TheManaDrain.com
May 25, 2018, 04:19:26 am *
Welcome, Guest. Please login or register.

Login with username, password and session length
News:
 
   Home   Help Search Calendar Login Register  
Pages: [1]
  Print  
Author Topic: Math on Bazaar and Serum Powder  (Read 10919 times)
gkraigher
Full Members
Basic User
***
Posts: 705


View Profile
« on: January 27, 2014, 08:23:58 pm »

I have been goldfishing with my dredge deck a lot and might consider piloting it (over my beloved MUD deck) in the near future.  

This is the current list:

4 Golgari-Grave Troll
4 Stinkweed Imp
1 GolgariThug
4 dark blast
3 ichorid
4 bloodghast
4 narcomebia
3 nature's claim (which i chose main over ingot chewer as it deals with artifacts and leyline of the void)
3 unmask
1 ancestral recall
4 bridge from below
4 serum powder
4 cabal therapy
1 dakmor salvage
4 bazaar of baghdad
4 undiscovered paradise
4 gemstone mines (to get around city in a bottle)
4 petrified field


sideboard:
3 leyline of sanctity  (a game 3 sideboard card if my opponent plays ravenous trap, tormod's crypt, and/or spellbomb.  it also shuts down oath of druids and is a nuisance for tendrils)
4 leyline of the void
4 ingot chewer
1 nature's claim
3 wispmare


I would like to fit a 4th unmask in the main deck and to do so I am considering cutting the 4th serum powder.

Does anyone know how much this lowers the chance of drawing a bazaar in my opening hand?  Are there links to the probability charts for this online somewhere?  

I know it has a negative impact, but I am wondering how much.  If my chances go down by less than 3%, that seems like a fair trade.

Another question would be the impact of a 61st card on those probabilities as well.
« Last Edit: January 28, 2014, 08:43:45 am by gkraigher » Logged
gkraigher
Full Members
Basic User
***
Posts: 705


View Profile
« Reply #1 on: January 27, 2014, 09:09:58 pm »

I was able to find this post that claims with 4 bazaar and 4 powder it is a 93% chance.  He included a link to a program.  I don't know how to run the program, but if someone could  to figure out 4 bazaar and 3 powder, as well as 4 bazaar 4 powder and 61 cards, that would be awesome.  It looks like an entry in the code. 


forum:
http://forums.mtgsalvation.com/showthread.php?t=513548

the code:
https://github.com/Dritte/VintageDre...ganToBazaar.py
« Last Edit: January 27, 2014, 09:23:46 pm by gkraigher » Logged
Smmenen
2007 Vintage World Champion
Adepts
Basic User
****
Posts: 6392


Smmenen
View Profile WWW
« Reply #2 on: January 27, 2014, 09:40:11 pm »

This math was done years ago.  Look at my article archive
Logged

Samoht
Adepts
Basic User
****
Posts: 1392


Team RST


View Profile Email
« Reply #3 on: January 27, 2014, 10:56:18 pm »

It's Serum Powder, not Serum Visions.
Logged

Char? Char you! I like the play.
-Randy Bueller

I swear I'll burn the city down to show you the light.

The best part of believe is the lie
gkraigher
Full Members
Basic User
***
Posts: 705


View Profile
« Reply #4 on: January 27, 2014, 10:58:28 pm »

Quote
It's Serum Powder, not Serum Visions.

noted, thanks
Logged
gkraigher
Full Members
Basic User
***
Posts: 705


View Profile
« Reply #5 on: January 28, 2014, 01:51:37 pm »

i had a friend run the program, it spit out these results for getting a bazaar:


I'm going to run it with the default values
 python mulliganToBazaar.py
0.941681291934

I'm going to run it with 120 cards in deck (test for accuracy)
 python mulliganToBazaar.py
0.682516208486

I'm going to run it with 61 cards
 python mulliganToBazaar.py
0.937710368968

I'm going to run it with 60 cards and 3 serums
 python mulliganToBazaar.py
0.925610134261

I'm going to run it with 60 cards and 2 serums
 python mulliganToBazaar.py
0.907321468248

I'm going to run it with 60 cards and 1 serums
 python mulliganToBazaar.py
0.887037447507


I'm going to run it with 60 cards and 0 serums
 python mulliganToBazaar.py
0.865044198724


the rate stays close to 1 out of every 14 games you don't get a bazaar.
Logged
vaughnbros
Basic User
**
Posts: 1574


View Profile Email
« Reply #6 on: January 28, 2014, 04:14:06 pm »

There are a couple of big problems with reducing your chances of finding bazaar:
1.  A no bazaar hand game 1 is pretty much an automatic loss.
2.  These calculations don't factor in how many cards you have in your opener when you finally find bazaar.  Having a hand of just bazaar and nothing else relevant gives your opponent a good chance of stealing that game off you.
Logged
gkraigher
Full Members
Basic User
***
Posts: 705


View Profile
« Reply #7 on: January 28, 2014, 04:24:47 pm »

vaughnbros,

I totally agree with you.  I just wanted to know the underlying odds of finding a bazaar with minor tweaks from the accepted shell of the deck. 

The fact that the odds of finding a bazaar only went down 1% going up to 61 cards is a really interesting development.  I'll need to run the simulator, but it seems like going to 62 cards is basically the same as running 60 and 3 powder. 

If tabernacle at pendrell vale starts seeing more and more play, it might makes sense to run 62 cards and 4 ichorid. 

I've always wondered the effects of playing 61 cards in any given deck.  Whether it was to add a basic land to a shops deck (to get around ghost quarter) or simply run a random 1 of gitaxian probe just to get info.  If your deck aggressively mulligans (and dredge turbo aggressivly mulligans) it appears the odds do not swing at a significant level.

 In fact, it basically swings at the same rate as playing 1 additional fetch land over some other land. 
Logged
vaughnbros
Basic User
**
Posts: 1574


View Profile Email
« Reply #8 on: January 28, 2014, 06:50:20 pm »

>1% seems significant to me.  There is really never a reason to play 61 cards since there is always a worst card in your deck.  If I could run 59 cards Id do it in a second...

You seem a little too overly concerned with the lesser hate cards.  Tabernacle, City in a bottle, and Tormod's/Nihil/Rav trap only slow you down and can't beat you without help.   If you commit too much to trying to deal with these cards you are going to be blown out by the more powerful hate cards, leyline/cage/RiP.  Or you are going to hurt your game 1 percentages, which is so crucial to winning any match, but especially against big blue and combo.
Logged
gkraigher
Full Members
Basic User
***
Posts: 705


View Profile
« Reply #9 on: January 28, 2014, 09:43:47 pm »

to play gemstone mine over city of brass to fight city in a bottle actually makes a lot more sense than "conventional" wisdom says.  You will never tap a city of brass for more than 3 mana in a game, so it is the same card and you get to keep the life.

leyline of sanctity is actually a great card, as it also hurts oath of druids and tendrils along with stopping graveyard hate.  If my graveyard never gets blown out by a crpyt, that is much more effective than forcing the issue with an ingot chewer.

The only card I don't run that many people do is chain of vapor.  A card that is a stickily worse answer to all the hate cards, as they get to recast them, and is only better if your opponent has a blightsteal in play.  

ichorid is the only way of beating a tabernacle without running cards like wasteland and strip mine.


Let's break down the math on 61 cards vs 60 cards.

I play in a 9 round tournament, with a top 8.  Assuming I go to the finals, here is the math assuming I play the most games possible (playing less actually lowers the final number):

9 rounds * 3 games per round = 27 games
top 8 = 3 rounds * 3 = 9 games

total 36 games.  

with 61 cards, I have a 93.7% of getting a bazaar, so I expect to get it in 33.73 games.

with 60 cards, I have a 94.3% of getting a bazaar, so I expect to get it in 33.84 games.  

That is .11 games in a full tourney difference.  Does that still seem significant to you?
« Last Edit: January 28, 2014, 10:01:55 pm by gkraigher » Logged
vaughnbros
Basic User
**
Posts: 1574


View Profile Email
« Reply #10 on: January 28, 2014, 10:03:39 pm »

I've played and won a lot of games with dredge... I am well aware of the power level of all the cards you are playing.

to play gemstone mine over city of brass to fight city in a bottle actually makes a lot more sense than "conventional" wisdom says.  You will never tap a city of brass for more than 3 mana in a game, so it is the same card and you get to keep the life.

But you do tap it for mana a lot more than 3 times most game 2 and 3's...

leyline of sanctity is actually a great card, as it also hurts oath of druids and tendrils along with stopping graveyard hate.  If my graveyard never gets blown out by a crpyt, that is much more effective than forcing the issue with an ingot chewer.

The problem is most decks only run 1 tormods/nihil spellbomb if any at all.  You have to devote 3 card slots in your deck for the chance to have it your opening hand to cancel out their singleton hate piece.  Its just not worth it.  3/4 Unmasks are more than enough to deal with that type of hate.  Against oath you are much better off with nature's claims and wispmares because they destroy oath and their leylines.  Against storm you are much better off with mental misstep because it stops them at the beginning of their combo not the end of their combo.  Misstep is also great against grafdigger's cage, deathrite shaman, pithing needle, nihil spellbomb, as well as many other 1 drops.

The only card I don't run that many people do is chain of vapor.  A card that is a stickily worse answer to all the hate cards, as they get to recast them, and only is better if your opponent has a blightsteal in play.  

I never advocated chain, but sometimes all you need is EoT bounce, use bazaar, untap use bazaar again to pull off a win.

ichorid is the only way of beating a tabernacle without running cards like wasteland and strip mine.

You don't at all.  I've beaten tabernacle more times than I can count without even having ichorid in my deck.  You can pay 1 to keep a creature, so essentially every land you have gives you another creature to fight through it.  When your opponent is at 10 or below your bloodghasts also gain haste.

Let's break down the math on 61 cards vs 60 cards.

I play in a 9 round tournament, with a top 8.  Assuming I go to the finals, here is the math assuming I play the most games possible (playing less actually lowers the final number):

9 rounds * 3 games per round = 27 games
top 8 = 3 rounds * 3 = 9 games

total 36 games. 

with 61 cards, I have a 93.7% of getting a bazaar, so I expect to get it in 33.73 games.

with 60 cards, I have a 94.3% of getting a bazaar, so I expect to get it in 33.84 games. 

That is .11 games in a full tourney difference.  Does that still seem significant to you?

So what you are saying is your deck is worse at 61 cards, but not significantly worse?  There is no benefit to playing 61 cards at all.  The 61st card is by definition the worst card of your deck.  Why play it all if you don't have to?  I'd be drawing it in my opening 7 nearly 10% of games, so in your example that's 4 times I'll see the worst card of my deck in my opening hand instead of potentially a better one.
« Last Edit: January 28, 2014, 10:11:40 pm by vaughnbros » Logged
gkraigher
Full Members
Basic User
***
Posts: 705


View Profile
« Reply #11 on: January 29, 2014, 11:10:04 am »

Vaughnbros,

I appreciate the posts a lot.  I am still trying to figure this deck out and all your comments are very helpful.  

I originally had misstep in the board, and I do love that card.  You are probably right, I should not be playing leyline of sanctity when this is still a valuable card.  I will find a way to run 4 unmasks to deal with the type of hate, like you recommend.  

I am surprised to hear you have won games against tabernacle without ichorid.  That seems very difficult to me, considering all they need to do is have a blocker.  I view tabernacle as a silver bullet against the deck, then thought about what I could do if one was in play.  I realize the bloodghasts get there eventually, but to get there is super difficult, especially without ichorid.  The only decks that run tabernacle seem to be shops, maybe landstill and combo could play it also.

I must misunderstand how this deck is played in games 2 and 3 after the sideboard, because I still cannot see how tapping a city of brass more than 3 times actually happens (without a tabernacle in play).  
Logged
msg67183
Basic User
**
Posts: 929



View Profile Email
« Reply #12 on: January 29, 2014, 11:14:14 am »

Vaughnbros,

I appreciate the posts a lot.  I am still trying to figure this deck out and all your comments are very helpful.  

I originally had misstep in the board, and I do love that card.  You are probably right, I should not be playing leyline of sanctity when this is still a valuable card.  I will find a way to run 4 unmasks to deal with the type of hate, like you recommend.  

I am surprised to hear you have won games against tabernacle without ichorid.  That seems very difficult to me, considering all they need to do is have a blocker.  I view tabernacle as a silver bullet against the deck, then thought about what I could do if one was in play.  I realize the bloodghasts get there eventually, but to get there is super difficult, especially without ichorid.  The only decks that run tabernacle seem to be shops, maybe landstill and combo could play it also.

I must misunderstand how this deck is played in games 2 and 3 after the sideboard, because I still cannot see how tapping a city of brass more than 3 times actually happens (without a tabernacle in play).  

You are casting anti hate, as well as casting creatures through their hate. When I used to play Dredge I found that playing a very Disruptive version with Mental Missteps and Unmasks won games simply by preventing the hate from hitting the table.
Logged

Bloomsburg Tournaments:

1 Win
3 Finals
2 Top 4
2 Top 8

Outside Bloomsburg:

Winter Grudge Match lV Top 4

Creator of The Mana Drain Vintage League.

Website for The League:

http://tmdvl.github.io

Zombies ate your brains!
vaughnbros
Basic User
**
Posts: 1574


View Profile Email
« Reply #13 on: January 29, 2014, 11:22:45 am »

I must misunderstand how this deck is played in games 2 and 3 after the sideboard, because I still cannot see how tapping a city of brass more than 3 times actually happens (without a tabernacle in play).  

Basically I end up winning almost half of my game 2/3's by hard casting bloodghasts, stinkweed imps, and the like.  Then the other half actually answering their hate and beating them with bazaar.  I wrote this a year ago: http://www.themanadrain.com/index.php?topic=44964.0  It might give you some tips as to some lines of play and some other options for card choices.
Logged
gkraigher
Full Members
Basic User
***
Posts: 705


View Profile
« Reply #14 on: January 29, 2014, 12:19:27 pm »

That is awesome.  I never thought to not play golgari grave troll and play more thugs.  makes a ton of sense, especially with 4 unmasks.  the interaction with ingot chewer is really awesome. 

City of brass clearly gets the nod over gemstone mine again.  They only have 1 city in a bottle anyway and you still have paradise to get rid of it. 

What do you think about the main deck choice of nature's claim over ingot chewer?  I know chewer is better against a lot of matchups, but nature's claim does enough and can allow you to get around a leyline of the void some dredge player could be running main. 
« Last Edit: January 29, 2014, 12:22:11 pm by gkraigher » Logged
vaughnbros
Basic User
**
Posts: 1574


View Profile Email
« Reply #15 on: January 29, 2014, 02:37:36 pm »

What do you think about the main deck choice of nature's claim over ingot chewer?  I know chewer is better against a lot of matchups, but nature's claim does enough and can allow you to get around a leyline of the void some dredge player could be running main. 

I'd say the main difference is claim is better against oath while chewer can be used to get bridge triggers.  Both require you to have mana game 1, which makes them generally not very useful until games 2/3.
Logged
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Valid XHTML 1.0! Valid CSS!
Page created in 0.063 seconds with 20 queries.