Hot take: Most of the time when people complain about the meta being uninteractive, they are just bad at combat maths. Or assigning roles based on matchup and game state
A card gets N points for every element of the opponent's gamestate it can target or affect (i.e. interact with). So a card like Pyretic Ritual gets 0 points because it purely contributes to your gameplan. Same with Gifts Ungiven because it only effects your resources, despite targeting an opponent.
Push gets 1 point because it can target and destroy creatures. Bolt gets 3 because it targets players, walkers, and creatures. Death's Shadow gets 3 points because it can attack players and walkers and also block creatures. Ooze gets 4 points becauae it does everything Shadow can do plus target cards in a GY.
So a list of 4 Bolt, 3 Push, and 3 Ooze gets 27 points. A list of 4 DRit, 4 PyRit, and 4 Gifts gets 0 points.
There are lots of ways to change the scoring system to reflect different nuances of interaction. But I'm confident if we can agree and improve on such a system, we would reach an acceptable, accurate, and largely objective measure of interactivity.
so...what about discard or counterspells? do they count every possible card type they can hit? cryptic command should be the highest value a card can reach if that is the case. unless there is something that can hit every card type and the player im not thinking of.
snapcaster would be a 3 i guess, though giving it the same rating as ambush viper seems odd. modal cards would have high values, including walkers. liliana uptick would have to be rated as interacting with the opponents hand, the same goes for any other forced discard.
pump spells would be zero i guess, and mill cards interact with the opponents deck.
based on this type of system im gonna predict taking turns is one of, if not the most, interactive decks in modern.
Private Mod Note
():
Rollback Post to RevisionRollBack
Modern: UWGSnow-Bant Control BURGrixis Death's Shadow GWBCoCo Elves WCDeath and Taxes (sold)
Let me see if I can get a feel for how people in this thread would react to various 'feelsbad' scenarios. My hunch is that people often look back at 'feelsbad' moments and let those experiences influence their perception of a deck. This is gonna get a little wordy, I'll do my best to provide a tldr at the end.
A) Player 1 successfully casts Mistbind Clique during Player 2's upkeep and champions Vendillion Clique. This results in tapping all of Player 2's lands, and making it very difficult to make much out of his/her turn. In this scenario, how much 'feelsbad' does Player 2 experience?
B) Player 1 is at 6 life. He/she casts two Death's Shadow, and already has a Gurmag Angler on the battlefield. Player 1 has one Island and one Watery Grave left untapped. Player 2 has Tron, Sanctum of Ugin, and a Forest. The next turn, Player 2 casts Ugin, the Spirit Dragon, but Player 1 responds with Stubborn Denial. How much 'feelsbad' does Player 2 feel? More or less than Scenario A?
C) Same as Scenario B, except Player 2 casts Ulamog, the Ceaseless Hunger, and Player 1 does not have a response. How much 'feelsbad' does Player 1 feel? More or less than Scenario B and/or A?
In any of these scenarios, would any of you feel that one player's greivance has more weight than the others? Do 'feelsbad' moments take away from the legitimacy of a deck?
These scenarios are oversimplified, but I feel that a lot of players have experienced something similar. It never feels great to have the tide suddenly turn against you in a game, but it is interesting how players respond to how the tide turns. Scenarios B and C both appear to be plays that decide how a game ends, but Scenario A doesn't appear to look that way. However, not all decks are looking to blow the opponent out in one swift move. Some decks build an incremental advantage that doesn't look like much, but ultimately wins games.
Let me see if I can get a feel for how people in this thread would react to various 'feelsbad' scenarios. My hunch is that people often look back at 'feelsbad' moments and let those experiences influence their perception of a deck. This is gonna get a little wordy, I'll do my best to provide a tldr at the end.
A) Player 1 successfully casts Mistbind Clique during Player 2's upkeep and champions Vendillion Clique. This results in tapping all of Player 2's lands, and making it very difficult to make much out of his/her turn. In this scenario, how much 'feelsbad' does Player 2 experience?
B) Player 1 is at 6 life. He/she casts two Death's Shadow, and already has a Gurmag Angler on the battlefield. Player 1 has one Island and one Watery Grave left untapped. Player 2 has Tron, Sanctum of Ugin, and a Forest. The next turn, Player 2 casts Ugin, the Spirit Dragon, but Player 1 responds with Stubborn Denial. How much 'feelsbad' does Player 2 feel? More or less than Scenario A?
C) Same as Scenario B, except Player 2 casts Ulamog, the Ceaseless Hunger, and Player 1 does not have a response. How much 'feelsbad' does Player 1 feel? More or less than Scenario B and/or A?
In any of these scenarios, would any of you feel that one player's greivance has more weight than the others? Do 'feelsbad' moments take away from the legitimacy of a deck?
These scenarios are oversimplified, but I feel that a lot of players have experienced something similar. It never feels great to have the tide suddenly turn against you in a game, but it is interesting how players respond to how the tide turns. Scenarios B and C both appear to be plays that decide how a game ends, but Scenario A doesn't appear to look that way. However, not all decks are looking to blow the opponent out in one swift move. Some decks build an incremental advantage that doesn't look like much, but ultimately wins games.
if people count a match, game, or sequence not going their way or have a bad outcome as a feelbad then that is nonsense. of course there are negative feelings involved. no one likes losing or flooding on lands when the opponent is peeling perfects.
i think there is a stark difference between acknowledging that some play patterns elicit a negative reaction and whether that negative reaction is a strike against the deck. for example i hate ponza, i just dont enjoy playing against it. however that isnt saying i want the deck gone. if people get their thrills from stone raining people then thats cool. its not like the deck is too good.
Private Mod Note
():
Rollback Post to RevisionRollBack
Modern: UWGSnow-Bant Control BURGrixis Death's Shadow GWBCoCo Elves WCDeath and Taxes (sold)
@ktkenshinx, I think the solution might involve considering what decks use as pre-emptive solutions to prune branches off of an opponent's decision tree as well.
Since it's already established that a (competitive) deck is designed to accomplish three things:
Minimize the amount and/or significance of interaction the opponent can have on the gamestate
Maximize the amount and/or significance of its own interaction
Maintain as much consistency as possible in that minimization and maximization
We can see how all decks accomplish these things.
Bogles does so by using hexproof creatures and effectively making their creature threats much larger and/or resilient than the opponent's. That would be pre-emptive interaction with the opponent's decision tree.
That actually leads us to what I think may be the root of the misguided use of the word "interaction". People see the cards on the battlefield, in hand, or on the stack, and focus on that. But a deck can best interact by being more focused on an opponent's decision tree, which isn't a physical thing that we can see. People are very vision-oriented, so this isn't something that is as easily perceived and understood, I guess.
So rather than scoring cards on how they interact with other cards in the visual sense, it may be better to score them by how well they interact with another deck's decision tree.
It seems like it would be a huge headache going through so many cards and giving them a grade and point scale. It could be done, but almost no one has the time to do that.
It seems weird giving Death's Shadow such a high rating, as players would sometimes feel like it was almost a combo card in itself.
Something like Snapcaster should get about as many interaction points as possible, as the potential is so high.
I mean, technically something like Burning inquiry is really high on interaction, but if your opponent suddenly drops 2 4/4 Hollow Ones before you even make a land drop, how interactive and fair is that?
There's this weird, intrinsic interaction and fair scale, too.
Tron's blowout cards are super interactive, but they can be deployed so fast that it doesn't feel fair
Hot take: Most of the time when people complain about the meta being uninteractive, they are just bad at combat maths. Or assigning roles based on matchup and game state
This is probably the most true comment I've read. Having played both hollow one and affinity I'm not sure how someone can say affinity is anymore interactive then hallow one. I'm guessing they've never played either. Stock hallow one list usually run 6 pieces of interaction (collective brutality and lightning bolt) while affinity runs 4 pieces usually (playset of galvanic blast). Simple math states it has more interaction.
This conversation about interactiveness is much more interesting and unexplored than the tired ban and unban circle. I want to revisit an earlier analysis I made about evaluating interactivity based on cards targeting and interacting with opposing decks. A card like Bolt would get 3 points on this scale (1 each for targeting players, walkers, and creatures). Decay would get 4 for all the permanent types it can hit. Creatures would get points based on ability to attack, block, and any other interaction modes built in (e.g. Lavamancer). I think if you evaluated cards and decks on that axis you would have a pretty solid system to start quantifying interactivity.
I tried this before with solid first results, but ran into issues evaluating decks like Burn with tons of targeting options. Any ideas on this approach or approaches of your own?
I'm not sure I understand your method here. Could you walk us through something like a counterspell or discard spell?
A card gets N points for every element of the opponent's gamestate it can target or affect (i.e. interact with). So a card like Pyretic Ritual gets 0 points because it purely contributes to your gameplan. Same with Gifts Ungiven because it only effects your resources, despite targeting an opponent.
Push gets 1 point because it can target and destroy creatures. Bolt gets 3 because it targets players, walkers, and creatures. Death's Shadow gets 3 points because it can attack players and walkers and also block creatures. Ooze gets 4 points becauae it does everything Shadow can do plus target cards in a GY.
So a list of 4 Bolt, 3 Push, and 3 Ooze gets 27 points. A list of 4 DRit, 4 PyRit, and 4 Gifts gets 0 points.
There are lots of ways to change the scoring system to reflect different nuances of interaction. But I'm confident if we can agree and improve on such a system, we would reach an acceptable, accurate, and largely objective measure of interactivity.
I remember your method. It's smart. This method though runs the obvious problem of being a manual one, kind of misleading some times and quite subjective and must take each case individually into account. But if we all agree to those points, it could be great.
Just an observation. Do creatures have to get that many points? I mean, this way, Abzan Company gets many points, while it's not that interactive at the end of the day. Unless interaction in the form of creatures is meant to be heavier interaction than a deck filled with hand disruption spells, counterspells and spot removals.
More specifically, Vizier of remedies is a creature that, just like Death's Shadow, can get 3 points because it can attack players and walkers and also block creatures. In reality though, it's used as a purely uninteractive combo, coupled with Devoted Druid and Walking Ballista. I am not sure it's that right to give 3 points here.
That said, how many points does a creature heavy deck like Zoo or Abzan Company has and in the other hand how many points does Jeskai Control get? I think this may lead to a flaw, because Jeskai control must be heavier on interaction that Zoo or Abzan Company. If it is nearly at the same levels, this could mean that a creature like Vizier of remedies or Death's Shadow should get less points, so that we can balance the situation.
Gkourou raises some important points I think in this discussion, the level of interactiveness of a card is also contextual. In Burn Lightning Bolt can definitely hit all three, but something like 90%-95% of the time it just hits face.
In Zoo style decks, creatures can definitely block and/or attack planeswalkers but most likely they will go face. In creature based combo decks, most creature will just sit around looking at the board around them until it's time for them to fulfill their part of the combo.
I agree with what others pointed out, that it would be a huge amount of work, but I don't think it would be unmanageable. It could easily even be crowdsourced if you create the correct platform. Something like someone setting up a platform where people can register decks, rate the cards and get a result, and then the results could be compared. That would be interesting and could be realized within a relatively small period of time if enough people from the forum contribute. However, I wouldn't know how to create such a platform, my programming skills definitely rate 0 in the interactive spectrum.
That being said, the contextual problem will remain, and it would need to be clarified that we would not be looking to contextualize cards in their interactivness. This means that Grapeshot still gets a 3 (face, creature, PW), even though it usually just goes face.
Ym1r the nice aspect to that post is that it points out that just because a deck has a linear gameplan does not mean it is simple or easy to pilot because there are still loads of decisions to make. You weren't taking it that far I realize, but part of the "linear vs interactive" discussion does generally involve the claim that games between the latter are more complex and interesting. At a high-level of proficiency, burn is a difficult deck to play, in large part because you have to get over the hump of understanding that smaller percentage of scenarios where "bolt to face" is not automatically the best play.
Interactive vs non-interactive, in terms of casual conversation about Modern, is like the game Road Rash (any of you remember that old Sega Genesis game?).
A non-interactive deck is like a racer who focuses on getting the fastest bike with the best parts and has a game-plan of just cruising past the competition without getting into any scuffles.
An interactive deck is like the racer who brings a crowbar to the race to beat the tar out of other racers.
Lots of really interesting ideas and conversation here. Rather than create a giant quote wall, I'm just going to respond to topics:
Re: Complexity/ease of the grading method
I don't think the method has to be easily implemented to be viable. If the method happened to be time consuming but still produced accurate results, that's fine with me. Metagame compilations were time consuming but were quite accurate for what they claimed to be doing.
Re: subjectivity/objectivity of the grading method
This method should absolutely not be subjective at all. There should be clear, objective measures of when cards get interaction points and they should be consistently applied. For instance, it is 100% objective that Bolt can target three elements of an opponent's gamestate (board, walkers, creatures). We start getting muddier when something like Shadow having 3 modes (attack player, walker, and block) gets reduced because it's "effectively" just a dumb beater, or that Vizier gets reduced because it's effectively a combo piece.
Re: combo cards getting too many points
This should be accounted for by the other cards in the deck. For instance, Electromancer, Baral, and Grapeshot each get a bunch of points because they can attack, block, and target different stuff respectively. So an average Storm deck would get about 30 points for that pile of cards (3 points for each copy of the cards). But the rest of Storm would see a huge dropoff in points: Manamorphose, Gifts, Dritual, PRitual, PiF, SV, Sleight, etc. all contribute literally 0 points to the deck. This should lead to combination decks having far fewer points than non-combination ones after a final calculus.
Re: the "Burn problem"
I like the idea of the synergy approach to this (i.e. diminishing returns on effects that are identical to existing effects), but I think this creates false positives/negatives for other decks. For example, a deck that was all counterspells might lose points here even though I'm sure we would all agree it is mostly interactive. I think scoring cards by the different types of permanents they can interact with might adjust for this over the long run. So Bolt can still get 3 points, but something like Mana Leak would get 6 points because it can hit all 6 non-land card types. Same with Thoughtseize/IoK. Something like Cryptic would get even more points because it has different modes that effect all types of cards including lands. This might equalize out the point spread in the long run.
Re: limiting decision trees
I'm not sure how this method would actually be scored with multiple decks, so maybe I'd need to see it to know whether it's appropriate or not. Does anyone have an example of how two random decks would score on it? For instance, Bogles (a known non-interactive deck) vs. Jeskai Control (a known interactive deck)?
Interactive vs non-interactive, in terms of casual conversation about Modern, is like the game Road Rash (any of you remember that old Sega Genesis game?).
This all presupposes that every form of interaction is always better than any un-interactive strategies.
No, no it does not. You can add your own biases to the conclusion as you please, but the method proposed by KTK does not attempt to measure "good", "bad", or "better". Just interaction.
Now certainly there are people who will strenuously argue that interactive is good and uninteractive is bad (and probably some vice versa) but that's a value judgment. If there were an objective classification such as KTK's scoring system, we could at least agree on the definition were using when discussing how a metagame sorts out across the spectrum of interaction.
I feel like some people are giving non interaction a bad name.
If the top players at my LGS are playing Lightning Bolt, why would I play a deck that folds to Lightning Bolt? If the players are playing a lot of creature removal in slower decks, why would I play a handful of slow creatures that lose to that? If players are playing super quick goldfish creature decks, why would I play something that durdles and dies on turns 3 and 4? Maybe that's fun for some people; not me.
I fear some of you will never have that feeling when your opponent has 2-3 cards in hand (presumably Lightning Bolt and Fatal Push, among others) when you're playing Bogles or Grishoalbrand. I don't want to give you a chance to interact with every single card in your hand. Hollow One and Gurmag Angler dodge Bolt/Push, so that's what gives them more resilience.
Private Mod Note
():
Rollback Post to RevisionRollBack
Legacy - Sneak Show, BR Reanimator, Miracles, UW Stoneblade
Premodern - Trix, RecSur, Enchantress, Reanimator, Elves https://www.facebook.com/groups/PremodernUSA/ Modern - Neobrand, Hogaak Vine, Elves
Standard - Mono Red (6-2 and 5-3 in 2 McQ)
Draft - (I wish I had more time for limited...)
Commander - Norin the Wary, Grimgrin, Adun Oakenshield (taking forever to build) (dead format for me)
This all presupposes that every form of interaction is always better than any un-interactive strategies.
No, no it does not. You can add your own biases to the conclusion as you please, but the method proposed by KTK does not attempt to measure "good", "bad", or "better". Just interaction.
Now certainly there are people who will strenuously argue that interactive is good and uninteractive is bad (and probably some vice versa) but that's a value judgment. If there were an objective classification such as KTK's scoring system, we could at least agree on the definition were using when discussing how a metagame sorts out across the spectrum of interaction.
What is the goal, then, of measuring interaction? I'm sorry if I missed it in the lat few pages.
This conversation about interactiveness is much more interesting and unexplored than the tired ban and unban circle. I want to revisit an earlier analysis I made about evaluating interactivity based on cards targeting and interacting with opposing decks. A card like Bolt would get 3 points on this scale (1 each for targeting players, walkers, and creatures). Decay would get 4 for all the permanent types it can hit. Creatures would get points based on ability to attack, block, and any other interaction modes built in (e.g. Lavamancer). I think if you evaluated cards and decks on that axis you would have a pretty solid system to start quantifying interactivity.
I tried this before with solid first results, but ran into issues evaluating decks like Burn with tons of targeting options. Any ideas on this approach or approaches of your own?
The reason you're running into issues is that you're using the wrong heuristic. A couple years ago I discussed this idea with you briefly. I've since moved on to other systems of trying to solve the game because of the data input this would involve but I think you want to look at this slightly different. It's not really about scoring cards as high/low but scoring cards at all. In the system I envision, you create a series of rules such as:
Trades at parity (no more than CMC + 1)
Turns off card
Trades or better in combat
And a few other rules (in the AI I have that plays Magic I have several rules like this)
Then, you rate card A vs card B in a matrix. If A passes all of the above rules against B it gets 1 point, if it fails any of them it gets a 0.
You can then sum the cards along their rows. The cards that pass the most cases, will naturally be the most interactive against a general meta. If you restrict the columns to include only certain cards, representing certain decks (or excluding certain decks) then you can make a more nuanced case about particular matchups. You can even weight all of this by meta percent to make card and deck building suggestions for an expected metagame.
This system could even identify new cards that don't yet see play by discovering interactions not currently being considered.
Given enough manpower I am 100% certain the system can work. The problem is that it has a very prohibitive data entry requirement and it scales exponentially, so as new cards are introduced to the format it becomes harder and harder to maintain. If we had a 10 card format it would require a 10x10 matrix for 100 points of data, a 100 card format is 100x100 or 10,000 points of data. The difference between a 400 and 500 card format (a 25% increase in the cards) 160,000 vs 250,000 pieces of data, a 56% growth in inputs required.
Currently I'm working with a small team (though the project is stalled until late may/early june) trying to implement this system in Legacy. Legacy currently has the smallest competitive cardpool,so it makes sense in that format. We've come to the conclusion though that it can only work if we write a system that can parse oracle text to determine function. Doing it manually is out of the question unless there's a very dedicated community to doing so. Based on what I've observed from your previous metagame updates before you quit doing it... the community to build and maintain these statistics doesn't exist.
The good news though, is that automatic card parsing SHOULD make the project apply to any existing or arbitrary format and the hope is that we would be able to build a database of decks with cards they may include, and then tweak to any specific meta.
This would result in an answer to a slightly different question from what you're asking though. You're asking about interactivity while my proposed answer is what lines up well. Sometimes two decks don't interact but still line up well against each other... Scapeshift vs Grapeshot for example.
Also, this approach would not be able to handle combo.
Here's my current rules for the similar approach I just outlined that I'm currently using in Legacy:
This example was rating Nissa, Vastwood Seer in Nic Fit (our goal was to rate finishers only) but you can get a general idea from this with the rules involved (this also applies to ktk's proposed system):
General criteria:
1 - 6 mana or less
1 - Card advantage
1 - Can GSZ (optional?)
I feel like some people are giving non interaction a bad name.
If the top players at my LGS are playing Lightning Bolt, why would I play a deck that folds to Lightning Bolt? If the players are playing a lot of creature removal in slower decks, why would I play a handful of slow creatures that lose to that? If players are playing super quick goldfish creature decks, why would I play something that durdles and dies on turns 3 and 4? Maybe that's fun for some people; not me.
I fear some of you will never have that feeling when your opponent has 2-3 cards in hand (presumably Lightning Bolt and Fatal Push, among others) when you're playing Bogles or Grishoalbrand. I don't want to give you a chance to interact with every single card in your hand. Hollow One and Gurmag Angler dodge Bolt/Push, so that's what gives them more resilience.
There is no issue in playing a non-interactive deck. People may label or assign negative value to that label, but it just is a label.
Bogles is not interactive, its quite literally (Game 1) the most non-interactive deck you can imagine. Thats simply fact.
That doesnt mean its a bad thing, or a poor choice, or without logical purpose behind running the deck.
Simply dont try and call it interactive, as in the colloquial term used by many (most?) Magic players.
@Aazadan that system aligns more with thnkr's definition of interaction. where any quality that makes a card or deck 'good' is a form of interaction. more interaction in a deck makes it better, therefore the best decks are the most interactive.
the difference is that you are attempting to solve a format given a card pool and a metagame. whereas this thread, and what wizards cares about, are gameplay considerations which go beyond winning or losing to include more abstract elements like fun or enjoyment.
some people are obviously biased, but in the context of format health its about breadth of gameplay aimed at maximizing the enjoyment for the most players possible. if one set of play patterns is overly represented that generally isnt a good sign. a format defined by nothing but midrange/control grindfests is just as bad as nothing but overtly linear strategies in aggro/combo.
the problem is that most of this is largely subjective, and ultimately depends on the vision that wizards has for the game or format since they are the arbiters. their goal is obviously attracting and keeping players engrossed in the game for as long as possible, while also balancing the competitive aspect to ensure that skill and decision making are large enough factors in success.
Private Mod Note
():
Rollback Post to RevisionRollBack
Modern: UWGSnow-Bant Control BURGrixis Death's Shadow GWBCoCo Elves WCDeath and Taxes (sold)
This all presupposes that every form of interaction is always better than any un-interactive strategies.
No, no it does not. You can add your own biases to the conclusion as you please, but the method proposed by KTK does not attempt to measure "good", "bad", or "better". Just interaction.
Now certainly there are people who will strenuously argue that interactive is good and uninteractive is bad (and probably some vice versa) but that's a value judgment. If there were an objective classification such as KTK's scoring system, we could at least agree on the definition were using when discussing how a metagame sorts out across the spectrum of interaction.
What is the goal, then, of measuring interaction? I'm sorry if I missed it in the lat few pages.
To be fair, I don't think it's been actually stated. The goal of measuring interaction with a clearly defined model is an objective measurement for each deck. That can then be used to roughly compare decks, evaluate the overall mix of a metagame across time, etc. At this point we're stuck with "I know it when I see it" evaluations of interaction and that quickly bogs down in arguments over the trees rather than taking in the forest.
There are a lot of second order applications from there. For example, determining which type of metagames WoTC has historically found acceptable and then using it as a predictor for future bans/unbans. Or evaluating how metagames cycle (or don't) between interactivity and uninteractivity.
@Aazadan & ktkenshinx, That method is actually exactly what was used for the development of Lantern. You can see me explaining it here.
EDIT: While I'm at it, to avoid double-posting, working on Lantern is also what I feel helped me come to understand the game better, to include having a better grasp of what it truly means for a deck to interact or have linear characteristics, and why decks are best designed to minimize the opponent's ability to interact effectively, maximize it's own ability to interact effectively, and do those two things consistently.
I feel that we can understand why people currently use those terms when they do by understanding human behavior. As I mentioned before, our species is heavily influenced by what we are able to directly observe - Physical objects and their movement. When someone plays a card in this game that drastically reduces the number of branches on an opponent's decision tree, we don't get to see those branches get pruned. We see cards moving around some, and that's about it. But if we can see past the movement of those cards, and use our minds to perceive the invisible changes in respective decision trees, I feel that we can come to a deeper understanding of the game.
So our dilemma, when we want to understand the game better, and when we want to talk with others about understanding the game better, is that we must be able to overcome our own human instinct to focus only on what we can see and we have to avoid using (as someone recently best described it) dog whistle terms for criticizing what we don't like, usually due to some feeling of entitlement for what we might consider fair and just.
@Aazadan that system aligns more with thnkr's definition of interaction. where any quality that makes a card or deck 'good' is a form of interaction. more interaction in a deck makes it better, therefore the best decks are the most interactive.
Not necessarily. There's a distinction to be made between interacts and follows the rules. Slippery Bogle beats Fatal Push but I wouldn't say that makes Bogle an interactive card. It just means that it's positioned well against it and that's what the system I outlined is designed to measure. Interaction is a subjective term because highly interactive could mean fun or unfun. Lantern is very interactive, Bogles is not, so fun clearly transcends interactivity. Furthermore there exists good decks on both ends of the spectrum. I'm not attempting to say interaction is good or bad and instead looking only at positionings of groups of cards against other groups of cards.
The idea of such a system is to solve a metagame quickly, test what should be banned, test what should be unbanned, try new cards in old decks, and figure out what is currently positioned well, and what is not positioned well in a non subjective manner. Like I said though, I don't believe the community is capable of providing the required data set here (and most of the data needs to be filled in before any results can be reliable) so I've been looking at automated methods of doing this. I've made some progress on that front, but life obligations currently prevent me from working on it further for a few weeks minimum.
So are you saying that the design of the Bogles deck isn't to pre-emptively interact with a potential opponent's decision tree, to reduce as many possible branches as possible?
EDIT: @tronix I think that's a misconstrued version of what I'm saying. I'm not saying any quality that makes a card or deck "good" is a form of interaction. I think it would be more precise to say that what reduces the amount of interaction that a potential opponent might have is what makes it "good".
so...what about discard or counterspells? do they count every possible card type they can hit? cryptic command should be the highest value a card can reach if that is the case. unless there is something that can hit every card type and the player im not thinking of.
snapcaster would be a 3 i guess, though giving it the same rating as ambush viper seems odd. modal cards would have high values, including walkers. liliana uptick would have to be rated as interacting with the opponents hand, the same goes for any other forced discard.
pump spells would be zero i guess, and mill cards interact with the opponents deck.
based on this type of system im gonna predict taking turns is one of, if not the most, interactive decks in modern.
UWGSnow-Bant Control
BURGrixis Death's Shadow
GWBCoCo Elves
WCDeath and Taxes(sold)A) Player 1 successfully casts Mistbind Clique during Player 2's upkeep and champions Vendillion Clique. This results in tapping all of Player 2's lands, and making it very difficult to make much out of his/her turn. In this scenario, how much 'feelsbad' does Player 2 experience?
B) Player 1 is at 6 life. He/she casts two Death's Shadow, and already has a Gurmag Angler on the battlefield. Player 1 has one Island and one Watery Grave left untapped. Player 2 has Tron, Sanctum of Ugin, and a Forest. The next turn, Player 2 casts Ugin, the Spirit Dragon, but Player 1 responds with Stubborn Denial. How much 'feelsbad' does Player 2 feel? More or less than Scenario A?
C) Same as Scenario B, except Player 2 casts Ulamog, the Ceaseless Hunger, and Player 1 does not have a response. How much 'feelsbad' does Player 1 feel? More or less than Scenario B and/or A?
In any of these scenarios, would any of you feel that one player's greivance has more weight than the others? Do 'feelsbad' moments take away from the legitimacy of a deck?
These scenarios are oversimplified, but I feel that a lot of players have experienced something similar. It never feels great to have the tide suddenly turn against you in a game, but it is interesting how players respond to how the tide turns. Scenarios B and C both appear to be plays that decide how a game ends, but Scenario A doesn't appear to look that way. However, not all decks are looking to blow the opponent out in one swift move. Some decks build an incremental advantage that doesn't look like much, but ultimately wins games.
if people count a match, game, or sequence not going their way or have a bad outcome as a feelbad then that is nonsense. of course there are negative feelings involved. no one likes losing or flooding on lands when the opponent is peeling perfects.
i think there is a stark difference between acknowledging that some play patterns elicit a negative reaction and whether that negative reaction is a strike against the deck. for example i hate ponza, i just dont enjoy playing against it. however that isnt saying i want the deck gone. if people get their thrills from stone raining people then thats cool. its not like the deck is too good.
UWGSnow-Bant Control
BURGrixis Death's Shadow
GWBCoCo Elves
WCDeath and Taxes(sold)Since it's already established that a (competitive) deck is designed to accomplish three things:
We can see how all decks accomplish these things.
Bogles does so by using hexproof creatures and effectively making their creature threats much larger and/or resilient than the opponent's. That would be pre-emptive interaction with the opponent's decision tree.
That actually leads us to what I think may be the root of the misguided use of the word "interaction". People see the cards on the battlefield, in hand, or on the stack, and focus on that. But a deck can best interact by being more focused on an opponent's decision tree, which isn't a physical thing that we can see. People are very vision-oriented, so this isn't something that is as easily perceived and understood, I guess.
So rather than scoring cards on how they interact with other cards in the visual sense, it may be better to score them by how well they interact with another deck's decision tree.
Lantern Control
(with videos)
Uc Tron
Netdecking explained
Netdecking explained, Part 2
On speculators and counterfeits
On Interaction
Every single competitive deck in existence is designed to limit the opponent's ability to interact in a meaningful way.
Record number of exclamation points on SCG homepage: 71 (6 January, 2018)
"I don't want to believe, I want to know."
-Carl Sagan
UWGSnow-Bant Control
BURGrixis Death's Shadow
GWBCoCo Elves
WCDeath and Taxes(sold)It seems weird giving Death's Shadow such a high rating, as players would sometimes feel like it was almost a combo card in itself.
Something like Snapcaster should get about as many interaction points as possible, as the potential is so high.
I mean, technically something like Burning inquiry is really high on interaction, but if your opponent suddenly drops 2 4/4 Hollow Ones before you even make a land drop, how interactive and fair is that?
There's this weird, intrinsic interaction and fair scale, too.
Tron's blowout cards are super interactive, but they can be deployed so fast that it doesn't feel fair
In Zoo style decks, creatures can definitely block and/or attack planeswalkers but most likely they will go face. In creature based combo decks, most creature will just sit around looking at the board around them until it's time for them to fulfill their part of the combo.
I agree with what others pointed out, that it would be a huge amount of work, but I don't think it would be unmanageable. It could easily even be crowdsourced if you create the correct platform. Something like someone setting up a platform where people can register decks, rate the cards and get a result, and then the results could be compared. That would be interesting and could be realized within a relatively small period of time if enough people from the forum contribute. However, I wouldn't know how to create such a platform, my programming skills definitely rate 0 in the interactive spectrum.
That being said, the contextual problem will remain, and it would need to be clarified that we would not be looking to contextualize cards in their interactivness. This means that Grapeshot still gets a 3 (face, creature, PW), even though it usually just goes face.
UB Faeries (15-6-0)
UWR Control (10-5-1)/Kiki Control/Midrange/Harbinger
UBR Cruel Control (6-4-0)/Grixis Control/Delver/Blue Jund
UWB Control/Mentor
UW Miracles/Control (currently active, 14-2-0)
BW Eldrazi & Taxes
RW Burn (9-1-0)
I do (academic) research on video games and archaeology! You can check out my open access book here: https://www.sidestone.com/books/the-interactive-past
A non-interactive deck is like a racer who focuses on getting the fastest bike with the best parts and has a game-plan of just cruising past the competition without getting into any scuffles.
An interactive deck is like the racer who brings a crowbar to the race to beat the tar out of other racers.
Re: Complexity/ease of the grading method
I don't think the method has to be easily implemented to be viable. If the method happened to be time consuming but still produced accurate results, that's fine with me. Metagame compilations were time consuming but were quite accurate for what they claimed to be doing.
Re: subjectivity/objectivity of the grading method
This method should absolutely not be subjective at all. There should be clear, objective measures of when cards get interaction points and they should be consistently applied. For instance, it is 100% objective that Bolt can target three elements of an opponent's gamestate (board, walkers, creatures). We start getting muddier when something like Shadow having 3 modes (attack player, walker, and block) gets reduced because it's "effectively" just a dumb beater, or that Vizier gets reduced because it's effectively a combo piece.
Re: combo cards getting too many points
This should be accounted for by the other cards in the deck. For instance, Electromancer, Baral, and Grapeshot each get a bunch of points because they can attack, block, and target different stuff respectively. So an average Storm deck would get about 30 points for that pile of cards (3 points for each copy of the cards). But the rest of Storm would see a huge dropoff in points: Manamorphose, Gifts, Dritual, PRitual, PiF, SV, Sleight, etc. all contribute literally 0 points to the deck. This should lead to combination decks having far fewer points than non-combination ones after a final calculus.
Re: the "Burn problem"
I like the idea of the synergy approach to this (i.e. diminishing returns on effects that are identical to existing effects), but I think this creates false positives/negatives for other decks. For example, a deck that was all counterspells might lose points here even though I'm sure we would all agree it is mostly interactive. I think scoring cards by the different types of permanents they can interact with might adjust for this over the long run. So Bolt can still get 3 points, but something like Mana Leak would get 6 points because it can hit all 6 non-land card types. Same with Thoughtseize/IoK. Something like Cryptic would get even more points because it has different modes that effect all types of cards including lands. This might equalize out the point spread in the long run.
Re: limiting decision trees
I'm not sure how this method would actually be scored with multiple decks, so maybe I'd need to see it to know whether it's appropriate or not. Does anyone have an example of how two random decks would score on it? For instance, Bogles (a known non-interactive deck) vs. Jeskai Control (a known interactive deck)?
URStormRU
GRTitanshift[mana]RG/mana]
Cattleprod!!!
I loved that game.
Modern: Storm
Legacy: ANT
No, no it does not. You can add your own biases to the conclusion as you please, but the method proposed by KTK does not attempt to measure "good", "bad", or "better". Just interaction.
Now certainly there are people who will strenuously argue that interactive is good and uninteractive is bad (and probably some vice versa) but that's a value judgment. If there were an objective classification such as KTK's scoring system, we could at least agree on the definition were using when discussing how a metagame sorts out across the spectrum of interaction.
If the top players at my LGS are playing Lightning Bolt, why would I play a deck that folds to Lightning Bolt? If the players are playing a lot of creature removal in slower decks, why would I play a handful of slow creatures that lose to that? If players are playing super quick goldfish creature decks, why would I play something that durdles and dies on turns 3 and 4? Maybe that's fun for some people; not me.
I fear some of you will never have that feeling when your opponent has 2-3 cards in hand (presumably Lightning Bolt and Fatal Push, among others) when you're playing Bogles or Grishoalbrand. I don't want to give you a chance to interact with every single card in your hand. Hollow One and Gurmag Angler dodge Bolt/Push, so that's what gives them more resilience.
Premodern - Trix, RecSur, Enchantress, Reanimator, Elves https://www.facebook.com/groups/PremodernUSA/
Modern - Neobrand, Hogaak Vine, Elves
Standard - Mono Red (6-2 and 5-3 in 2 McQ)
Draft - (I wish I had more time for limited...)
Commander -
Norin the Wary, Grimgrin, Adun Oakenshield (taking forever to build)(dead format for me)What is the goal, then, of measuring interaction? I'm sorry if I missed it in the lat few pages.
URStormRU
GRTitanshift[mana]RG/mana]
The reason you're running into issues is that you're using the wrong heuristic. A couple years ago I discussed this idea with you briefly. I've since moved on to other systems of trying to solve the game because of the data input this would involve but I think you want to look at this slightly different. It's not really about scoring cards as high/low but scoring cards at all. In the system I envision, you create a series of rules such as:
Trades at parity (no more than CMC + 1)
Turns off card
Trades or better in combat
And a few other rules (in the AI I have that plays Magic I have several rules like this)
Then, you rate card A vs card B in a matrix. If A passes all of the above rules against B it gets 1 point, if it fails any of them it gets a 0.
You can then sum the cards along their rows. The cards that pass the most cases, will naturally be the most interactive against a general meta. If you restrict the columns to include only certain cards, representing certain decks (or excluding certain decks) then you can make a more nuanced case about particular matchups. You can even weight all of this by meta percent to make card and deck building suggestions for an expected metagame.
This system could even identify new cards that don't yet see play by discovering interactions not currently being considered.
Given enough manpower I am 100% certain the system can work. The problem is that it has a very prohibitive data entry requirement and it scales exponentially, so as new cards are introduced to the format it becomes harder and harder to maintain. If we had a 10 card format it would require a 10x10 matrix for 100 points of data, a 100 card format is 100x100 or 10,000 points of data. The difference between a 400 and 500 card format (a 25% increase in the cards) 160,000 vs 250,000 pieces of data, a 56% growth in inputs required.
Currently I'm working with a small team (though the project is stalled until late may/early june) trying to implement this system in Legacy. Legacy currently has the smallest competitive cardpool,so it makes sense in that format. We've come to the conclusion though that it can only work if we write a system that can parse oracle text to determine function. Doing it manually is out of the question unless there's a very dedicated community to doing so. Based on what I've observed from your previous metagame updates before you quit doing it... the community to build and maintain these statistics doesn't exist.
The good news though, is that automatic card parsing SHOULD make the project apply to any existing or arbitrary format and the hope is that we would be able to build a database of decks with cards they may include, and then tweak to any specific meta.
This would result in an answer to a slightly different question from what you're asking though. You're asking about interactivity while my proposed answer is what lines up well. Sometimes two decks don't interact but still line up well against each other... Scapeshift vs Grapeshot for example.
Also, this approach would not be able to handle combo.
Here's my current rules for the similar approach I just outlined that I'm currently using in Legacy:
This example was rating Nissa, Vastwood Seer in Nic Fit (our goal was to rate finishers only) but you can get a general idea from this with the rules involved (this also applies to ktk's proposed system):
General criteria:
1 - 6 mana or less
1 - Card advantage
1 - Can GSZ (optional?)
Combat:
1 - Beats Strix
Beats Angler
Beats TNN
Beats Griselbrand
1 - Beats Delver
1 - Beats Jace
1 - Beats Leovold
Beats Marit Lage
1 - Beats TKS
Beats Reality Smasher
1 - Beats DRS
1 - Beats Lingering Souls
1 - Beats Pyromancer
Beats Tarmogoyf
Resiliency
1 - Beats Fatal Push @2
1 - Beats Fatal Push @4
1 - Beats STP
Beats Bolt
1 - Beats Deed
1 - Beats Deluge
Beats Decay
Beats Pulse
Beats FoW
1 - Beats Spell Snare
1 - Beats Spell Pierce
1 - Beats Daze
1 - Beats Stifle
1 - Beats Terminus
1 - Beats Tabernacle
Beats Punishing Fire
1 - Beats Jitte
1 - Beats Chalice
1 - Beats Kommand
Beats Discard
Evasion:
Relevant protection
Evasion ability
1 - Goes wide
There is no issue in playing a non-interactive deck. People may label or assign negative value to that label, but it just is a label.
Bogles is not interactive, its quite literally (Game 1) the most non-interactive deck you can imagine. Thats simply fact.
That doesnt mean its a bad thing, or a poor choice, or without logical purpose behind running the deck.
Simply dont try and call it interactive, as in the colloquial term used by many (most?) Magic players.
Spirits
the difference is that you are attempting to solve a format given a card pool and a metagame. whereas this thread, and what wizards cares about, are gameplay considerations which go beyond winning or losing to include more abstract elements like fun or enjoyment.
some people are obviously biased, but in the context of format health its about breadth of gameplay aimed at maximizing the enjoyment for the most players possible. if one set of play patterns is overly represented that generally isnt a good sign. a format defined by nothing but midrange/control grindfests is just as bad as nothing but overtly linear strategies in aggro/combo.
the problem is that most of this is largely subjective, and ultimately depends on the vision that wizards has for the game or format since they are the arbiters. their goal is obviously attracting and keeping players engrossed in the game for as long as possible, while also balancing the competitive aspect to ensure that skill and decision making are large enough factors in success.
UWGSnow-Bant Control
BURGrixis Death's Shadow
GWBCoCo Elves
WCDeath and Taxes(sold)To be fair, I don't think it's been actually stated. The goal of measuring interaction with a clearly defined model is an objective measurement for each deck. That can then be used to roughly compare decks, evaluate the overall mix of a metagame across time, etc. At this point we're stuck with "I know it when I see it" evaluations of interaction and that quickly bogs down in arguments over the trees rather than taking in the forest.
There are a lot of second order applications from there. For example, determining which type of metagames WoTC has historically found acceptable and then using it as a predictor for future bans/unbans. Or evaluating how metagames cycle (or don't) between interactivity and uninteractivity.
EDIT: While I'm at it, to avoid double-posting, working on Lantern is also what I feel helped me come to understand the game better, to include having a better grasp of what it truly means for a deck to interact or have linear characteristics, and why decks are best designed to minimize the opponent's ability to interact effectively, maximize it's own ability to interact effectively, and do those two things consistently.
I feel that we can understand why people currently use those terms when they do by understanding human behavior. As I mentioned before, our species is heavily influenced by what we are able to directly observe - Physical objects and their movement. When someone plays a card in this game that drastically reduces the number of branches on an opponent's decision tree, we don't get to see those branches get pruned. We see cards moving around some, and that's about it. But if we can see past the movement of those cards, and use our minds to perceive the invisible changes in respective decision trees, I feel that we can come to a deeper understanding of the game.
So our dilemma, when we want to understand the game better, and when we want to talk with others about understanding the game better, is that we must be able to overcome our own human instinct to focus only on what we can see and we have to avoid using (as someone recently best described it) dog whistle terms for criticizing what we don't like, usually due to some feeling of entitlement for what we might consider fair and just.
Lantern Control
(with videos)
Uc Tron
Netdecking explained
Netdecking explained, Part 2
On speculators and counterfeits
On Interaction
Every single competitive deck in existence is designed to limit the opponent's ability to interact in a meaningful way.
Record number of exclamation points on SCG homepage: 71 (6 January, 2018)
"I don't want to believe, I want to know."
-Carl Sagan
Not necessarily. There's a distinction to be made between interacts and follows the rules. Slippery Bogle beats Fatal Push but I wouldn't say that makes Bogle an interactive card. It just means that it's positioned well against it and that's what the system I outlined is designed to measure. Interaction is a subjective term because highly interactive could mean fun or unfun. Lantern is very interactive, Bogles is not, so fun clearly transcends interactivity. Furthermore there exists good decks on both ends of the spectrum. I'm not attempting to say interaction is good or bad and instead looking only at positionings of groups of cards against other groups of cards.
The idea of such a system is to solve a metagame quickly, test what should be banned, test what should be unbanned, try new cards in old decks, and figure out what is currently positioned well, and what is not positioned well in a non subjective manner. Like I said though, I don't believe the community is capable of providing the required data set here (and most of the data needs to be filled in before any results can be reliable) so I've been looking at automated methods of doing this. I've made some progress on that front, but life obligations currently prevent me from working on it further for a few weeks minimum.
EDIT: @tronix I think that's a misconstrued version of what I'm saying. I'm not saying any quality that makes a card or deck "good" is a form of interaction. I think it would be more precise to say that what reduces the amount of interaction that a potential opponent might have is what makes it "good".
Lantern Control
(with videos)
Uc Tron
Netdecking explained
Netdecking explained, Part 2
On speculators and counterfeits
On Interaction
Every single competitive deck in existence is designed to limit the opponent's ability to interact in a meaningful way.
Record number of exclamation points on SCG homepage: 71 (6 January, 2018)
"I don't want to believe, I want to know."
-Carl Sagan