- mehungary
- Registered User
-
Member for 15 years, 6 months, and 27 days
Last active Sun, Jan, 17 2021 12:55:36
- 0 Followers
- 716 Total Posts
- 18 Thanks
-
Aug 26, 2014mehungary posted a message on Off Topic: Gaming the $ystem for ValueI agree that it is the best way to get a whole playset of a set. However in doing my own method of purchasing just what i want plus a playset of uncom and comm i can get within a budget of about 150$. Since i have no desire to have every single card from each set and that 150 is 3 times cheaper than the mtgo method. Also some times mtgo is just a expensive as the paper set, when buying singles. I did a full set price comparison for journey into nyx with mtgo and paper, and found that it was more expensive than paper with chaff removed (***** rares and mythics). If you really want a whole set then yes its great, if you don't want the whole set then there are cheaper and faster options. Also if your willing to forgo the mythic's you can get a full 4x card set for about the same price as the mtgo method its not as good but you get almost everything with less waiting. I did much thinking about this method and decided against it simply because you really need to commit a lot of money in order to get value, and i can get what i need for less and avoid the hassle. That said if i really love a set i will definitely use the mtgo method, but i will really need to like the set.Posted in: Articles
-
Feb 3, 2014mehungary posted a message on Launch Giveaway!My Favorite card has to be from my first deck from time spiral. Jedit Ojanen of Efrava I just love the flavor as well as how awesome he was when i played against my brothers green decks.Posted in: Announcements
- To post a comment, please login or register a new account.
Chosen because I find the animations awesome and hilarious.
I had an edited Dakkon previously but with the switch to the new forum format the background was the wrong color. So I used Snip to find something new and interesting.
The long term player - Lots of decks, lots of cards, Lots of odd and strange stuff. Old enough to be a mature player, and played enough games to make the ego of winning a minor motivation. These players are still playing the game because they like playing the game most of all. These players also have a variety of decks that are at different power levels so that they can change decks to ensure parity.
The entrant - Few cards, 1-3 decks. They originate from anywhere. They are still adapting to the culture and style of a group or the format.
The social player - Plays with friends, might not even own their own cards. Willing to play and enjoys playing the game. Focuses on multiplayer games and only plays with known people.
The Psychographic profiles, With Timmies, Johnnies, and Vorthos being the largest groups. Spike is a vocal group on these forums, however I have yet to actually play with someone who identified as that in EDH irl.
You have 4 player matchups, with potentially ~30 decks to test so nCr = 27405 matchups. I am using combinations without repetition for each match up. The reason why the match ups are not compressed further is because we assume that 4 player games have the participants making contextual choices relevant for each combination. We need atleast 30 of each of these matchups to even begin considering them as actually meaningful (100 or more tests would be better). So about 822150 games are required to test, and if each game takes about at best case .5 hours, thats still 411075 hours to test the matchups. Then once you have all this data you will need to test the null hypothesis that the mean win rate for each matchup is = 25%, if we reject the null hypothesis then in that matchup its not just chance that determines the winner. If we fail to reject the null then the winner of those matchups are determined via chance. Then we can also test to see if the win rate is consistent between matchups, by testing the null hypothesis that the two or more are equal, if we reject this then the deck performs differently in different matchups, where if we fail to reject the null then the deck performs the same regardless of matchup. From these tests we can then confirm real patterns with supporting evidence from which a tiered list can be extrapolated from.
For this solving the format attempt, we would assume that each deck wins by chance until proven otherwise.
Our starting point would be population probability = .25 and population Variance = 0.1875 which we assume from the null hypothesis.
Now for an even bigger problem to know the next level of the meta, let's limit to only one deck per comander. So 392 decks total. The number of possible matchups with 4 player games is nCr = 968870630 matchups. We apply the same rules as before and we get, 29066118900 games and 14533059450 hours of playing at best case.
Now for the entirety of the decks that are in the database. 8.7*10^14 matchups, 2.61*10^16 games for significance, and 1.3*10^16 hours at a fast pace.
So now that I have framed the scales of actually solving the format you might realize that in order to get significant results from the first model you would need to devote about 62% of an average human's lifespan, or about 10 weeks of 1000 people working full time in order to have something that would be a significant amount of information. This is the reason why people have not done this in any meaningful way. The problem is just bigger than it is worth to solve. One reason why 1v1 is mostly solved is because with 1 deck per general we only get 76636 combinations in the 4 player model we have 12642.5 times more than 1v1. Looking at the ~30 decks suggested here, for 1v1 we get 435 matchups our 4 player matchups are 63 times more numerous. The scale for 1v1 is tiny compared to 4 player multiplayer, not to also mention that in multiplayer there is more than just the decks themselves being played which would be omitted with just testing the win rates. We also have a problem of bias within the data because if we have a small group of player doing this homework we will not be able to claim that our samples are random, as the fewer people who participate the greater their personal bias will be in the data, which is something we would want to avoid if we wanted to have meaningful data with regards to the meta instead of individual player skill.
Now you could use less than 30 samples per matchup however you would have to use a t distribution instead of a normal distribution, which would make it harder to reject the null hypothesis; that they are just winning by chance. The fewer samples the lower the degrees of freedom and thus the harder it would be to reject the null at a significant level. So in order to show that your matchup was not just chance but was the deck itself that won you would need enough samples to make it easier to reject the null. The whole point of this solving the format is to show that the win rates are not just random chance, because with no distinction from random chance there is no discernible metagame. And in order to have confidence that there is a metagame we need to have enough evidence to say that it is not just random chance. Now that i have explained why such a large amount of games per matchup need to be played in order to more easily show that there is in fact a metagame, it becomes a large problem of actually playing those games and eliminating player bias. Which in itself is a huge task. The conundrum of actually doing this task is apparent hopefully by now if you have read this far.
One alternative method would be to use a regression model. It would not show if the win rates were chance alone but it could show the relationship between that of a deck with relation to all the other decks. The scale would be at minimum which would not be the best 30^2 or 900 games with 30 games for each deck. One issue is that the whole thing is probability and as a result you would want to likely be running a probit or logit model. These models would provide useful predictions although the coefficients of the variables would not be directly readable, so unless you knew how to interpret the model it would not serve any use. There would also be other issues such as removing insignificant variables, and dealing with omitted variables bias, heteroskedasticity, correlation, and others, and manipulating the model to fix these problems if they arise as well as refine the model for better readability.
In either method I personally do not see enough value to justify doing the work to make them useful. What value is there? For the scale of the problem and how much effort is required to meet the minimum standards so we can solve the commander 4 player meta. How much is this metagame worth? not much, not enough to spend 411075 hours playing games to get enough data to test if there is even a meta or if it is all chance.
TL:DR
Solving the multiplayer metagame with accuracy and significance is not worth the time and resources it would take to do so as it is many many times larger than that of the 1v1 metagame on the same scale. Thus the 1v1 metagame is likely the best estimation of the multiplayer meta simply because it is measurable within a reasonable amount of work and scale for this format.
An uncontrollable interaction where the effects repeat endlessly resulting in a draw unless stopped, as an infinite combo.
I would define combo as:
A Controllable interaction of card(s) that can be repeated many times, possibly with the potential to be repeated forever but not required, and the goal to end the game, either directly or enabling a win condition. It usually involves short cutting. I would consider this to be, just combo, as it must resolve and thus can not be truly infinite. I would also state that all this must happen in a single turn, or be repeatable effects that can virtually lock others out of the game permanently such as, repeatable extra turns, mindslaver effects, and brine elemental + Vesuvan Shapeshifter effects ect.
Combo, control, and LD are all considered anti causal methods of play. But making a casual tournament by banning these methods of playing while still having the same win conditions as normal doesn't dissuade people from pursuing an effective way to win. If the tournament organizers want to have a casual thing then they would need to have disincentives for building the best deck(s) that win in standard methods. I think that a points system that includes things other than wins as means to obtain prizes makes a decent incentive for players to build decks that don't focus solely on winning the match. Hopefully incentives to build to earn points from things other than winning matches provides enough of a disincentive to play the best deck.
If I were to make a casual friendly tournament:
I would ban all tutors (or have a small cap). I would also ban all combos with the total CMC of all the cards in the combo less than 10, so any combo that has a combined cmc greater or equal to 10 would be legal. Then I would add a points system where you can "win" through other methods than the conventional way. I would then operate a case by case banning of cards to control the meta to ensure parity between players. For the points system I would need to brainstorm with the playgroup or community to ensure that they feel the added rules for winning prizes are fair and encourage them to play in a way that makes the game as interesting as possible while also making the game accessible to new players, or players who are unwilling to spend the money/time required to be competitive. It would be interesting to note that it would be likely that unless the players are all casual and not interested in metagaming that these rules would also expands to reduce the use of other "unfriendly to casual" forms of winning such as control or LD. Another idea would be that points would have diminishing returns. Earning points in only one section becomes less effective each time they are earned, encouraging a player to change their decks to earn points in other areas as well.
An example of this system would be points for; Style or theme, Winning, Kills, Most creatures played, ect.. Where each time you earn points from one of these sections you earn less the next time. At the end of the tournament your points determine the reward(s).
Personally I find the casual tournament idea to be silly, as they are somewhat contradictory. This might be very difficult to pull off with an open group. A closed group i think would be more successful at having the idea work.
Sisay a very nice option as well, although some of the best cards you can run in that deck can eat up your money fast, making him more expensive to play. Although as many other people have pointed out his toolbox nature is very powerful. Sisay is a powerful yet boring general, likened to Zur or Momir vig decks where they play mostly the same every game with long turns of solitary and searching.
My bias is for Dralnu. Sisay is boring imo. Although other GW generals would be fun. I'm currently making a Tolsimir Wolfblood deck.
Would i like a type of removal to stop annoying commanders for most of the game? yes! Is tuck the best method? well it was the only method... I suppose in a general sense never having to fear the permanent loss of your commander is a good thing, and a bad thing depending on what commanders your play and face.
XColorlessX = Kozilek, Butcher of Truth
WWhiteW = Darien, King of Kjeldor - Just my fav white weenie deck
UBlueU = Muzzio, Visionary Architect
BBlackB = Chainer, Dementia Master
RRedR = Zirilan of the Claw - Work in progress
GGreenG = Kamahl, Fist of Krosa
WUAzoriusUW = Grand Arbiter Augustin IV
WBOrzhovBW = Teysa, Orzhov Scion - Nasty token combo board control
WRBorosRW = Aurelia, the Warleader
WGSelesnyaGW = Tolsimir Wolfblood - Surprisingly easy to make huge creatures fast and for low cost.
UBDimirBU = Dralnu, Lich Lord - Value for days. Lots of flashback and copies
URIzzetRU = Jhoira of the Ghitu
UGSimicGU = Momir Vig, Simic Visionary - Tutor + combo, what's not to like
BRRakdosRB = Kaervek the Merciless
BGGolgariGB = Savra, Queen of the Golgari - Dredge + mass revival, combos really nice with greater good
RGGruulGR = Ulasht, the Hate Seed
WUBEsperBUW = Sydri, Galvanic Genius - Did i mention i love artifact decks because I love this deck and its artifacts
WUGBantGUW = Rafig of the Many - Bant good stuff
WRGNayaGRW = Mayael the Anima - Naya monsters
UBRGrixisRBU = Thraximundar - Grixis good stuff
BRGJundGRB = Sek'Kuar, Deathkeeper
RWUJeskaiUWR = Narset, Enlightened Master
WRBMarduBRW = Zurgo Helmsmasher
BWGAbzanGWB = Ghave, Guru of Spores - Thalids and saprolings for fun
GUBSultaiBUG = Vorosh, the Hunter - Proliferate madness also mostly for silliness
RUGTemurGUR = Riku of Two Reflections - Spell slinging madness, nothing like an 80+ dmg fireball copied for each opponent out of nowhere
WUBRGPentacolorGRBUW = Sliver Overlord - A sliver toolbox, only missing the queen
My vote is Dralnu because inspite of the drawback he can be terrifying if he doesn't die to removal right away. Tasigur, the Golden Fang looks like a nice general to build around a self mill theme. Sidisi, Brood Tyrant I think would work well with him depending upon the build. Maybe get a zombie or two then get a nonland card back of your opponent's choice back with both out. They are vastly different cards and they might synergize well together in the same deck although there are already plenty of options that bring sorceries/instants and creatures back to your hand or into play/cast, 4 mana for a repeatable Mausoleum Turnkey effect still seems meh to me. The opponents choice thing makes it kind of sucky especially since they have lots of choice due to more types, so you'll get a card that you don't necessarily want for 4 mana. Why not use Arcanis the Omnipotent or Archivist or the other 100's of repeatable "draw a card(s)" cards because the effect is similar to not getting what you want but still getting something for an activated cost. If you're going for self mill then there are also better options such as dredge cards. Still these ideas wouldn't stop me from building whatever deck i wanted with him as the general if i really wanted to use him.
A nasty combo with Tasigur: I would use mindslaver on my turn, then activate Tasigur on your opponent's turn. While you are taking his turn you get to choose what goes back to your hand, guess what it is... mindslaver! because you get to make all the choices that the other player would make during their normal turn. So with this 2 card combo with one being the general and the other easily tutored with black and blue Tasigur could be a scary general.
My reasoning of why i never got involved with any competitive format. The cost for entry was very high but the rewards were the same as casual. Or more accurately put, I don't get enough added happiness beyond casual (which is cheap to play) for what competitive costs to play (hugely expensive). Especially when compared to other competitive luxury entertainments such as video games, movies, or hanging out with friends. In fact these days MTG is becoming unattractive even for casual play when I ponder my opportunity cost against the other forms of entertainment. I honestly wonder how 5$ pack of mtg cards provides me more entertainment than a 5$ indie game. However 3 packs for 15$ for a draft is about the cost of a movie in the theater and provides about the same length of entertainment in time, so a draft is time effective entertainment. Now is a 400$ EDH deck worth the money? for me no. I would need to get 80 hours of playtime with just that deck to justify the cost in relation to a draft or a movie. If each game is an hour then you would need to play that deck 80 times to get a comparable return, and if you play the deck only 3 times a weekend (or 3 hours a weekend) you would need to use that deck for 27 weeks in order to be competitive with the other forms of entertainment. I do not play that often and also don't have just one deck, but have 16 EDH decks. Now most of my decks are around 200-100$ so I only need to play them for about 30 -/+ 5 hours each (about 30 games) which is a good amount before i get bored of the deck and move onto another.
So no I don't have any anxieties about MTG as a hobby and the costs of playing it, because I make decisions based upon not just the cost of the cards in my budget but also the costs of the other things that I can enjoy with that money. I will never own a goyf or a JTMS simply because for their cost they do not offer me enough enjoyment to justify the purchase. While Thalids and other odd and strange decks offer me cheap decks that also entertain me at a price that is efficient with my time and money when compared to other things. My conclusion will be different from other people's conclusions, because of how we value MTG and the other things that compete with it will be different.
So that is an interesting result that in a four player game where all else is held equal the likelihood of winning is 1/4. This is the expected value for actual winning percentages. So 80% win rate seems extreme when compared to the expected win percentage. Its not impossible but I would be very skeptical of anyone who claims a win rate that is far from the expected value. IMO 80% win rate is a lie. If you look at the stats from WOTC regarding win/loss ratios the highest level players have only a win rate above 50% near 60% (And this is for 1v1 games). So the even best players don't seem to travel very far from the expected value. Skill only increases your chances of winning by a maximum of ~10%.
Applying a very skilled player to the formula is adding another variable where the probability of not losing at any one of the stages is increased by 10%? which means that P(winning and Skilled) ~= .41 or 41% so The best players at this game will likely have a win ratio of 41% in a 4 player games, however this may be an inflated number because we know that the general probability of winning is 1/4 or 25% and to have a 10% higher chance may not directly be translated to the addition of more players. Meaning that i don't know the actual impact that skill has in this scenario. For all i know the factor of a skilled player could range from 25% (no effect) to 41% (very skilled) win rate. This could depend upon the disparity of the skill levels of all the players in the game, where in a 1v1 scenario its easier to see the difference in skill levels. While in a 4 player game there are more variables because there are more players. If they are all highly skilled players then they will be closer to the 25%. If one player is highly skilled and the rest are poorly skilled then a 41% win ratio for one of the players makes sense. A player that has a win ratio higher than 41% in 4 player ffa games with at least 20-30 or more games played, is an anomaly.
The thing with probabilities is that they show the most likely events that may happen but they not guarantee that an unlikely event can't happen. Although the more unlikely the event the more justifiably skeptical you can be at the idea. Such that I would consider that having a win ratio greater than 35% in this scenario to be an overestimation or a boastful lie. And if you make a claim that you achieve an extreme win percentage I would still doubt you unless you had more than at least 20-30 games of data to back up your claim. Less than that number is not considered to be statistically relevant. If you make a claim based upon insufficient data then you do not have a truthful representation of your claim making your statements a lie, or at least a grossly inaccurate claim as fact which is essentially a lie.
TLDR: 80% win rates in multiplayer games is almost impossible. 25% plus/minus 10% is a range that is a believable win rate for 4 player multiplayer ffa games. 25% win rate is the most likely win percentage if all else is equal and win rates should not be too far from this number to be believable.
While its all subjective i do enjoy the silliness if group hug from time to time just because it creates rare and unusual games, i wouldn't want to play with them all the time.
Side note:
This is how most of this Thread reads and its funny to me.
Magic is a serious game and must be taken seriously! How dare those hippo players do something odd and indirect! KingMAKERS! Not playing the game My way! ManaDoublers! Howling Mines! Unbalancing effects! Chaos Effects!
Casual is no different than other sports or competitive things when parity is involved, watchers and players have more enjoyment when the match outcome is uncertain.
Since casual is not regulated in an official way, and neither does its amatuer nature force people to find an optimized metagame, the players must self regulate their small group. This self regulation is the biggest hurdle I find for competitive players to understand when they move to casual play. The reason why self regulation is so huge in casual and nonexistent in competitive is because of choice. Casual players retain the right to refuse to play against anything they do not want to and do not enjoy, while competitive players do not have the choice to refuse what the metagame is playing without giving up playing entirely. Also Casual has the ability to shape the metagame not by what wins the most, but by a social contract of unofficial and local rules and taboos. Competitive formats do not have an etiquette of socially acceptable strategies, rather the organization who owns the events controls the acceptable strategies via ban lists, the players have no common agreement between themselves but instead accept the will of an authority. The thing with casual is that it is socially complex while competitive is not. In order to play in a casual group you must adapt socially to what that group does and how the individuals in that group feel about certain things. Casual only has one driving factor for play, entertainment. Bringing back parity which in other sports is shown to make games more interesting the closer the two teams are be a likely winner. Thus for casual, if you feel that your deck will crush, or be utterly crushed you need to adapt. If you find that your deck is offending reasonable people, find a way to play that is less so in order to maintain a list of players who are willing to play with you. After all casual isn't to be played by youself and other players are a scarce resource required to enjoy the game.
For the most part in order to solve the difficulty of matching power levels when the people I play with are diverse as well, i bring many decks. A variety of decks at differing levels of strength from Zur to Darien, provides me the ability to adjust the power required to make the game as interesting as possible and at the same time also allowing me to change to a different deck if the one i just played was considered to oppressive.
Personally I find it odd that in my old group i had the silly weird decks that won sometimes, and now those same decks in my new group are powerful and scary. Sometimes I tell them that they aren't that good. Othertimes its because i am using older cards that the new people haven't even heard of, and not knowing for them is the same as scary and powerful. A recent example is when i played meekstone, sure it slows down big aggro but its not too hard to deal with, however it put a target on my head for a while, which confused me due to two other players having large token armies. Lucky for me meekstone bought me time to win that game, so in the end, i suppose it really was something scary lol.