The paperstrip method reloaded

by wickeddarkman on 03 December 2019

Main Deck (60 cards)

Sideboard (15 cards)

Instants (1)

Artifacts (3)

Submit a list of cards below to bulk import them all into your sideboard. Post one card per line using a format like "4x Birds of Paradise" or "1 Blaze", you can even enter just the card name by itself like "Wrath of God" for single cards.

Deck Description

Is a way to use paperstrips and your brain as a supercomputer!
You will need no skills in ordinary programming, as well as no mathematical skills (unless you consider +1 and -1 to be too heavy for you)

Currently the method takes time, but being a computational construct it should be subject to Moores law, meaning it should double in processing power every second year, and I'm proud to state that it's been holding so far.

The average game takes 36 minutes when all the paperwork is accounted for, and I usually run 200 games per run, which translates into 120 hours or 5 days doing nothing else. I am alone in doing this, but I am certain a team of two could work faster.

The power of paper as a tool is more or less forgotten these days with computers everywhere, after all computers have the processing edge because we forgot to expand our ideas on how paper can be used.

This might change when enough people generates papercomputer like the one I'm about to teach you how to use, because the concept isn't tied to mtg, it can be used on ordinary cards as well with some work, which means you will be able to develope a powerful poker playing computer to have at your side while you play online. If you are also an app developer you could build an app that supports your play immediately, making you able to compete against top level poker players.
Contact me if you need help starting a system up, and I'll give you enough ideas on how to do it.

But before that is gonna happen you will need to understand how I use it to create magic-decks for me.

A team of friends will enable you to work faster, and will be useful when you need innovation. It's possible to use paperstrips alone, but then you have to work at autism level, preferably with OCD tendencies.

Testdecks as close to your meta as possible, the more accurate, the better.
If you can make proxies of everything it will be a good idea to keep cards safe. The paperstrip method can actually wear cards so much that they lose value. You wont believe how much damage sliding a paperstrip into a sleeve can cause after 1000 games.

Patience or curiosity will take you through the more boring parts of the method, but if you create too many side-rules that you feel like a slave, cut yourself some slack and try to let intuition be your only rule, except for the alpha rule. The more fun you are having, the faster you will get results.

The method uses a point based system to evaluate cards. You will reward cards with a point when they're good, and you remove a point when they are bad. In addition to this, you will use what I call the alpha rule. The cards with most points will need to be rewarded most often, while cards with few points must be neglected more.
You must play the deck as you would play it normally, but in any situation where you are in doubt, or if you can choose between two copies of a card, you must cast the card with most points and reward it with a point.

Here's an example. In the above deck, there is 4 moxen. Since the meta is prepared for them, they are less valuable than lands, except at the early game and against land destruction. So my reward system for the moxen are: a mox not played or used during a game gets a useless-point. If used solely for colourless mana it gets a colourless point, and in the case where it is used for its colour during play it gets a coloured point. Here's the score for each moxen at the moment.

Jet: 3 useless, 5 colourless, 10 coloured.
Pearl: 5 useless, 6 colourless, 9 coloured.
Ruby: 3 useless, 18 colourless, o coloured.
Sapphire: 6 useless 8 colourless, 6 coloured.

The high number of colourless points on ruby is because of the alpha rule. Since there are no red spells in the deck ruby was only used for colourless mana, so every time I had ruby and another mox in hand, ruby was played first if I only needed a mox for colourless.

Mox sapphire is also used for colourless and has most points in being useless, that means there is a potential to swap it with mox emerald to build the manabase towards green as well.

There are many ways to build up the reward/punishment system, so here is my key guidelines.

During games cards can only gain points once unless they are somehow re-cast.
If a spell has impact on the game it gets a point.
If a spell wasn't cast during game and is in the finishing hand it loses a point.
Towards the end of the game you will be eager to cast spells so they wont lose points, so you must be careful of what you deem as useful. For example it isn't useful to cast a land destruct spell just before you kill them while they have 12 lands.
A spell being countered or destroyed makes the opponent use a card, so a point is rewarded.
A spell discarded gains a point if it got chosen by the opponent, while it loses a point if you choose what was discarded. Random discard and all-discard is neither rewarded or punished.
To keep track if a permanent is used, place a counter on it, and when the card impacts the game, remove the counter and reward the card with a point and continue the game. At the end of the game all cards with these counters on lose a point.

During games it is possible to use strips as counters, and you can specialize these strips. In the example with moxen I use a coloured strip a colourless strip and a dice. The dice Mark's that the mox have had no impact on the game, but as soon as it is used for mana it gets one of those strips. A coloured strip removes the colourless strip. I use a similare system to evaluate what my lands become,
And this is what build my manabase for me. Due to the alpha rule, a land will strive towards becoming either a basic land or a dual land if you set it up well enough.

There are virtually no limits to how you set up this part of the method, so I can only recommend that you set up rules you think will work the best in the long run, but keep the rules the same for each version of the deck you want to expand, and only change rules between play tests, otherwise you will get data distortion, and basically have to start over. If you discover that you've made a fault, scrap the data from that run, but before you start over you could try to rethink your current rules and see if you can innovate something so it wasn't all for nothing. Don't forget to let your intuition have some space. In theory the whole method could be run by points and intuition.

You are about to have your first measurement run, so be sure to have some rules on paper before you start. Your first run should be done to see how the method works to mark out bad cards within your deck, so don't worry too much about precision. You need to battle a single testdeck 5 times. Usually I run against 20 testdecks and at least 10 games against each, but you shouldn't have such a large run to start with.
During those 5 games you must play as close to the rules of how you reward/punish as possible so cards generate or lose points. These games must all be made without sideboarding.

After you played all games you must divide the entire deck into piles based on how many points each card have, and remember it's possible for cards to get negative points. Now you should be able to see what cards were most effective during the games. This gives you valuable information. (I usually write down how many points individual cards gains against each testdeck, that way I can see what can be removed when sideboarding against that deck)

Now you remove the 9 worst cards or more if some cards scored equally bad
Remove the proxies and strips from these and insert blank paperstrips. I refer to these as "the 9" and they are to become new cards during the innovation phase. Now you must reset all other strips to 0. Everything that is not part of the 9 is your new "core" your deckcore should be written down so you can go back to previous designs if you somehow mess up the method.

At any time you can replace any of the 9 with anything you want and probably have an improved deck already, just remember that the more games you play against as many testdecks as you can, the better the measurements will be.

This is where the fun begins and things get a little more complex.
During the games against your testdecks you were probably clever enough to write down your wins and losses and which of the enemy cards were causing you trouble. If you didn't just do it next time.

Now you must play against the testdecks in order with the most difficult decks first.
If you scored equal wins in against some, try to look at past notes or just decide which you think is the worst matchup.

During these games the 9 blank strips have the potential to become everything you want them to be, and when you play a strip as a spell write the name of it on the strip and give it a point. Each time in the future you play that strip as that card, put a point close to the name. If you for any reason want to play the strip as something else, write that card name and give it a point too. When you play enough games there will be a clear pattern to what you want to play, just keep yourself within the 4 off each card limit if it applies to the card. Each strip has enough space for 8 card names and their points, so knock yourself out and don't worry about the fact that your deck gains an advantage at top decking whenever you draw or use one of the 9 strips.

During all of this you also re-measure the old cards gaining more data on those cards, getting an idea of what should go the next time.

When you are done with all the measuring you simply look at what cards gained most points in all. The best 9 cards can now be put into the deck, and the next 15 best results (if you used that many cards) becomes your sideboard. I currently use between 54 to 63 card names on these strips, and it's possible to double the number by also using the backside of a strip. I usually run 10 games against 200 decks which means I get a picture of what the deck needs through 200 games. This usually gives me a clear answer to what the deck needs for both the 9 new main board cards as well as the 15 sideboard cards. If you take notes on what gained points against each deck you will be able to use your data on what was worst against a deck and you will know both what to sideboard out as well as what to board in.

This is the basis of the paperstrip method, I'm working on a process to merge the measuring phases with the innovation phases which will cut the number of games into half, doubling the processing capability and again keep up with Moore's law.

Since this is a programmable tool there is virtually no limit to how you can "program" the paperstrip method. You have the overall rules of the reward/punishment pointsystem and you have the strips themselves. Some of my oldest strips only had two modes, cut and keep and cards gained points in these sections and I would subtract cuts from keeps to get the value of a card. These days I use several modes on strips to evalute them, lands and moxen have coloured, colourless, and useless as a guide system while library of Alexandria has a lot of small rooms where I can write how many cards it drew during games and what interrupted it. It's rooms are large enough for me to write its use during all 200 games.

I sometimes use a variant of the strip system where the cards are given specific categories, like draw, burn, creatures, mill or whatever I want to focus on. The 9 blank strips are then split up into two or more categories and I can then gain data on how a deck should be overall. This way you can optimise the number of cards you need to support for example an advanced combo and how much of the deck needs to be mana or defence.

If anyone wants to know more about past strip designs, or how I would make a strip for specific cards you just write me here, and I'll answer within 14 days.

Deck Tags

  • paperstrips
  • Article Deck
  • Deckbuildingtool
  • Any format
  • Competitive
  • wickeddarkman
  • DK

Deck at a Glance

Social Stats


This deck has been viewed 377 times.

Mana Curve

Mana Symbol Occurrence


Deck Format


NOTE: Set by owner when deck was made.

Card Legality

  • Not Legal in Standard
  • Not Legal in Modern
  • Not Legal in Vintage
  • Not Legal in Legacy

Deck discussion for The paperstrip method reloaded

I will make a deck based on this do you have any recomendations?

Posted 29 April 2020 at 21:28


Burn or combo. The faster the deck is, the less time the whole process takes.
At any time as questions, this can be done in so many ways.

If you don't want to start from scratch, you could use your radios deck.

Posted 29 April 2020 at 22:36


Do you mean Rakdos? Also might sacrifice how effective it is to get results faster

Posted 29 April 2020 at 23:55


Damn autocorrect. Yeah, rakdos.
And yes you can "cheat" with the system, but perfection takes time, so you might get weak results.
I'm pretty rigid in my own behaviour, letting evolution take 100% control of it, but no matter what % you set the point system is really what's important.

Also try to be as omnipotent as possible. Look several turns ahead by looking at the top of decks, this will make the deck give you what you need almost every time mechanically. If the opponent attacks and makes a horrible mistake take the game back to before the mistake. The better you play the enemy decks, the more resistant your own deck will become.

Posted 30 April 2020 at 04:54


bookmarked to read later to much meat here for my sleep deprived brain

Posted 29 April 2020 at 21:33


Go for it :)

Posted 29 April 2020 at 22:37


=0 when AI starts writing AI scripts and carrying them out, scarry stuff! but a great read thanks for writing this up and sharing it. dark you ever play anarchy online?

Posted 30 April 2020 at 11:23


It's so different in its builds.
I never play online, I started paperstrips when my computer died on me and it kept evolving.
I have a phone now but waste time on strips instead of games :)

Posted 30 April 2020 at 16:52