Diminishing returns on "better" cards

[Avacyn Believer]

So, I am sitting here away from home (deployment), playtesting some new cards for my main deck so I can comment on them in my primer and maybe order them, and it starts to dawn on me that the deck doesn't feel like it plays any differently.

Have you ever felt that way?

Admittedly the cards I am testing are more for the later game but it is not the first time I felt this way, like I am making very marginal gains. I think it feels more obvious because less than two years ago when I was playtesting while I couldn't play it was a significant change to the speed of the deck. I wouldn't say the deck peaked, but I do feel at a plateau, at least without drastically changing some strategy.

Maybe I just need a break from trying to find "better" cards to swap.

[pokken]

I am at the point with most decks where -

1) upgrades are incremental
2) there are so many cards I want to play that would be perfectly fine

Part of that I think is that there's just such a huge volume of high power cards now -- I've been lately cutting tutors extensively to just play more good cards just to get more slots to do stuff :D

[TheGildedGoose]

Once you reach a certain point of deckbuilding and getting a firm grasp of the fundamentals of the format, this is the fate of your decks. It becomes a game of inches until some new card is printed and blows it all apart, at which point you get to rebuild and reconsider past options. While I really like the creative phase of deckbuilding, tuning is my strength, and though there are certainly diminishing returns, you may discover new lines of play that will absolutely win you games as you become more experienced and consider new interactions. It's incredibly rewarding to do so.

[PrimevalCommander]

Eventually if a deck gets worked on enough, it reaches a point of optimization that it will take a complete retool of the core synergy of the deck to make a dramatic difference in play style. Many of my decks are like that as well, and even those incremental upgrades can take several outings to reveal themselves. This leads me to build new decks every so often to restart the process and get the creative juices flowing again. Once I reach about 10-12 decks I'll find the least played few and retire them to make room for new. Though that will be difficult now since I have become fond of all my current decks, there is still one or two that get much less attention than the others.

[materpillar]

To be fair, without a large amount of tutors your deck is going to play fairly different from game to game. Feeling the impact of one card change to another is going to be somewhat rare since your deck is so large unless the card improvement is massive.

[DirkGently]

This is part of why I hate tuning and instead just start building a new deck.

I'll finish the decklist in a hurry, tell myself "it's fine, I'll see how it plays and then move stuff around" and then 9 times out of 10 I don't bother. Just apply lessons learned to the next build.

[Serenade]

Yeah, we're getting a new, better version of the general in six months anyway. I don't tune anything anymore.

[Mookie]

Broadly speaking, most deck tuning takes the form of hill climbing optimization - you do some testing, then you make minor tweaks based on those tests. If you have too little of some effect, you add more. If you have too much of something else, you cut some. Eventually, you'll converge at some sort of local optima, and any additional changes will be fairly minor.

However, just because you have found a local optima doesn't mean you have found the global optima - there is probably some better version of your deck out there if you're willing to make some drastic changes and rework some basic assumptions. If you converge back to the same optimized list, then that just means your changes weren't drastic enough.

Alternatively, you can try playing against different people. You've probably optimized for your local meta... but if you start playing with a different group, you'll probably find that the things that work look very different.

Finally, I'll note that some decks converge quicker than others. It's pretty easy to solve tutor-heavy decks like Vannifar, Evolved Enigma that play the same every time... but testing and tuning takes longer if you're only seeing a small portion of your deck every game. Similarly, some decks are juggling more things, and are more unstable as a result - decks that only care about one thing (like, say, enchantress) are often easier to tune than something like Kaalia of the Vast that is trying to juggle giant beaters + haste + protection + ramp to cast beaters.

Anecdotally, I've found my Animar, Soul of Elements and Sharuum the Hegemon decks to be pretty stable - they only care about creatures and artifacts, respectively, and my changes to them tend to be relatively minor. I did make a somewhat major change to the former recently though when I dropped Primal Surge, and I've been slowly changing things due to the fallout from that change.

On the flip side, I've found my Teysa, Orzhov Scion and Samut, Voice of Dissent decks have been in constant flux. The former wants to balance between token producers, sac outlets, and other aristocrats stuff. I've gone through two major changes recently when I tore out all the infinite combos and shifted to something more midrange-y, and again when I shoved in Relic of Legends, We Ride at Dawn, and a bunch of other legend synergies. Meanwhile, Samut has been balancing between creatures with tap abilities, untap effects, and the perpetual allure of just playing giant beaters like Inferno Titan. Lots of changes there.

...of course, you can also just retire a deck for a year or so and come back later - every time a new Modern Horizons or Commander Legends set comes out, the format changes pretty drastically.

[Cyax]

As other people in the thread have pointed out, the source of the largest increases in a deck's power comes from modifying the core strategy.

Frankly, some commanders, cards, and strategies are way better than the others. As an extreme example, converting an Obeka, Brute Chronologist into a Grixis turbo combo pile with Thassa's Oracle, Ad Nauseam, and Underworld Breach instantly turns a casual deck into a top-tier cEDH list.

But even barring the extreme examples, there are more efficient ways to do what casual decks want to do. For example, running multiple tutors to consistently play The One Ring early is a much better play for card draw than simply playing a Phyrexian Arena. Similarly, other game aspects such as ramping or strategic archetypes have huge variances in strength depending on what you do.

On the other hand, upgrading cards like Murder into Hero's Downfall is a small upgrade because you are still paying three mana to kill just one creature. Even changing a Counterspell into Force of Will won't help your deck much until it handles the most important concerns of mana, card advantage, speed, and consistency. Only then will smaller details between cards become relevant, which from the implications of your question is the aspect that you are focusing on prematurely.

[darrenhabib]

I've definitely gotten to the point with my primers that I'll often try cards that make them moderately worse, but I like trying new cards "for science".
It might represent a new angle on the deck, which isn't just done best with a simply swap, but a few card changes to make that card idea more synergistic.
But the changes come at some cost. Anyway it's pretty hard to break into my primers at this stage, but I think one card maybe every two sets will make it in.

[3drinks]

I've been saying this for years, most new cards are just side-grades at best, so don't bother buying (especially don't buy them at insane presale scalper prices). Unless it's a dockside crackstortionist, smothering tits, or teferi's pro, which have warped the format significantly around their existence...

Otherwise you're really just comparing Hero's Downfall to Bitter Triumph. Yeah, one of these is better but it's not raising your win %, statistically speaking, to play one over the other.

[Avacyn Believer]

cutting tutors extensively to just play more good cards

Generally speaking I agree and encourage less tutors in decks to have more fun, but in my Avacyn deck it's the only way I found (for now) to make Voltron fast enough, so I try not to cut them, but there is some balancing of the right number of tutors.

you may discover new lines of play that will absolutely win you games as you become more experienced and consider new interactions.

I'd say that is what I am waiting for right now. There hasn't been enough cards printed yet that would, together, make me think I could approach winning with Voltron Avacyn differently.

Once I reach about 10-12 decks I'll find the least played few and retire them to make room for new.

Such is the life cycle of EDH players

To be fair, without a large amount of tutors your deck is going to play fairly different from game to game.

The wall I've been running into with this deck is not the number of tutors available, but the kind of tutors available in Boros. I can tutors for Equipment and Auras for days, but not so much other things, at least not as consistently.

This is part of why I hate tuning and instead just start building a new deck.

That's fair. To be honest I don't tune my other decks all that much, nowhere near as often as this Primer deck. I guess I feel obliged to at least try to I can keep the set reviews honest.

if you're willing to make some drastic changes and rework some basic assumptions.

Not wrong, I am not really considering drastic changes, at that point I'd more so consider building a different deck, even if it is with the same commander.

I should of kept a list of the different LGS and groups I played this deck against (I travel decent amount because of work). I like to think its optimised to the strategy to play well in any meta. I do find it fun to play in different places and see how people play against the deck.

...of course, you can also just retire a deck for a year or so and come back later

There can be no retirement from devotion to Avacyn

a small upgrade because you are still paying three mana to kill just one creature.

Otherwise you're really just comparing Hero's Downfall to Bitter Triumph. Yeah, one of these is better but it's not raising your win %, statistically speaking, to play one over the other.

That is relatable. I do end up looking at new printed Equipment that is good, and all I can really think is that yeah, it is good, but the Equipment I already play does the job, so swapping feels like swapping for sake of changing something. If that makes sense.

Anyway it's pretty hard to break into my primers at this stage, but I think one card maybe every two sets will make it in.

I feel that. And on some level it is an interesting experience because I am so casual about all of my other decks that I never imagined I'd have this "problem".

[TheGildedGoose]

Otherwise you're really just comparing Hero's Downfall to Bitter Triumph. Yeah, one of these is better but it's not raising your win %, statistically speaking, to play one over the other.

Let's say there's a mono-black deck that wants to run seven pieces of spot removal.

Deck A runs the following:
Murder
Hero's Downfall
Swift End
Poison the Cup
Rend Flesh
Eyeblight's Ending
Defenestrate

Deck B runs the following:
Bitter Triumph
Infernal Grasp
Heartless Act
Go for the Throat
Power Word Kill
Baleful Mastery
Shoot the Sheriff

Which deck do you think will win more a statistically significant percentage of the time? Inches become feet, feet become miles.

I tend to think in terms of packages these days. Land ramp, mana rocks, tutors, removal, counterspells, hell, even win conditions in certain archetypes. The whole deckbuilding process is pretty demystified to me now, at least in terms of what works for me, which is kind of boring.

[illakunsaa]

This is why I kinda hate the maximum decksize rule.

Why cant I just turn my decks into blobs made out of love?

You could also just make a new variant of the same deck. I did it and now I have like 10+ turbofog decks.

[Toshi]

it starts to dawn on me that the deck doesn't feel like it plays any differently.

Have you ever felt that way?

I think when it comes down to Primer level decks, i comes down to subtle and miniscule changes and therefor the perceived difference will diminish.
In Teshar, Ancestor's Apostle i went from Excise the Imperfect, Generous Gift, Forsake the Worldly and Lapse of Certainty to Bovine Intervention, Get Lost, Lost to Legend and Reprieve.
Will that deck feel inherently better now? Not really. But what i realized overall is, that once you shave of more and more less optimal cards your perception of a deck might become more bland, since everything is redudant, available and viable. It takes broken stuff like Dockside Extortionist for me to actually feel an actual change of urgency in a game.

[Jemolk]

There's a reason I have almost fifty decks at this point. This is that reason. There's only so much tinkering to be done, but I still want to build, adjust, and play around with card choices. And at the same time, I get far too attached to my decks to disassemble them. I do still tinker with my decks quite a bit -- and sometimes I make "upgrades" that are more sidegrades or even downgrades from a power level perspective, and are merely more flavorful or thematic or something, which slightly expands the space for adjustment. But overall, decks eventually stabilize to a point where I like the core strategy more than the alternatives, and improvements are rare and insignificant, so to get creative with things, I have to build something wholly new.

[3drinks]

You'll take my disenchant|parl when you pry it from my cold, dead fingers.

[PrimevalCommander]

You'll take my disenchant|parl when you pry it from my cold, dead fingers.

Same with my Vindicate|26480

[Toshi]

I make "upgrades" that are more sidegrades

That one goes right into my vocabulary

[3drinks]

You'll take my disenchant|parl when you pry it from my cold, dead fingers.

Same with my Vindicate|26480

I had this one. Such beautiful art. Then I sold it when it was at $50. I wanna say I don't miss it when Vindicate these days is $1 bulk rare...but here I am carrying a Disenchant marked up ~1400%

[3drinks]

Otherwise you're really just comparing Hero's Downfall to Bitter Triumph. Yeah, one of these is better but it's not raising your win %, statistically speaking, to play one over the other.

Let's say there's a mono-black deck that wants to run seven pieces of spot removal.

Deck A runs the following:
Murder
Hero's Downfall
Swift End
Poison the Cup
Rend Flesh
Eyeblight's Ending
Defenestrate

Deck B runs the following:
Bitter Triumph
Infernal Grasp
Heartless Act
Go for the Throat
Power Word Kill
Baleful Mastery
Shoot the Sheriff

Which deck do you think will win more a statistically significant percentage of the time? Inches become feet, feet become miles.

I tend to think in terms of packages these days. Land ramp, mana rocks, tutors, removal, counterspells, hell, even win conditions in certain archetypes. The whole deckbuilding process is pretty demystified to me now, at least in terms of what works for me, which is kind of boring.

Kinda changing the formula here from my point when, yeah, of course making your entire removal suite +1mv from where it should be will statistically affect your win % in a tangible non-zero percent of games, but point taken. My example should've used Terminate|pls to Bitter Triumph for a more concise, apples::apples comparison. The point was trading one two mv good removal for another shiny new two mv good removal doesn't statistically move the needle enough to be worth burning the mental calories to make the change in the first place. Despite this new removal being "better", it's still trading two mana and a card for another card an opponent has. Unless it hits any card type with no drawback (or mitigated drawback, i.e. AssTroph), naturally.

[TheGildedGoose]

... doesn't statistically move the needle enough to be worth burning the mental calories to make the change in the first place.

This, I think, is the crux of your argument, but in terms of whether or not a given level of diminishing returns is "worth" it or not is pretty subjective.

Not to bring up old %$#%, but I think the discussion recently in the Unreleased Card thread about a particular cycle of cards is representative of this. Will playing those cards significantly raise your win percentage individually, like adding Torment of Hailfire or Skullclamp or whatever would? No, but it's a question of degree, and no card exists in a vacuum. A massive change in one card and a slight change in ten could still have the same effect in terms of win rate.

[Rumpy5897]

It's pretty easy to tell if you're cut out for the tuner life. If your deck does the thing and your response is "that was awesome, how do I get this to happen more consistently?" then I have good news for you. Mookie's hill climbing optimisation description of a build congealing is absolutely spot on.

But what happens when everything is pretty much in shape and you don't really have anywhere to go? Then you're largely combing through new sets and making these inch improvements, unless you invent a new hill to pursue. I'm actually rather unhappy when sets offer too many new includes, as this is often indicative of a power creep bout. OTJ/BIG ended up granting me an above average number of new things to run. We're also entering an era where redundancy creep means you need to actually think about how much of a particular effect to run rather than just dime it, which is going to force me to reevaluate some deck constituents at some point. MH3 is bringing a new Zulaport. There are now more Rampant Growths/Cultivates than a year ago. How do my decks respond to this?

It also helps to be intimately familiar with the dynamics of the decks. This is something I'm guilty of deteriorating at, as I'm not blasting as much EDH as I was in the past, and my attention is divided between more decks as I uncovered the double-edged sword of meta-accepted deck building not killing most of my contraptions in a matter of months. Slumbering Keepguard came out in WOE, I gave it a spin in Daxos and realised it plays well there, but wasn't sure what the cut was. Ultimately the answer took a good half year to appear, and involves doing something completely different than just nixing an existing card advantage/selection piece. I managed to gather enough data to point me toward enjoying a particular thing happening in game that was a bit different from my regular game plan, and I figured out how to get that to happen more consistently without actively shooting myself in the foot to get there. Only took Necrodominance to appear in MH3 to make it possible

Oh, and this sweeping sounding Daxos rework? I'm literally swapping two cards. Tuner life incarnate.