|
6 months ago ::
Nov 26, 2012 - 11:55AM
#81
|
|
|
Please stop referring to the one time when you kept 4/5/6 cards and you ended up winning all the same. No one is saying it isn't possible, only that it isn't probable based on the data collected and an abstract understading of the game. When you talk about single games there is no way to properly gauge the strength of your opponent. It would be like having a football team play with 10 men on the field instead of 11 and then observing how well they did. If they were playing a pee-wee team they might end up winning all the same. Against a pro team they would almost certainly lose, even if they were another pro team themselves.
1. I hope the please stop comment isn't in reference to me specifically because I'm pretty sure this is the first time I have replied to one of these posts of yours.
2. Pretty hypocritical to say, "stop referring to the time..." when you very frequently use anecdotes from recent matches you have played in your posts. I get that you weren't saying it's impossible, only improbable. The point I was trying to make was that you made it sound like it's pretty much a done deal if you mull and your opponent doesn't and I had recently had multiple instances of success to the contrary.
Overall, I agree that 27% is too low, but I'm not totally convinced that is the correct number for the majority of Magic players (non-pro) and I don't think your data is robust enough to make a strong enough case. Although, I don't think there is a good way to get that kind of data in a deterministically relevant way and any sort of conclusion on the issue will probably have to involve some guess work and rely entirely on pure statistics.
3. Funny that you use that analogy and talk about not knowing the strength of opponents considering you play almost exclusively Swiss drafts and numerous people have confirmed that the skill level there is quite a bit lower than 8-4s.
I'm not trying to tell you to stop thinking this way or to give up on something you feel very passionate about. I generally find reading your threads and opinions pretty entertaining. I think the reason people get annoyed with you is because of how often you bring these things up and more importantly how you tend to blow off/shoot down any suggestions and basically act like you are right now matter what and just looking for vindication/people to agree with you.
|
|
|
|
6 months ago ::
Nov 26, 2012 - 3:24PM
#82
|
Date Joined:
Jun 15, 2009
|
1) The please stop comment is directed at everyone in general. I do not feel that you personally have used anecdotes too often.
2) I use anecdotes because they make my writing more colorful, I don't believe they prove anything.
3) I enjoy Swiss for a number of reasons, but my skill level is high enough I could play 8-4s if I wanted to. The last time I played 8-4 I drafted a killer mid-range deck and lost two games in round two to a speedy Rakdos deck to end the tournament with a record of 1-1 and no prizes. This experience fairly well encapsulates what I dislike about playing 8-4.
See, now the anecdotes are in Italics.
As for why it may appear like I'm shooting down/blowing off suggestions... it's generally because I've already considered and discarded those positions prior to posting about the matter online. I'm not looking for people to echo back my own positions, I'm looking for someone to find something I missed.
|
|
|
|
6 months ago ::
Nov 26, 2012 - 3:51PM
#83
|
Date Joined:
Jul 15, 2012
|
Those were all of the Limited GP Top 8 Games from 2010-2012 where one player kept 7 cards and their opponent kept less than 7. I could look at 2009 to push the number of games examined above 100, but there was very little variation in win rates for the three years I've already looked at.
I mean that the number of games is not idicative of the whole spectrum. You need something like 1000 games. Also, because you are looking at GP Top 8, your results could be biased due to the nature of your sample.
|
|
|
|
6 months ago ::
Nov 26, 2012 - 4:10PM
#84
|
Date Joined:
Jun 15, 2009
|
Those were all of the Limited GP Top 8 Games from 2010-2012 where one player kept 7 cards and their opponent kept less than 7. I could look at 2009 to push the number of games examined above 100, but there was very little variation in win rates for the three years I've already looked at.
I mean that the number of games is not idicative of the whole spectrum. You need something like 1000 games. Also, because you are looking at GP Top 8, your results could be biased due to the nature of your sample.
I would be much happier having access to a database of online game results, but so far no one has given that to me.
|
|
|
|
6 months ago ::
Nov 26, 2012 - 7:12PM
#85
|
Date Joined:
Apr 14, 2008
|
So, it was hit on earlier, and this thread is something beyond a joke at this point, but lets take this back to an earlier point of contention:
The selection for when the mull occurs and why it hurts needs to be addressed.
Decks with a janky base, etc., are going to mull more. Those decks are more likely to lose and are also more likely to be controling decks that'll get blown out by fast aggro more often. 3 color decks are also more likely to hate their 6 card hand.
As such, winning 27% of time time isn't valuable as a flat statistic, there is too many recursive effects, especially at times when janky manabases are tempting.
And, y'know, there's the point where we're at least an order of magnitude off on having a meaningfull sample size.
Luckily, we have stop-having-fun guys to remind us that having anything more than 60 cards in your deck is tantamount to being a rapist and anyone considering it should be strung up by their ****  .
|
|
|
|
6 months ago ::
Nov 26, 2012 - 7:36PM
#86
|
Date Joined:
Jun 15, 2009
|
I know what you are saying regarding the likelihood of mulliganing being tied to mana base, but the results I collected were from the Limited GPs Top 8s from 2010-2012. A simple evaluation of the deck listings will show the vast majority of the decks were 2 color, 17 land, 23 spell splits. Reading through the tournament reports you will see that the majority of hands that were mulliganed were mulliganed due to having the wrong number of lands.
So no, deck construction cannot explain the variance in game results that is explained by mulligans/starting hand size.
As for meaningful sample size, that depends on how wide you want your confidence interval to be. I could go further back in time to gather more samples, but the average win rates for muliganers vs people keeping their starting 7 varies remarkably little from year to year.
|
|
|
|
6 months ago ::
Nov 28, 2012 - 7:10PM
#87
|
Date Joined:
Jun 15, 2009
|
Out of curiosity I looked at the 2010-2012 Standard GPs and found essentially the same results as for the Limited GPs. There the people keeping their starting 7 were winning 52-21 for a 71.2% win rate. Or conversly the people mulliganing their starting hands are winning 28.8% of the time. Which is odd because someone else said they looked at high level Constructed events and found the mulliganers collectively winning 38% of the time and the 6v7 matchup winning nearly 50% of the time. I guess I'll have to get to the bottom of this.
|
|
|
|
6 months ago ::
Nov 28, 2012 - 10:38PM
#88
|
Date Joined:
Sep 10, 2006
|
Get to the bottom of what? Convoluted logic?
|
|
|
|
6 months ago ::
Nov 29, 2012 - 4:21AM
#89
|
Date Joined:
Jun 15, 2009
|
For the last time ShadowC, this was looking at tournament results and counting how often a person keeping 7 cards won against an opponent who kept less than 7 cards. There's nothing convoluted or difficult about this, it is a straight observation.
|
|
|
