play the deck better.  If anything, this applies even more to a combo deck.