Amazing sequencing, eeking out small advantages in many different ways, and 3 Byes is the key to success?
There's something else that people are forgetting.
How many GPs you attend will also determine your chances of doing well at these tournaments. I went to 2 GP locations last year. The average Pro player/Grinder goes to 10-20 GPs per year. They will have a better chance to win, get top 8, get top 32, Day 2, win a match, than someone like myself even if the play skill is equal, which it's certainly not. How many people here attend 10 GPs per year? I have teammates that do it, so I have some kind of perspective, but it's not the same as me doing it.
*Players like Matt Nass realize that they can't make it to too many GPs, so they took the preparation super seriously. Then they had some luck and good play along the ride.
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)
Thats the thing though, no field is 40% Tron. Its not even remotely close. So to say that Pro's somehow have an inside track on metagame prep, is a steep one for me to accept. They dont have any more or less data than we do.
I think he's just saying that a gathering of minds get together to plan a GP. It would be the same for Grinders that often go to many GPs or even teams that go together often.
I've noticed that some of the people on my team are not as helpful to me as to others, but still just hearing what's going on in their heads quite often helps me towards preparation. I mean if someone has a compelling argument for not playing Titanshift, I wouldn't have done it. I mean, in the end, you have to make up your own mind, but you can bounce ideas off other experienced players. I would rather test vs. my team members 1 day a week than play 2 weekly tournaments. The results are just much more helpful in many ways. (I mean, it's more "fun" to play in 2 weekly events, but it's really not helping much toward GP play.)
You gotta understand this. Just talking and play testing with someone like Paulo Vitor Damo Da Rosa every week is going to improve your game. Anyone here would do much better with that as the only "change" they have. I don't have access to that, so I try to top 8 local tournaments where in the top 8, I can face solid GP competition.
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)
When was the last time a group of pro's spiked an event though, I took a break for a bit during the period that lead to Probe being banned, but the last time before that, was Eldrazi Winter, which EVERYONE figured out, not just pro's.
No deck since (again, unless it was during the period I wasnt paying attention to) has broken through in force, except...JDS? Which was due to a MTGO grinder, the pro's just picked up that individuals deck if I remember right.
JDS had its time in the sun. GDS had its time. Etc, etc. Nothing has broken the format, no meta game is really breaking out to abuse, at any level that we as followers of the format can not also identify...but then you may not even play against it!
As cfusionpm mentioned, when they played the side events of GP Vegas, no Tron, no KCI, all 'fair' midrange type decks, and UR Breach was a great choice.
The 'meta' was Humans, UWR, and Tron though, so how does that happen?
You cannot meta game modern at such a fine level, Pro or not.
So Pro's are simply more lucky at dodging match ups? I dont buy that for a moment.
They also get to dodge the true out-of-nowhere matchups that could happen in rounds 1-3. But success breeds more success. Needing to go 10-2 is alot easier than 13-2.
So Pro's are simply more lucky at dodging match ups? I dont buy that for a moment.
They also get to dodge the true out-of-nowhere matchups that could happen in rounds 1-3. But success breeds more success. Needing to go 10-2 is alot easier than 13-2.
I can get behind this, for sure. Especially the importance of byes, but no, I'm not at all convinced that even the testing of tournament groups, has any great relevance. Unless something is truly broken, teams always seem to come to different conclusions, because no, even pro's (as noted in Ross' article) are biased.
So Pro's are simply more lucky at dodging match ups? I dont buy that for a moment.
They also get to dodge the true out-of-nowhere matchups that could happen in rounds 1-3. But success breeds more success. Needing to go 10-2 is alot easier than 13-2.
Again, this sort of comment is not an indictment of Modern or any Modern dynamics. This is merely a commentary on the importance of byes in all events.
At this point, we have a very clear picture of how top players succeed in all formats. We've seen the same effect on both the SCG and the GP/PT circuit at this point; top players perform consistently in Modern as they do in other formats, and they do so at the same level of performance. So now the argument has shifted away from pro players to average players without byes. This is unsurprising as the Modern criticisms constantly shift when confronted with new data. I am positive that if we broke down the numbers, we would find that the effect of byes/no byes in Standard/Modern is statistically identical. I say this with confidence because basically everything else has been identical between the formats after we break down the numbers, so I see no reason why this won't be the case either. If I'm wrong, I'll happily admit it, and I'll check some numbers to see if it's true. I also am fairly certain that after I run this analysis, and if I find it to be the case that byes matter the same in all formats, some of the critics will AGAIN shift arguments again to show that Modern is high variance, low skill, matchup dependent, etc. Either these things don't exist in Modern OR the best players have found ways to overcome them.
I firmly believe that most of this returns to players who are dissatisfied with their personal performance in Modern and are looking for things to blame. It is easier to blame these external elements, up until they are disproven, than it is to focus on personal areas of improvement. I would 100% buy these criticisms if we didn't see such consistent success by the best players. But we do, and that success is true across formats and in the same degree.
Double-post but I wanted the numbers to stand alone. If this needs to be merged, feel free.
I realized it's pretty easy to compare standings vs. byes with the GP data we have, so I just threw this together tonight. I wanted to see if byes have a greater, lesser, or equal effect on players in Modern vs. Standard. To do so, I looked at standings vs. number of byes for the first five Modern and Standard GP of 2018. For Modern, this was Sao Paulo, Barcelona, Las Vegas, Hartford, and Phoenix. For Standard, this was Pittsburgh, Singapore, Copenhagen, Birmingham, and Seattle. For each event, I found the total number of byes for all players in the first three rounds. I then took the average of their overall points broken down in each of four categories: players with 1, 2, 3, and 0 byes respectively. I did this for all events, averaged them within formats, and then compared them between formats. Here is what I found:
Modern - 0 Byes: 9.78 points Modern - 1 Bye: 14.77 points Modern - 2 Byes: 18.29 points Modern - 3 Byes: 26.52 points
Standard - 0 Byes: 9.85 points Standard - 1 Bye: 14.58 points Standard - 2 Byes: 18.79 points Standard - 3 Byes: 26.39 points
In all categories, there is less than a .5 point difference between the point standings in Modern and Standard. If you have 3 byes in Standard, you average 26.39 points at an event. In Modern, you average 26.52 points. Got 0 byes? In Standard, you'll average 9.85 points. In Modern, you'll average 9.78 points. Needless to say, these differences are not statistically significant at all and are barely different in absolute numbers.
This all but categorically disproves the allegation that byes matter more in Modern because you can dodge bad matchups and have better performance. Byes have an IDENTICAL effect in Standard or Modern. If byes mattered more in Modern, we would see players with more byes doing better in Modern than in Standard. We don't see this at all. Instead, we see players with the same number of byes averaging the same standings as those across format lines. This also supports the argument that the formats are comparably skill-testing from a performance standard, although I admit the required skills may have differences as well as overlaps. Between this analysis and the previous one I posted, we now see that top players have consistent performance in both Modern and Standard, and we now see that byes have an identical effect between formats. We can certainly spin this into an indictment of the disproportionate effect of byes at top events; I'm all for that. But I think we can safely move beyond this untenable notion that byes matter more in Modern because they help people dodge high variance matchups early. Byes have the same effect in Standard and Modern on your win rate at a GP.
So...next topic? Control is fine, Modern is no less skill testing, and byes are a huge advantage regardless of format, while we know Wizards feels Modern is healthy.
Other than my eternal quest to find a true replacement of Twin, I'm not sure what else I can even gripe about, besides Tron/Stirrings...
This also supports the argument that the formats are comparably skill-testing from a performance standard
This seems like a big logical leap. Why do you claim this one is the case? And what exactly do you mean?
We can surely say that we can safely move beyond this untenable notion that byes matter more in Modern because they help people dodge high variance matchups early.
But, we can NOT make assumptions on how much more skill testing Standard is (when compared to Modern), based on those data. Right?
It doesn't prove the argument. It just supports it. The common claim, at least from certain critics, is that Standard or Legacy requires more skill than Modern. And yet, in every analysis I've posted here when we actually compare formats, we find that the formats all have the same numbers. Top players have the same top MWP. The MWP ceilings are the same. They have consistent performance between events. Byes matter equally between formats. There is simply no performance evidence that Modern is any more or less skillful than any other format. If Modern was a less skill-intensive format than Legacy or Standard, we would see evidence of this in at least one of these areas. We would see byes matter more in Modern than Standard. We would see a lower MWP ceiling (if top players were victim to variance) or a higher MWP ceiling (if the format was easier to win at). We would see pros at GP do worse in Modern than in Standard (if variance determined their matchups) or better (if Modern was easier to succeed at). There would be some evidence of this, and yet, we don't see it at all. All of these analyses push harder and harder against the notion that Modern requires less skill than Standard.
Are there any numbers-based arguments for Modern requiring less skill? Those arguments are something like 0 for 5 at this point, and if there are more of them I'd love to see and test them.
This also supports the argument that the formats are comparably skill-testing from a performance standard
This seems like a big logical leap. Why do you claim this one is the case? And what exactly do you mean?
We can surely say that we can safely move beyond this untenable notion that byes matter more in Modern because they help people dodge high variance matchups early.
But, we can NOT make assumptions on how much more skill testing Standard is (when compared to Modern), based on those data. Right?
the 'from a performance standpoint' rider is the key to his statement.
let me clarify for you:
modern is probably more skill intensive than standard. (more decks, more cards, more variety, shorter games mean individual decisions matter more in your average game etc. etc.). I have no statistical basis for this, but it seems intuitive based on what we know... so... "probably".
What's important though, and what KTK was saying, is that this doesn't affect relative performance between formats.
simplifying it; I guess you'd say that a decent player who puts the time into mastering Standard would have an equal amount of success in Modern if they replicated their efforts there. This is backed up again and again by looking at the spread of repeat-success from those pros and players who put the time in. Even on a local level, you'll probably see this, but it's harder to discuss because it would mostly be anecdotal.
What you may be alluding to is a potential difference between the learning curves of the two formats. I have no evidence to support this, but my own experiences have led me to consider that Modern probably has a steeper 'starting point' on the learning curve than Standard. Ultimately, they'll end up in the same place, but i have this intuition that Modern is initially a trickier format to start playing. Perhaps this difference might account for newer players to the format slinging around arguments about variance and blaming their losses on diversity etc. Who knows. That's wild speculation, but conceptually it seems to fit the mold and make sense.
this is actually a super interesting topic of discussion. =)
the 'from a performance standpoint' rider is the key to his statement.
let me clarify for you:
modern is probably more skill intensive than standard. (more decks, more cards, more variety, shorter games mean individual decisions matter more in your average game etc. etc.). I have no statistical basis for this, but it seems intuitive based on what we know... so... "probably".
What's important though, and what KTK was saying, is that this doesn't affect relative performance between formats.
I'd argue something completely different from recent pages of discussion. Modern requires a significant amount more work to be effective in the format compared to Standard.
Skill in Magic can be measured in multiple ways, but what I find are two issues that arises frequently; The ability to recognize areas that need improvement, then acting upon to create those improvements. There are hundreds of players I have come across, both intelligent and somewhat so, who have difficulty in acknowledging their own failures. On the other side of the coin, there are people I find who are able to recognize their problems, but recreate them without any signs of improvement towards a solution.
It's why I find the skill argument lacking when comparing formats. It's just deeply layered problem solving.
On the face of things Shmanka, you're talking about general human problem solving and evaluation skills. It's an interesting topic and incorporates all sorts of interesting academic areas of study, but it isn't really what we were talking about?
Anyway let's evolve beyond the 'standard vs modern' dichotomy and talk another other stuff.
Anyone here like.. Weirdly excited to see infect back as a deck people are willing to play? Fundamentally I don't think it was ever bad, but it's certainly having a mini Renaissance and while I have never really played the deck for myself, it's way cool to see it back again.
Private Mod Note
():
Rollback Post to RevisionRollBack
Modern: G Tron, Vannifar, Jund, Druid/Vizier combo, Humans, Eldrazi Stompy (Serum Powder), Amulet, Grishoalbrand, Breach Titan, Turns, Eternal Command, As Foretold Living End, Elves, Cheerios, RUG Scapeshift
On the face of things Shmanka, you're talking about general human problem solving and evaluation skills. It's an interesting topic and incorporates all sorts of interesting academic areas of study, but it isn't really what we were talking about?
Anyway let's evolve beyond the 'standard vs modern' dichotomy and talk another other stuff.
Anyone here like.. Weirdly excited to see infect back as a deck people are willing to play? Fundamentally I don't think it was ever bad, but it's certainly having a mini Renaissance and while I have never really played the deck for myself, it's way cool to see it back again.
I only played against infect a couple of times while on blue tron, so it was obviously a *****show not in my favor. I like it, though. It preys on tron and is one of the best linear decks at turn 3 kills, which lets it race things like storm and KCI. Yet, because it is creature based, it struggles against midrange and control (outside of the occasional "oops I win" opener). I'd consider building a GB variant of hierarch wasn't disgustingly expensive. I simply cannot justify spending more than $30 on an individual card.
So...next topic? Control is fine, Modern is no less skill testing, and byes are a huge advantage regardless of format, while we know Wizards feels Modern is healthy.
Other than my eternal quest to find a true replacement of Twin, I'm not sure what else I can even gripe about, besides Tron/Stirrings...
Ya I mostly agree. I've tried blue moon, adnauseum, cheerios and other combo decks but I haven't found a deck I liked as much as twin. I've told myself for a long time now that I would be happy enough with the ban list if BBE and twin could just get unbanned. We are half way there!
It would be cool to have stuff like SFM, gsz or others unbanned but I'm not as motivated by any others. Sittings and tron are annoying but I'm not ready to take action against them yet.
The data on win percentages in different formats is interesting. I've posters say that Modern has a lot of 70/30 & 65/35 matchups while legacy and standard have more 55/45 match ups and far fewer decks. IF (I don't think I've seen data proving this) this is true is it possible the wide range of decks in modern still leads to the same win percentage because playing a wide range of 70/30 & 65/35 match ups would be equivalent to playing a small range of decks with 55/45 match ups?
1) Standard(which I dislike as a format), does not produce any(or very few) non-games(mainly cards from 8th/9th edition). There are no turn 2 Blood Moons, there are no Ensnaring Bridge that mean odds are heavily against you if you are on a creature deck or else, you don't care it at all if you are a combo deck, no cards like Choke that one can side in, draw and make him a favourite to win.
2) In Standard, there are little to none 35-65/65-35 matchups. More matchups are close to 50/50 or 45-55/55-45 than Modern. This, creates more games that are based on skill and decisions, rather than games where you are back against the wall if you play vs that.
The data on win percentages in different formats is interesting. I've posters say that Modern has a lot of 70/30 & 65/35 matchups while legacy and standard have more 55/45 match ups and far fewer decks. IF (I don't think I've seen data proving this) this is true is it possible the wide range of decks in modern still leads to the same win percentage because playing a wide range of 70/30 & 65/35 match ups would be equivalent to playing a small range of decks with 55/45 match ups?
I imagine this is the next frontier of the Modern criticism. Not saying this to throw any shade at you two, particularly at ThinkingChimp who phrases it as a question and not a reality. I'm just pretty sure this is where the matchup lottery, high variance, low skill argumentation is going to move next.
Thankfully, we already have the numbers to check this: we just need to look at MWPs in tournaments. If Modern had a lot of polarized matchups, we would see highly variable MWP over smaller samples of events, and then we would see those normalize over time to the same MWP bracket as Standard. Standard's would be normalized the entire time. We would be able to measure this in the variance of the population, which might have MWP that averages to the same as a Standard format, but would have greater variance overall.
For example, in a population of 20 events where the player alternates a 45% MWP and a 55% MWP, the variance is only .002. But in a population of 20 events where they alternate 70% and 30%, the variance explodes to .04. The average of both is the same at 50% MWP. We can also check the quartile ranges of those populations, which would reflect the greater variability: in the 45-55 matchup, the 1st quartile would be .45 and the third quartile would be .55. In the 70-30 matchup format, it would be far wider at .30 and .70. I will test this later today and see what I find. I already know the variance is identical for the formats (.039 in both). But I have not tested the quartile ranges.
1) Standard(which I dislike as a format), does not produce any(or very few) non-games(mainly cards from 8th/9th edition). There are no turn 2 Blood Moons, there are no Ensnaring Bridge that mean odds are heavily against you if you are on a creature deck or else, you don't care it at all if you are a combo deck, no cards like Choke that one can side in, draw and make him a favourite to win.
2) In Standard, there are little to none 35-65/65-35 matchups. More matchups are close to 50/50 or 45-55/55-45 than Modern. This, creates more games that are based on skill and decisions, rather than games where you are back against the wall if you play vs that.
The data on win percentages in different formats is interesting. I've posters say that Modern has a lot of 70/30 & 65/35 matchups while legacy and standard have more 55/45 match ups and far fewer decks. IF (I don't think I've seen data proving this) this is true is it possible the wide range of decks in modern still leads to the same win percentage because playing a wide range of 70/30 & 65/35 match ups would be equivalent to playing a small range of decks with 55/45 match ups?
I imagine this is the next frontier of the Modern criticism. Not saying this to throw any shade at you two, particularly at ThinkingChimp who phrases it as a question and not a reality. I'm just pretty sure this is where the matchup lottery, high variance, low skill argumentation is going to move next.
Thankfully, we already have the numbers to check this: we just need to look at MWPs in tournaments. If Modern had a lot of polarized matchups, we would see highly variable MWP over smaller samples of events, and then we would see those normalize over time to the same MWP bracket as Standard. Standard's would be normalized the entire time. We would be able to measure this in the variance of the population, which might have MWP that averages to the same as a Standard format, but would have greater variance overall.
For example, in a population of 20 events where the player alternates a 45% MWP and a 55% MWP, the variance is only .002. But in a population of 20 events where they alternate 70% and 30%, the variance explodes to .04. The average of both is the same at 50% MWP. We can also check the quartile ranges of those populations, which would reflect the greater variability: in the 45-55 matchup, the 1st quartile would be .45 and the third quartile would be .55. In the 70-30 matchup format, it would be far wider at .30 and .70. I will test this later today and see what I find. I already know the variance is identical for the formats (.039 in both). But I have not tested the quartile ranges.
I can not believe for a second that Modern has so many matchups, close to 50-50's.
Does anybody have numbers like those?
Edit: I know those decks got banned, but you can find more numbers, in the following announcements.
For example, the current Mono red aggro or RB midrange decks, have mainly 50-50 matchups.
Edit2: I think you are all making a huge mistake. Instead of focusing on whether Standard is more skilltesting than modern is(which it is), you/we should focus on the diversity. Standard will never have that perfect diversity modern has.
The one balances each other out.
The last time we had large N MTGO data was 2015. At that time, no deck had matchups outside of the 45-55 range. Not a single one. Even super polarizing decks like Bogles and Infect were at 52% and 53% MWP respectively. I guarantee that if we had the same dataset today, we would see all top decks in the 45-55 overall MWP range. But we don't have those MTGO numbers so we can't test this.
I guess you haven't played the KCI vs Tron matchup then. To claim this one is not lopsided is craziness!
What about Tron vs Jeskai/Mardu Pyromancer?
What about GDS vs KCI?
What about Jeskai vs KCI, where Matt Nass said he won vs 5 Jeskai decks (5-0) during one GP of his?
Lantern vs any creature based aggro deck?
None of those are 50/50 or close. You should stop pretending everything is great in modern. I love it, but I guarantee all of those are lopsided matchups(at least 60-40 or more).
It's like I am hearing apples are oranges at the moment.
Modern makes up for this, by having the best diversity of all times though and that's it's strengh! It is great!
There definitely are lopsided matchups in Modern.
Turns vs. Tron.
Grishoalbrand vs. Tron.
Tron vs. GBx. Yes, I said it.
Grixis Control vs. Lantern.
Humans vs. Storm.
Burn vs. Storm.
Ad Nauseam vs. Tron.
In fact, there are so many that I've come to appreciate it somewhat when a matchup is somewhat close. But somewhat close in a linear meta often means the difference between someone mulliganing or not. It can come down to draws during the game. But then that happens in ANY meta, not just a linear one.
@idSurge - the difference between 40/60 and 45/55 is winning 8 of 20 matches or 9 of 20 matches. Over a long period of time, this could mean the difference of top 8ing and being able to play for 1st or not. I will use myself as an example. Recently in the past 8 months or so, outside of PPTQs, I have had a lot of trouble making the top 8. I usually am close near the end, but lose to the "guy who ends up making the top 8" instead of me. If I won just 1 more match in 20, I could easily see making more top 8s and I try super hard to finish since nothing feels like 1st place. (I will admit that my ratio of 2nd places to 1st places is something like 10 to 1 though, lol)
Besides, it's more the slight differences in how Pros play each matchup. Gaining 3-5% in every matchup around the board is what separates top 8 contenders from those who fall short. I know this, as I personally did it with Bogles, a terrible deck that did not deserve to get 2 PPTQ wins in 3 PPTQs for me. I just simply knew how to play it, knew matchups because I played and tested like crazy, and knew what to concentrate on and what to concede to. I knew what texture of each game/match would give me the best chances to win.
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)
@foodchainsgoblins, I agree to everything you said.
Forgot the storm vs humans matchup and storm vs burn. Those are top tier decks.
I probably shouldn't bring up less common decks (<1% of the meta), but #1, I know those matchups somewhat and #2, people play what they want to play, so many of us run into these "odd" types of decks sometimes and it kiiiiiills us. Imagine you're a Tron player who missed 3 Byes because you barely missed Gold this year. In round 3, your first round, you face Daniel Wong (quad sleeved turns deck/timewalkingonsunshine here at mtgs) and he absolutely obliterates you. It takes a strong person to put that behind them and grind the last 13 rounds.
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)
@foodchainsgoblins, I agree to everything you said.
Forgot the storm vs humans matchup and storm vs burn. Those are top tier decks.
I probably shouldn't bring up less common decks (<1% of the meta), but #1, I know those matchups somewhat and #2, people play what they want to play, so many of us run into these "odd" types of decks sometimes and it kiiiiiills us. Imagine you're a Tron player who missed 3 Byes because you barely missed Gold this year. In round 3, your first round, you face Daniel Wong (quad sleeved turns deck/timewalkingonsunshine here at mtgs) and he absolutely obliterates you. It takes a strong person to put that behind them and grind the last 13 rounds.
Yeah...the solution really is to get rid of the byes. Now that widespread GP trials are gone, the only purpose they serve is to help get pros deeper into tournaments. Zero value for WOTC.
I have nothing to back this up, just a gut feeling from anecdotal "evidence." But perhaps the overall, smoothed, aggregate MWPs are similar for very different reasons. For example, it's no secret that Modern has swingy matchups that can be incredibly one-sided, and require the stars to align in order to go the opposite of expectation. That generates variance in WP based on something outside your control. In Standard, many of the games are kind of grindy midrange matches, where even "aggro" and "control" can end up in stalls that turn into top deck wars. So whoever draws the card they need off the top of their deck first wins. This generates variance in WP based on something outside of your control as well.
Again, nothing hard to base this on (especially not being super familiar with *CURRENT* Standard meta, but something noticed from recent past Standard metas. The difference could be simplified (correctly or not) to:
Standard: Low matchup variance, high card draw/top deck variance (due to poor cantrips and minimal redundant effects), skill used to outplay opponents in close games and adds large % boost.
Modern: High matchup variance, relatively low card draw/top deck variance (due to abundance of cantrips and decks running multiple copies of redundant effects), skill used to make card/deck choices as well as fine nuances to add a noticeable, but smaller % to outcome.
I definitely don't know for sure, but it definitely irks me to see the comparison made that Modern has the "same" variance as Standard, without delving into the attributing factors or WHY.
I think its questionable in the extreme to even bother trying (without numbers to prove it) to show that Modern's wider pool of decks, the 'match up lotto' is a significant factor at this point.
Is it a factor? Sure. Is it a MEANINGFUL factor? Considering all the actual numbers ktk has provided?
I finished a little more aggregation for the top 10 players in the world, comparing their 2017-2018 performance in Standard and Modern. I thought this would be more productive than talking hypothetically about whether or not Modern is more high variance and/or lower skill than Standard. The net Standard MWP for this group of players is 63%. The net Modern MWP for this group of players is 67%. There is no significant difference in those samples and I believe most of the 4% points difference is in the relatively greater number of Standard events relative to Modern events. By all accounts, their MWP in both formats is identical.
Great work on this, especially considering no one else wants to run the numbers (myself included). I understand using confidence bounds to make sure we're only considering statistically relevant results, but it seems a bit off to assume that a 4% difference means they're essentially the same. Being a pedant, your results would show we can't say with confidence that the formats have different MWP%; that doesn't mean we can confidently say they're the same.
Put another way, how large would the difference in MWP have to be to confidently say they're significantly different? Because I have a feeling, given the sample size, that could be a number approaching a 10% difference between formats or more. Which is an astronomical value given the floor for pros in a format is essentially a 50% MWP.
TLDR: Pedantic question -- if we can't comfortably say a 4% difference in MWP is relevant using the available data, that could mean we're working from too small a sample size to reach significantly granular conclusions?
Private Mod Note
():
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
There's something else that people are forgetting.
*Players like Matt Nass realize that they can't make it to too many GPs, so they took the preparation super seriously. Then they had some luck and good play along the ride.
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)Spirits
I've noticed that some of the people on my team are not as helpful to me as to others, but still just hearing what's going on in their heads quite often helps me towards preparation. I mean if someone has a compelling argument for not playing Titanshift, I wouldn't have done it. I mean, in the end, you have to make up your own mind, but you can bounce ideas off other experienced players. I would rather test vs. my team members 1 day a week than play 2 weekly tournaments. The results are just much more helpful in many ways. (I mean, it's more "fun" to play in 2 weekly events, but it's really not helping much toward GP play.)
You gotta understand this. Just talking and play testing with someone like Paulo Vitor Damo Da Rosa every week is going to improve your game. Anyone here would do much better with that as the only "change" they have. I don't have access to that, so I try to top 8 local tournaments where in the top 8, I can face solid GP competition.
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)No deck since (again, unless it was during the period I wasnt paying attention to) has broken through in force, except...JDS? Which was due to a MTGO grinder, the pro's just picked up that individuals deck if I remember right.
JDS had its time in the sun. GDS had its time. Etc, etc. Nothing has broken the format, no meta game is really breaking out to abuse, at any level that we as followers of the format can not also identify...but then you may not even play against it!
As cfusionpm mentioned, when they played the side events of GP Vegas, no Tron, no KCI, all 'fair' midrange type decks, and UR Breach was a great choice.
The 'meta' was Humans, UWR, and Tron though, so how does that happen?
You cannot meta game modern at such a fine level, Pro or not.
Spirits
They also get to dodge the true out-of-nowhere matchups that could happen in rounds 1-3. But success breeds more success. Needing to go 10-2 is alot easier than 13-2.
http://www.cubetutor.com/viewcube/13649 - My all foil cube.
I can get behind this, for sure. Especially the importance of byes, but no, I'm not at all convinced that even the testing of tournament groups, has any great relevance. Unless something is truly broken, teams always seem to come to different conclusions, because no, even pro's (as noted in Ross' article) are biased.
Spirits
Again, this sort of comment is not an indictment of Modern or any Modern dynamics. This is merely a commentary on the importance of byes in all events.
At this point, we have a very clear picture of how top players succeed in all formats. We've seen the same effect on both the SCG and the GP/PT circuit at this point; top players perform consistently in Modern as they do in other formats, and they do so at the same level of performance. So now the argument has shifted away from pro players to average players without byes. This is unsurprising as the Modern criticisms constantly shift when confronted with new data. I am positive that if we broke down the numbers, we would find that the effect of byes/no byes in Standard/Modern is statistically identical. I say this with confidence because basically everything else has been identical between the formats after we break down the numbers, so I see no reason why this won't be the case either. If I'm wrong, I'll happily admit it, and I'll check some numbers to see if it's true. I also am fairly certain that after I run this analysis, and if I find it to be the case that byes matter the same in all formats, some of the critics will AGAIN shift arguments again to show that Modern is high variance, low skill, matchup dependent, etc. Either these things don't exist in Modern OR the best players have found ways to overcome them.
I firmly believe that most of this returns to players who are dissatisfied with their personal performance in Modern and are looking for things to blame. It is easier to blame these external elements, up until they are disproven, than it is to focus on personal areas of improvement. I would 100% buy these criticisms if we didn't see such consistent success by the best players. But we do, and that success is true across formats and in the same degree.
I realized it's pretty easy to compare standings vs. byes with the GP data we have, so I just threw this together tonight. I wanted to see if byes have a greater, lesser, or equal effect on players in Modern vs. Standard. To do so, I looked at standings vs. number of byes for the first five Modern and Standard GP of 2018. For Modern, this was Sao Paulo, Barcelona, Las Vegas, Hartford, and Phoenix. For Standard, this was Pittsburgh, Singapore, Copenhagen, Birmingham, and Seattle. For each event, I found the total number of byes for all players in the first three rounds. I then took the average of their overall points broken down in each of four categories: players with 1, 2, 3, and 0 byes respectively. I did this for all events, averaged them within formats, and then compared them between formats. Here is what I found:
Modern - 0 Byes: 9.78 points
Modern - 1 Bye: 14.77 points
Modern - 2 Byes: 18.29 points
Modern - 3 Byes: 26.52 points
Standard - 0 Byes: 9.85 points
Standard - 1 Bye: 14.58 points
Standard - 2 Byes: 18.79 points
Standard - 3 Byes: 26.39 points
In all categories, there is less than a .5 point difference between the point standings in Modern and Standard. If you have 3 byes in Standard, you average 26.39 points at an event. In Modern, you average 26.52 points. Got 0 byes? In Standard, you'll average 9.85 points. In Modern, you'll average 9.78 points. Needless to say, these differences are not statistically significant at all and are barely different in absolute numbers.
This all but categorically disproves the allegation that byes matter more in Modern because you can dodge bad matchups and have better performance. Byes have an IDENTICAL effect in Standard or Modern. If byes mattered more in Modern, we would see players with more byes doing better in Modern than in Standard. We don't see this at all. Instead, we see players with the same number of byes averaging the same standings as those across format lines. This also supports the argument that the formats are comparably skill-testing from a performance standard, although I admit the required skills may have differences as well as overlaps. Between this analysis and the previous one I posted, we now see that top players have consistent performance in both Modern and Standard, and we now see that byes have an identical effect between formats. We can certainly spin this into an indictment of the disproportionate effect of byes at top events; I'm all for that. But I think we can safely move beyond this untenable notion that byes matter more in Modern because they help people dodge high variance matchups early. Byes have the same effect in Standard and Modern on your win rate at a GP.
So...next topic? Control is fine, Modern is no less skill testing, and byes are a huge advantage regardless of format, while we know Wizards feels Modern is healthy.
Other than my eternal quest to find a true replacement of Twin, I'm not sure what else I can even gripe about, besides Tron/Stirrings...
Spirits
It doesn't prove the argument. It just supports it. The common claim, at least from certain critics, is that Standard or Legacy requires more skill than Modern. And yet, in every analysis I've posted here when we actually compare formats, we find that the formats all have the same numbers. Top players have the same top MWP. The MWP ceilings are the same. They have consistent performance between events. Byes matter equally between formats. There is simply no performance evidence that Modern is any more or less skillful than any other format. If Modern was a less skill-intensive format than Legacy or Standard, we would see evidence of this in at least one of these areas. We would see byes matter more in Modern than Standard. We would see a lower MWP ceiling (if top players were victim to variance) or a higher MWP ceiling (if the format was easier to win at). We would see pros at GP do worse in Modern than in Standard (if variance determined their matchups) or better (if Modern was easier to succeed at). There would be some evidence of this, and yet, we don't see it at all. All of these analyses push harder and harder against the notion that Modern requires less skill than Standard.
Are there any numbers-based arguments for Modern requiring less skill? Those arguments are something like 0 for 5 at this point, and if there are more of them I'd love to see and test them.
the 'from a performance standpoint' rider is the key to his statement.
let me clarify for you:
modern is probably more skill intensive than standard. (more decks, more cards, more variety, shorter games mean individual decisions matter more in your average game etc. etc.). I have no statistical basis for this, but it seems intuitive based on what we know... so... "probably".
What's important though, and what KTK was saying, is that this doesn't affect relative performance between formats.
simplifying it; I guess you'd say that a decent player who puts the time into mastering Standard would have an equal amount of success in Modern if they replicated their efforts there. This is backed up again and again by looking at the spread of repeat-success from those pros and players who put the time in. Even on a local level, you'll probably see this, but it's harder to discuss because it would mostly be anecdotal.
What you may be alluding to is a potential difference between the learning curves of the two formats. I have no evidence to support this, but my own experiences have led me to consider that Modern probably has a steeper 'starting point' on the learning curve than Standard. Ultimately, they'll end up in the same place, but i have this intuition that Modern is initially a trickier format to start playing. Perhaps this difference might account for newer players to the format slinging around arguments about variance and blaming their losses on diversity etc. Who knows. That's wild speculation, but conceptually it seems to fit the mold and make sense.
this is actually a super interesting topic of discussion. =)
I'd argue something completely different from recent pages of discussion. Modern requires a significant amount more work to be effective in the format compared to Standard.
Skill in Magic can be measured in multiple ways, but what I find are two issues that arises frequently; The ability to recognize areas that need improvement, then acting upon to create those improvements. There are hundreds of players I have come across, both intelligent and somewhat so, who have difficulty in acknowledging their own failures. On the other side of the coin, there are people I find who are able to recognize their problems, but recreate them without any signs of improvement towards a solution.
It's why I find the skill argument lacking when comparing formats. It's just deeply layered problem solving.
Anyway let's evolve beyond the 'standard vs modern' dichotomy and talk another other stuff.
Anyone here like.. Weirdly excited to see infect back as a deck people are willing to play? Fundamentally I don't think it was ever bad, but it's certainly having a mini Renaissance and while I have never really played the deck for myself, it's way cool to see it back again.
I only played against infect a couple of times while on blue tron, so it was obviously a *****show not in my favor. I like it, though. It preys on tron and is one of the best linear decks at turn 3 kills, which lets it race things like storm and KCI. Yet, because it is creature based, it struggles against midrange and control (outside of the occasional "oops I win" opener). I'd consider building a GB variant of hierarch wasn't disgustingly expensive. I simply cannot justify spending more than $30 on an individual card.
Ya I mostly agree. I've tried blue moon, adnauseum, cheerios and other combo decks but I haven't found a deck I liked as much as twin. I've told myself for a long time now that I would be happy enough with the ban list if BBE and twin could just get unbanned. We are half way there!
It would be cool to have stuff like SFM, gsz or others unbanned but I'm not as motivated by any others. Sittings and tron are annoying but I'm not ready to take action against them yet.
I imagine this is the next frontier of the Modern criticism. Not saying this to throw any shade at you two, particularly at ThinkingChimp who phrases it as a question and not a reality. I'm just pretty sure this is where the matchup lottery, high variance, low skill argumentation is going to move next.
Thankfully, we already have the numbers to check this: we just need to look at MWPs in tournaments. If Modern had a lot of polarized matchups, we would see highly variable MWP over smaller samples of events, and then we would see those normalize over time to the same MWP bracket as Standard. Standard's would be normalized the entire time. We would be able to measure this in the variance of the population, which might have MWP that averages to the same as a Standard format, but would have greater variance overall.
For example, in a population of 20 events where the player alternates a 45% MWP and a 55% MWP, the variance is only .002. But in a population of 20 events where they alternate 70% and 30%, the variance explodes to .04. The average of both is the same at 50% MWP. We can also check the quartile ranges of those populations, which would reflect the greater variability: in the 45-55 matchup, the 1st quartile would be .45 and the third quartile would be .55. In the 70-30 matchup format, it would be far wider at .30 and .70. I will test this later today and see what I find. I already know the variance is identical for the formats (.039 in both). But I have not tested the quartile ranges.
The last time we had large N MTGO data was 2015. At that time, no deck had matchups outside of the 45-55 range. Not a single one. Even super polarizing decks like Bogles and Infect were at 52% and 53% MWP respectively. I guarantee that if we had the same dataset today, we would see all top decks in the 45-55 overall MWP range. But we don't have those MTGO numbers so we can't test this.
This is still a card game with variance.
I've beaten Dredge never drawing Rest in Peace and that match up 'feels' terrible. What we think are the numbers do not bear out to what ktk states.
Spirits
There definitely are lopsided matchups in Modern.
Turns vs. Tron.
Grishoalbrand vs. Tron.
Tron vs. GBx. Yes, I said it.
Grixis Control vs. Lantern.
Humans vs. Storm.
Burn vs. Storm.
Ad Nauseam vs. Tron.
In fact, there are so many that I've come to appreciate it somewhat when a matchup is somewhat close. But somewhat close in a linear meta often means the difference between someone mulliganing or not. It can come down to draws during the game. But then that happens in ANY meta, not just a linear one.
@idSurge - the difference between 40/60 and 45/55 is winning 8 of 20 matches or 9 of 20 matches. Over a long period of time, this could mean the difference of top 8ing and being able to play for 1st or not. I will use myself as an example. Recently in the past 8 months or so, outside of PPTQs, I have had a lot of trouble making the top 8. I usually am close near the end, but lose to the "guy who ends up making the top 8" instead of me. If I won just 1 more match in 20, I could easily see making more top 8s and I try super hard to finish since nothing feels like 1st place. (I will admit that my ratio of 2nd places to 1st places is something like 10 to 1 though, lol)
Besides, it's more the slight differences in how Pros play each matchup. Gaining 3-5% in every matchup around the board is what separates top 8 contenders from those who fall short. I know this, as I personally did it with Bogles, a terrible deck that did not deserve to get 2 PPTQ wins in 3 PPTQs for me. I just simply knew how to play it, knew matchups because I played and tested like crazy, and knew what to concentrate on and what to concede to. I knew what texture of each game/match would give me the best chances to win.
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)I probably shouldn't bring up less common decks (<1% of the meta), but #1, I know those matchups somewhat and #2, people play what they want to play, so many of us run into these "odd" types of decks sometimes and it kiiiiiills us. Imagine you're a Tron player who missed 3 Byes because you barely missed Gold this year. In round 3, your first round, you face Daniel Wong (quad sleeved turns deck/timewalkingonsunshine here at mtgs) and he absolutely obliterates you. It takes a strong person to put that behind them and grind the last 13 rounds.
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)Yeah...the solution really is to get rid of the byes. Now that widespread GP trials are gone, the only purpose they serve is to help get pros deeper into tournaments. Zero value for WOTC.
Again, nothing hard to base this on (especially not being super familiar with *CURRENT* Standard meta, but something noticed from recent past Standard metas. The difference could be simplified (correctly or not) to:
Standard: Low matchup variance, high card draw/top deck variance (due to poor cantrips and minimal redundant effects), skill used to outplay opponents in close games and adds large % boost.
Modern: High matchup variance, relatively low card draw/top deck variance (due to abundance of cantrips and decks running multiple copies of redundant effects), skill used to make card/deck choices as well as fine nuances to add a noticeable, but smaller % to outcome.
I definitely don't know for sure, but it definitely irks me to see the comparison made that Modern has the "same" variance as Standard, without delving into the attributing factors or WHY.
UR ....... WUBR ........... WB ............. RGW ........ UBR ....... WUB .... BGU
Spells / Blink & Combo / Token Grind / Dino Tribal / Draw Cards / Zombies / Reanimate
Is it a factor? Sure. Is it a MEANINGFUL factor? Considering all the actual numbers ktk has provided?
No.
Spirits
Great work on this, especially considering no one else wants to run the numbers (myself included). I understand using confidence bounds to make sure we're only considering statistically relevant results, but it seems a bit off to assume that a 4% difference means they're essentially the same. Being a pedant, your results would show we can't say with confidence that the formats have different MWP%; that doesn't mean we can confidently say they're the same.
Put another way, how large would the difference in MWP have to be to confidently say they're significantly different? Because I have a feeling, given the sample size, that could be a number approaching a 10% difference between formats or more. Which is an astronomical value given the floor for pros in a format is essentially a 50% MWP.
TLDR: Pedantic question -- if we can't comfortably say a 4% difference in MWP is relevant using the available data, that could mean we're working from too small a sample size to reach significantly granular conclusions?