The time immediately before a Pro Tour is always awkward for me as a writer. While all eyes are on StarCityGames.com Standard Open: Cincinnati to define Standard going forward, I have never seen a single turn of a single game of Standard with Return to Ravnica since I’ve been exclusively testing Modern and RTR Limited for Grand Prix San Jose and Pro Tour Return to Ravnica.
Most of you are interested in how to beat the new Standard decks in order to win the StarCityGames.com Standard Open in Providence this weekend, so my original plan was to just look at the lists from the previous weekend and talk about where to go from here. Patrick Chapin and Ari Lax have done an excellent job reviewing the decks here and here. I could attempt to write about building decks to beat those decks, but without playing any games, they’d be a little loose. There are sketches, but they’d mostly be obvious.
Instead, I’ve decided to try something that I hope is a little more fun this week.
When I lamented that I had no idea what to write about, Andrew Cuneo mentioned that he’d been trying to get Reid Duke to write “The Importance of Cheating in Playtesting.” I thought about that and decided that I could write a little about it but not a full article. The idea that came from it seemed like it’d be valuable though, so I wondered if other bad ideas for articles might offer a chance to discuss some useful concepts.
So of course, I turned to Twitter and asked for bad ideas for articles.
But let me start from the beginning.
The Importance of Cheating in Playtesting
I’ve heard of players agreeing to play games in which it’s understood that both players will try to get away with anything they can. This can be useful practice to learn to catch cheaters if you’re really worried about that, but I don’t think it’s very interesting strategically.
However, I do think a lot of people could learn answers to specific questions they have about matchups or cards by “cheating” more in playtesting. Specifically, you don’t have to play every game in playtesting a real game of Magic.
It’s fairly common for players to play with alternate mulligan rules in testing (I don’t, but there are arguments for and against the practice). Some people say your first two or three mulligans should go to six cards and that you should never mulligan below five because you don’t learn anything from playing a game where one player has such a huge advantage. I think doing this consistently puts you in danger of letting a deck that isn’t quite consistent enough appear playable a little too long, and I also think there’s value in learning to play from behind occasionally.
I think fewer people play a game with a preset opening hand or start with a particular card in their hand and shuffle the rest of the deck, and this is where I think a lot of people are testing inefficiently. If you want to know how effective a particular sideboard card is, you can just play games where it starts in your hand rather than merely playing post-board games. Obviously, this won’t tell you how likely you are to win games after sideboarding, but it will tell you if that card does what you want it to when you draw it.
Alternatively, if you’re wondering how important it is to have a Delver of Secrets in your opening hand in a certain matchup, you could play a set of games where any time you have a Delver in your opening hand you have to take a free mulligan (or just draw seven from a 56-card deck and then shuffle the Delvers in) and then another set where you start with a Delver in your hand. Maybe in that set you only start with a six-card hand to help you get an idea of whether it’s worth aggressively mulliganing for Delver.
The point is that people often playtest games to get specific information that they could get much more efficiently by taking steps to ask their decks the specific questions they want answered instead of just playing random games.
The other question this brings up is the idea of “take backs” in playtesting. Some people play all their games as if they’re playing in a tournament, and some people rewind any time they notice a mistake to correct their mistake. Both ways of playing have merit, so how do you decide which way to playtest?
Well, if you’re trying to build good habits and learn to play in tournaments, you should probably play as you would in a tournament setting. The argument for take backs is generally that if you’re trying to learn which deck is favored rather than which player is favored, you should allow take backs to try to minimize skill differential. This is dangerous, though, because it might end up increasing skill differential if one player is just more likely to pay attention and notice their own mistakes. Most of the time, I think it’s safest to play games straight up without take backs and trade decks if you have to help account for skill differential, but there are definitely times that call for exceptions, particularly if you’re playing with proxies, or one person is unfamiliar with the cards, or it’s someone’s first couple times playing a deck.
@Lansdellicious suggested:
The Best 65 Cards and Why You Should Play Them All
I don’t know what the best 65 cards are right now and I doubt the time is currently right, but a recurring theme in suggestions I got for bad ideas was “playing decks that are too big.”
I still think players are too resistant to going over the minimums in deckbuilding. These days, we see far more one-, two-, and three-ofs in decklists in Magic than we used to (a claim based on no real study, science, or evidence, just my extremely vague impression), and I think players have gotten a lot more comfortable with weird numbers in decklists and that they build better decks as a result.
This isn’t to say that I think that in the future, we’ll routinely see decks of all different sizes, but I think the idea of going over is slightly more taboo than it should be.
I’ve mentioned before that I think that if a deck has a consistent shell and some bullets that you can Tutor for, it can be right to go a few cards over 60 to make room for more bullets.
Consider the opposite case: if there was no minimum deck size, how many cards would you play in Standard? Well, I’m guessing you could build a pretty insane 20-30 card Zombie deck. It would be hard to compete with that deck with 60-card decks, but there are some decks that don’t work with many fewer cards.
The best evidence of this, I think, is Birthing Pod in Cube. I think most cubes shouldn’t contain Birthing Pod because it’s almost impossible to fit a robust Pod chain into a 40-card deck even if you happen to get all the pieces. (Realistically, there will always be some point on the curve where you only have one or no creatures, and it’ll cause a lot of frustration with your Pod.) I think a 40-card Birthing Pod deck is likely to just be worse than a 60-card Birthing Pod deck. I’m not sure if 50 would be better than 60, but there’s a chance that fifty isn’t the right number either. While we’re going up from 40 to find the perfect number, I don’t know for sure that that number isn’t greater than 60.
I think it’s extremely rare that a Constructed deck over 60 is right, but I do think it can happen.
Also, a reasonable number of tournament-winning decks have been 61 cards. I don’t think I’ve ever played a 61-card deck in a tournament, but I do completely agree with the reasoning that, while you know it isn’t optimal, you’re not sure which card you should cut. You think the danger of cutting the wrong one outweighs the advantage of having the smaller deck, so the best decision you can make based on the information you have at the time is to just play a 61st card. However, I’d only advocate making this concession with something like Reid Duke Grixis Control deck that didn’t play four-of any spells rather than something like Delver, Faeries, or Affinity that has some specific four-of that it really needs to draw.
In Limited, I think it’s right to play a 41st card somewhere between 0.1 and 1% of the time just to get your mana ratios right. That’s not very often, but it does come up. Also, there are specific archetypes and specific formats where playing a big deck is necessary for some reason (I did this multiple times in Innistrad Self-Mill), but that’s even more rare.
@TheBoozeCube suggested:
If It’s Good in Sealed, It’s Good in Legacy
This is a huge stretch, obviously, and I don’t have a lot to say about it, but I do routinely build Constructed decks based on interactions I learned in Limited, most recently the Diabolic Revelation / Elixir of Immortality engine I worked on. It’s not very likely that this would ever happen for Legacy, but it’s worth thinking about Limited decks/cards/interactions you’ve been impressed by if you’re trying to build a new Constructed deck.
@Kirblar024 suggested:
“Understanding Tiebreaker Math” by Todd Anderson
Good beats aside, this is actually something a lot of people don’t know and often need to know. I’m not Todd Anderson, so I can’t help with that part, but I can talk a bit about looking at a standings sheet and figuring out what to do.
I think a lot of people are intimidated by the idea of “tiebreaker math” and they hear a lot of stories about people doing it wrong, so they think they need to find a friend who’s “good at tiebreaker math” to ask for help.
I’ll let you in on a secret: for the most part, understanding your tiebreakers doesn’t really involve a lot of what most of you would probably think of as “math.”
For the most part, it’s just logically thinking through a pretty simple process.
When I’m in a situation where I might want to draw, I try to take a picture of the standings sheet on my phone as quickly as possible so that I can spend a minute with the data and potentially explain the situation to my opponent. This means I have some examples to discuss in the pictures saved on my phone.
In the last round of Grand Prix Boston-Worcester, I was in 48th place, so the questions were, “If I win, will I finish in the Top 32?”; “If I lose, will I finish in the Top 64?”; and “If I draw, will I finish in the Top 64?”
Sometimes you’ll find yourself in a position where winning and drawing give you the same prizes but losing costs you. In this situation, it seems pretty bad not to just draw. This is why it’s important to know how to do this “math.”
So when I was in 48th place, I had 33 match points, and my tiebreakers were 75.18%
I had the best tiebreakers of anyone with 33 points. Places 48-100 had 33 points, and 101-108 had 31 points. The highest tiebreaker of anyone with 30 points was 73.79%. This meant that even if he won and I lost, the player in 109th with 30 points couldn’t catch me, so I didn’t have to think about him or anyone lower than him.
The most points I could get by winning was 36. This meant I didn’t have to think about passing anyone who already had 37 points or 36 with much better tiebreakers than me. The top player with 36 points had a 72.51% tiebreaker, which meant that I could pass any player with 36 points if I won and they lost. The top player with 36 points was in 13th place.
Any time two people with 36 points drew, they would both go to 37, and I couldn’t pass them. But any time they played, if I won I would pass one of them. This is what makes the math tricky. You have to consider your best possible case, worst case, and expected case.
The best case for me would be if they all played and I won. So if everyone played, half of the people with 36-34 points would gain three points, and I couldn’t catch them. That would be half of 13th place-47th place. There were 35 people there, so the best case would be if eighteen of them lost, I won, and I moved up eighteen places to finish in 30th. If one match drew or if the player who got paired up won, I’d be 31st, and if two more matches drew, I’d be 33rd, which meant I wouldn’t get Top 32 prize. So if I was hoping to Top 32, I really needed to make sure there weren’t too many people who should obviously draw. The worst-case scenario if I played would be if all the 36s drew, all the 34s played, and I ended in 43rd place.
If I drew, I’d have 34 points. At this point, there’s a useful shortcut: drawing is horrible when you have the best tiebreakers because you only pass people who also have draws and your number of points, which isn’t very many. Conversely, drawing is generally pretty good when your tiebreakers are horrible. But if I drew and everyone else played, half of the people below me who could pass me would, which would mean half the people between 49th and 108th pass me, which would mean I’d drop 30 places and finish out of the money. So my decision here was obvious: I had to play, and if I won, I had to hope most other people also had to play.
This is the same principle that you use any time you need to look at tiebreakers. Basically it involves a lot of “I can pass half of this subset of people” and “half of these people can pass me.”
@TRenewer suggested:
Why You Should Choose to Draw When Playing Aggro Decks
Based on the exact phrasing, I don’t think this was intended as a beat on me for my decision to draw in the Top 4 of Pro Tour Philadelphia against Josh Utter-Leyton since I wasn’t exactly playing an aggro deck, but that is a decision a lot of people disagree with and can’t imagine the reasoning for.
There are a lot of different kinds of aggro decks that match up against each other in very different ways. In the extreme case, one might be all creatures with shadow and the other deck all vanilla creatures, the two decks have no way to interact with each other, they’re just racing, and the player who is on the play is at a huge advantage.
On the opposite end of the spectrum, both players’ decks are just lands and vanilla 3/1s, so creatures can never do anything other than trade one for one or hit a player. Whoever has the last 3/1 standing will win, so you just want to have an extra card to increase the odds that you’re that player.
The most often case that comes up in Limited that mirrors the first case is U/W flier aggro against big ground creature aggro, where being on the play is huge.
The case that most often resembles the second one is a R/B mirror where both players have a lot of removal, some creatures, and very little card advantage. While both decks might generally be aggressive in other matchups, against each other all their cards just trade, and they should choose draw.
My match against Josh wasn’t exactly like either of these cases, but I decided it was pretty similar to the second case. In order for me to win, I had to assemble my combo and match my counterspells against his removal spells. I needed to be able to cast as many counter-answers as he could cast answers to my combo. He was going to maximizing his answers, which meant he’d be siding out threats. Also, he couldn’t tap out to play threats because he needed to keep mana open to stop me from winning on the spot. This meant he couldn’t really pressure me, so the game wasn’t much of a race.
To account for this, I decided to side out lands, decreasing my already extremely low land count even further, to maximize my ability to compete in this attrition battle. This involved going to such a low land count that I needed an eighth card to reliably be able to draw enough lands to play my spells, and even if I did have lands and spells, I still felt like I just needed another spell to end up with the best of the inevitable one for one trades.
This is an article that could go on forever, but my teammates have gotten back from lunch, so I think it’s time for me to get back to working on Modern.
Thanks for reading,
Sam
@samuelhblack on Twitter