The LGS I play it runs a monthly modern 1k of almost exclusively local guys, very few travelers, considering pushing for a coalition for using the chess p/d system, anyone have experience/advice for running an event with that system?
The play/draw analogue in chess is white/black, and it is fairly easy.
The article is a bit technical but gives you all the information you need. Here is some of the most important:
"5.
Colour Allocation:
5.1
In the first round the colour assigned to player No.1 shall be decided by drawing a lot. All other odd numbered players in the top half of the initial list shall receive the same colour.
5.2
The difference of the number of black and the number of white games shall not be greater than 2 or less than –2.
5.3
A player shall not have the same colour three times in a row.
5.4
After pairing two players’ colours shall be assigned based on giving descending priority to:
giving both players their due colour
equalizing the numbers of black and white games played before
alternating the colours of both players regarding the first difference of their colour history going back from the previous round to the first round.
assigning his due colour to the player with the higher SB
assigning his due colour to the player with the higher Buchholz, and so on."
It's a bit hard to explain how to do it by hand in text form, but with any chess tournament organizing software you can do play/draw assignments and pairings automatically.
To summarize: Start the event by seeding all players (randomly given no ELO or other score to seed with) and randomly assigning player 1 the play/draw. Then, each odd numbered seed also gets the same allocation, and the even numbers the opposite. Then the pairings just need to be readjusted every round so that you follow the alternating play/draw system (5.4) while keeping each player in their SCORE GROUP (i.e the group of players who have the same score).
It would be, if not for people in this thread ignoring the data that is counter to their view that 'the BGx vs Tron matchup is almost unwinnable' (or similar statements).
For the scenario described, you either disagree or you agree with the assertions in the evaluation. It's not that complicated. Yes, we can get into the minutiae of evaluations having multiple conclusions that can be addressed separately, but that isn't what I was getting at. I was criticizing the dismissal of evidence on the basis that it disagrees with personal experience/anecdotes. That is not a rational position to take. If you disagree with an evaluation, that is fine. But you need supporting evidence, and anecdotes are insufficient,(edit) in order for your disagreement to be rational.
'Considering the data' is what happens while you choose to accept or refute it. I'm not clear how that is a third option. Holding that 'the data is insufficient to reach that particular conclusion' is refuting it btw. Your refutation could even be erroneous. But either way, you accept or refute.
1 Appealing to the self-proclaimed 'Mad Staticitian' is, itself, an appeal to authority. Kinda rings hallow against your little screed against such a "mistake".
2 Human beings are not rational actors so bringing it up in a game of hidden information is nonsense for robots.
3 Even if they were you've presented a false choice because you can do more than Refute or Accept new information.
1)I referenced the evaluation and data, and credited the source. I did not appeal to any individual as an authority. You are mistaken.
2)I'm sorry that you do not value being a rational actor. I do not view that as a criticism of myself, but rather a rather odd admission on your part. You do you, neighbor.
3)Sure, I'm not perfect either. What is the 3rd option for a rational actor? Is there a 4th option? I would like this concept to be fleshed out some more, I'm interested in this criticism.
It's more that in several years, not one single author or pro has ever claimed it is even or close to even, and is universally panned as atrocious.
This is a common mistake. Authorities can and will be wrong - even for prolonged periods. See many historical scientific discoveries that overturned centuries of 'common sense' and 'authority opinion'.
When someone does an evaluation and produces numbers that describe the relationship, one of two things should happen for any rational actor:
1)Refute the evaluation, or
2)Accept the paradigm described by the evalution.
One of those two things needs to happen, or you are not being rational. Other than statements from presumed authorities, what evidence do you have that refutes the data gathered by Ktkenshinx?
Re: Jund vs. Tron
Again, I don't know what the % is today, but I know with statistical certainty it was roughly 50/50 for about 5 months of 2015 during a time we had mined MTGO data for that matchup. I also knew many players said it was "unwinnable" in that same time period, despite the numbers clearly showing it was actually 50/50. If Reid said it was worse in that time period, Reid was wrong. I don't know what it looks like today but I expect it's similar. Overall, people are bad at guessing their matchups; there's way too much bias and opportunity to willfully or unknowingly lie to yourself and others.
Side note: Abzan is definitely worse against Tron than Jund. I think it was around 35/65.
I've been saying similar things RE: UW control for a long time. I think we are seeing the Dunning-Krueger effect at work here. People can't beat a matchup, and therefore conclude that it is unwinnable. Have a 25% record against tron? The matchup must be 25%.
For years while piloting UW control, I would hear about how "unwinnable" the tron/valakut/ad naus/storm matchups were. Then I would play them and discover that there was a great deal of play to the matchups, and they are actually fairly close with tight play.
It's not easy to see when a matchup is 55/45 vs 60/40 vs 65/35. I think we can safely conclude that those arguing against the numbers are simply applying their own experience to others and assuming it is universal. If the numbers show something different than one expects, it's time to re-evaluate the underlying assumptions imo.
That may be true in theory but I have found in practice that even if you know your deck better than your opponent. Ex. (BGx vs Tron). I think most people can agree that the BGx deck needs to clock opponent fast enough and be just disruptive enough so that the backbreaking spells don't come online. That knowledge goes out the window against god hands that even a new person signing up for a DCI number can pilot mindlessly.
Are you trying to say something of value here? Because 'god hands are hard/impossible to beat' sounds like a tautology to me. Yeah, god hands are called god hands because they are well... godly. This is saying basically nothing since god-hands are such a small % of draws. Are you trying to argue that god-hands are bad for the influence of skill the outcome of a game? I doubt that statement is terribly controversial (almost too obvious to point out really).
Your matchup knowledge only helps when your opponents keep loose or have average or below average hands.
I'm glad you acknowledge that most games of modern are about skill: "average or below average hands" are most games.
I think the "helps a top-tier deck" metric is problematic. Most people seem to interchange "helps a top-tier deck" with 'shares a colour with a top-tier deck' when they consider these things, and I wonder how much WotC does as well. A card like BBE or SFM would likely show in testing only a marginal gain at best for the decks that include them, calling into question just how much 'help' is a reasonable floor when used as a justification to remain banned.
But I doubt WotC would even make it to a testing phase where a card could prove itself innocuous if it shares a colour with a deck somewhere in the top tier. Which is strange because all but the most closed modern metas have had some kind of representation of each colour,
What about this tournament gives the impression that control isn't viable? What do you mean by viable? I didn't get to watch much, and don't see a control deck in the top8 - is that the basis for this statement?
On related topic: does the presence of abzan in the top8 make SFM also unlikely? Seems that the logic of BBE might apply to BGx because they can splash any colour as desired.
That being said, I doubt we could ever see a top8 totally absent of decks that could splash SFM, so is that even a realistic metric to use to decide if a card could be unbanned? Or maybe that dynamic implies that it was never an option.
A statistical analysis will tell you the difference in advantage / disadvantage for going first vs second, but that point is moot: there will always be some difference. If that difference is not controlled, the variance will influence events.
But, as I outlined already, you can fix it by making sure people alternate play / draw each round so that the difference is experienced equally (to the extent possible). This can be done without measuring the difference since it addresses any size difference. Once that is done, you can move on to other ways of addressing it if it still seems like a concern. Controlling the variance of who gets the advantage/disadvantage of going first will marginalize that variance from the event.
You don't want a random assignment imo (we have that now, it's just determined by dice rather than software). Instead, you want players to alternate games where they are the play / draw each round. The system should pair according to swiss rules, modified to allow same-score pairings to be altered to fix play-play pairings and draw-draw pairings.
IMO, the goal is to have play/draw alternate each round for each player. This is easily achieved in software used in other games. I bet I could even just use the chess software to pair a magic tournament to alter play - draw for each player (as much as possible while running a swiss tournament).
Of course you want to minimize the difference in experience between games on the play vs games on the draw, but you'll never achieve perfect parity. No matter what, the players should be alternating whether they are on the play or draw each round since there will always be some advantage to going first (or second!).
The chess pairing software takes into account who is on the play, and prints it out on the same sheet as the table pairings. There is no reason that this can't be done for magic - they do use pairing software after all. In chess there is a structure of rules to decide who gets to play first, and the software helps by repairing optimally to match those who last had opposite starts. This can all be done automatically and easily. I could do it by hand - I've done it enough by hand in chess that I could do it for a magic event fairly easily.
edit: In chess, if a game is played where players used the wrong colours, the game stands. Both parties are responsible for looking at their table number, I don't see why play/draw would be different.
You see this most commonly with G/W, G/W Company, GW, GW Company, Bant, Bant Company, Counters Company, Counter Company, Knightfall, Bant Knightfall, G/W Hate Bears, GW Hatebears, etc etc ...
This work is more detailed than it might seem at first, and the work is tedious to parse which Knightfall list has the Devoted Druid combo in it and which ones don't. Thanks for the effort!
[...]Shouldn't we get to have a real primer, instead of the actual "compendium" of cards in the opening page? I think this could be beneficial for the community, as most of the lists proposed are old. We're already past the Finks/RestoAngel era.
Of course, I'm by no means denigrating wpgstevo@'s work, and I thank him for opening the thread. But, personally speaking, I'd like to have a better insight of strategy/variations/lists/match-ups in the primer.[...]
I mentioned in a recent previous post that I intended to do an update, and asked for suggestions. In fact, I've asked for suggestions for the OP a number of times and never received a single one. If you have formatting suggestions, I'm all ears. Lets take suggestions to PMs though, so that the thread isn't about how the OP should look.
edit: Note almost all the cards listed in the OP have been taken directly from lists with results, along with some others that have been attempted at various times (Thopter/sword, for instance).
If you have suggestions on updating the OP, I'd like input! Send me some PMs!
Here is fide's document on how to do pairings: https://www.fide.com/component/handbook/?id=85&view=article
The article is a bit technical but gives you all the information you need. Here is some of the most important:
"5.
Colour Allocation:
5.1
In the first round the colour assigned to player No.1 shall be decided by drawing a lot. All other odd numbered players in the top half of the initial list shall receive the same colour.
5.2
The difference of the number of black and the number of white games shall not be greater than 2 or less than –2.
5.3
A player shall not have the same colour three times in a row.
5.4
After pairing two players’ colours shall be assigned based on giving descending priority to:
giving both players their due colour
equalizing the numbers of black and white games played before
alternating the colours of both players regarding the first difference of their colour history going back from the previous round to the first round.
assigning his due colour to the player with the higher SB
assigning his due colour to the player with the higher Buchholz, and so on."
It's a bit hard to explain how to do it by hand in text form, but with any chess tournament organizing software you can do play/draw assignments and pairings automatically.
To summarize: Start the event by seeding all players (randomly given no ELO or other score to seed with) and randomly assigning player 1 the play/draw. Then, each odd numbered seed also gets the same allocation, and the even numbers the opposite. Then the pairings just need to be readjusted every round so that you follow the alternating play/draw system (5.4) while keeping each player in their SCORE GROUP (i.e the group of players who have the same score).
For the scenario described, you either disagree or you agree with the assertions in the evaluation. It's not that complicated. Yes, we can get into the minutiae of evaluations having multiple conclusions that can be addressed separately, but that isn't what I was getting at. I was criticizing the dismissal of evidence on the basis that it disagrees with personal experience/anecdotes. That is not a rational position to take. If you disagree with an evaluation, that is fine. But you need supporting evidence, and anecdotes are insufficient,(edit) in order for your disagreement to be rational.
2)I'm sorry that you do not value being a rational actor. I do not view that as a criticism of myself, but rather a rather odd admission on your part. You do you, neighbor.
3)Sure, I'm not perfect either. What is the 3rd option for a rational actor? Is there a 4th option? I would like this concept to be fleshed out some more, I'm interested in this criticism.
When someone does an evaluation and produces numbers that describe the relationship, one of two things should happen for any rational actor:
1)Refute the evaluation, or
2)Accept the paradigm described by the evalution.
One of those two things needs to happen, or you are not being rational. Other than statements from presumed authorities, what evidence do you have that refutes the data gathered by Ktkenshinx?
For years while piloting UW control, I would hear about how "unwinnable" the tron/valakut/ad naus/storm matchups were. Then I would play them and discover that there was a great deal of play to the matchups, and they are actually fairly close with tight play.
It's not easy to see when a matchup is 55/45 vs 60/40 vs 65/35. I think we can safely conclude that those arguing against the numbers are simply applying their own experience to others and assuming it is universal. If the numbers show something different than one expects, it's time to re-evaluate the underlying assumptions imo.
I'm glad you acknowledge that most games of modern are about skill: "average or below average hands" are most games.
But I doubt WotC would even make it to a testing phase where a card could prove itself innocuous if it shares a colour with a deck somewhere in the top tier. Which is strange because all but the most closed modern metas have had some kind of representation of each colour,
On related topic: does the presence of abzan in the top8 make SFM also unlikely? Seems that the logic of BBE might apply to BGx because they can splash any colour as desired.
That being said, I doubt we could ever see a top8 totally absent of decks that could splash SFM, so is that even a realistic metric to use to decide if a card could be unbanned? Or maybe that dynamic implies that it was never an option.
But, as I outlined already, you can fix it by making sure people alternate play / draw each round so that the difference is experienced equally (to the extent possible). This can be done without measuring the difference since it addresses any size difference. Once that is done, you can move on to other ways of addressing it if it still seems like a concern. Controlling the variance of who gets the advantage/disadvantage of going first will marginalize that variance from the event.
IMO, the goal is to have play/draw alternate each round for each player. This is easily achieved in software used in other games. I bet I could even just use the chess software to pair a magic tournament to alter play - draw for each player (as much as possible while running a swiss tournament).
Of course you want to minimize the difference in experience between games on the play vs games on the draw, but you'll never achieve perfect parity. No matter what, the players should be alternating whether they are on the play or draw each round since there will always be some advantage to going first (or second!).
edit: In chess, if a game is played where players used the wrong colours, the game stands. Both parties are responsible for looking at their table number, I don't see why play/draw would be different.
edit: Note almost all the cards listed in the OP have been taken directly from lists with results, along with some others that have been attempted at various times (Thopter/sword, for instance).
If you have suggestions on updating the OP, I'd like input! Send me some PMs!