20 cards sideboard, not a bad idea?

[Nessaja]

If it's not proportional to the size of the metagame - which I agree with, and find a good thing, you have to prepare against what you expect, no?

No, because the amount of decks you expect exceeds the amount you can prepare for. It's pretty obvious if you don't read that sentence out of context.

Unexpected decks = no problem.
Expected decks that can't be prepared for just because there are more than 10 strategies to prepare for is potentially a problem.

[Grollub]

No, because the amount of decks you expect exceeds the amount you can prepare for. It's pretty obvious if you don't read that sentence out of context.

Unexpected decks = no problem.
Expected decks that can't be prepared for just because there are more than 10 strategies to prepare for is potentially a problem.

But wouldn't that just be a calculated risk assessment? You're preparing for what you believe to be the most likely matches.

I think I can follow your argument and understand your point, thank you for your patience, tho I respectfully disagree that this in fact is a problem. Our disagreement appears to be based on what you consider getting an unlucky match up (an exogenous variable) I consider getting burned on a judgement call (an endogenous variable).

[Nessaja]

I agree that this is calculated risk assessment. But then you get at the point where eventually, there are several matchups that you could prepare for and they're all more or less equivalent in terms of metagame pressence.

Like in the example with 8 decks that each cover 10% of the metagame and two decks that cover 9%. Your sideboard prepares you for 8 decks, which of the 10 decks do you choose to prepare for? You pick 8 decks and you hope you'll face those decks in the tournament, statistics tell you it's more likely to meet one of those 8 matchups over the 2 matchups you chose to let go, but that doesn't hold up when it's round 1 and you're facing one of those two matchups.

I don't call that endogenous because in principle, you didn't make a wrong metacall. You had to make a choice based on chance and you made the right call based on those statistics. So you lost against statistics, not neccesarily against skill.

Note that preparing, in this example, is not getting a 90% or better matchup against those decks. With preparing I mean having a decent chance bringing the matchup to a 50% matchup and let skill (and some luck) decide.

[Grollub]

But that's under the assumption that the metagame's decks frequency is distributed uniformly, which I find a rather courageous assumption - I haven't checked the latest coverages of the large tournaments, so please correct me if I'm wrong here.

If the metagame in fact is a uniform smattering of decks, I would agree that luck* can be a very deciding factor -- in which case I'd rather see dredge banned to free up sideboard slots than adding more. ;-)

* but considering we're playing a game with a randomized deck against another randomized deck, luck will always be a factor and is in my opinion one of the charms in regards to watching and playing MtG; watching a game in the finals getting shifted around due to an insane (perhaps series of) topdecks.

[Nessaja]

Well it's mostly under the assumption that I'm lazy and just made a simple model to illustrate my problem to bring the point across.

Even when decks are 3/4/5% you want to prepare for the decks that are the most common and then eventually, you need to make a choice and this choice can work out or not. Even when you made the correct choice in terms of metagame pressence, you can still very likely lose. Sure this is still the case with a 20 card sideboard but a little less so.

I agree that luck is a core part of the game, I just want the pendulum to swing a little away from the luck side because, in my opinion, the sheer size of the legacy metagame made the pendulum swing a bit towards the luck side.

[dontbiteitholmes]

Well it's mostly under the assumption that I'm lazy and just made a simple model to illustrate my problem to bring the point across.

Even when decks are 3/4/5% you want to prepare for the decks that are the most common and then eventually, you need to make a choice and this choice can work out or not. Even when you made the correct choice in terms of metagame pressence, you can still very likely lose. Sure this is still the case with a 20 card sideboard but a little less so.

I agree that luck is a core part of the game, I just want the pendulum to swing a little away from the luck side because, in my opinion, the sheer size of the legacy metagame made the pendulum swing a bit towards the luck side.

If this was true we wouldn't see the same people consistently top 8/16ing most of the Legacy events they enter.

[SpikeyMikey]

If this was true we wouldn't see the same people consistently top 8/16ing most of the Legacy events they enter.

You see this at small events because there's a very tight expected meta and because there is such a discrepancy in skill levels between players. At higher level events, you still see it, but not as much (yeah, pretty much every major event has 1-4 "names" in the T8, but there are dozens of "name" players out there) and there, a lot of it can be attributed to byes. Say it's round 8 of a 9 round SCG Open. I'm 6-2, on the bubble and I'm playing Gerry Thompson. Gerry is actually 3-2 but he's got 3 byes, so we're in the same bracket. Basically, it means that in a 9 round Legacy, I need to win 78% of my matches while Gerry has to win 67%. 8 rounds, I need 75%, he needs 60%. 15% can account for a lot of variance.

[John Cox]

You see this at small events because there's a very tight expected meta and because there is such a discrepancy in skill levels between players. At higher level events, you still see it, but not as much (yeah, pretty much every major event has 1-4 "names" in the T8, but there are dozens of "name" players out there) and there, a lot of it can be attributed to byes. Say it's round 8 of a 9 round SCG Open. I'm 6-2, on the bubble and I'm playing Gerry Thompson. Gerry is actually 3-2 but he's got 3 byes, so we're in the same bracket. Basically, it means that in a 9 round Legacy, I need to win 78% of my matches while Gerry has to win 67%. 8 rounds, I need 75%, he needs 60%. 15% can account for a lot of variance.

I top 8 at small (20-25 person) tournaments regularly, these tournaments have a completely random meta mainly due to most of the players having access to multiple decks. There is a skill discrepancy but after 2-3 rounds or so that completely disappears as all the bad players are x-1/ 0-2 and the good players are in the top bracket.
I disagree that it has so much to do with byes at larger events. I somehow doubt that all large events reward name players with byes and that that is why those players regularly make top 8 solely because of said byes.

[Admiral_Arzar]

You see this at small events because there's a very tight expected meta and because there is such a discrepancy in skill levels between players. At higher level events, you still see it, but not as much (yeah, pretty much every major event has 1-4 "names" in the T8, but there are dozens of "name" players out there) and there, a lot of it can be attributed to byes. Say it's round 8 of a 9 round SCG Open. I'm 6-2, on the bubble and I'm playing Gerry Thompson. Gerry is actually 3-2 but he's got 3 byes, so we're in the same bracket. Basically, it means that in a 9 round Legacy, I need to win 78% of my matches while Gerry has to win 67%. 8 rounds, I need 75%, he needs 60%. 15% can account for a lot of variance.

Note that the pros also generally dodge all the random decks that people are complaining about in this thread, as those decks have generally entered the loser's bracket by the time round 4 rolls around. Thus they only have to sideboard for the decks to beat and rarely get screwed by random pairings. I honestly think the bye system is flawed for this reason as it gives a huge advantage to certain players in 1-day large events like the SCG opens.

[thefringthing]

I'm a fan of the idea of making sideboards a fixed percentage of the maindeck size (rounded down, say). A 60-card maindeck gets 15 sideboard cards, so an 80-card deck could have a sideboard of 20 cards. (Perhaps the factor of a quarter could be changed to some other fraction.)