Balance - It's here.

39 posts / 0 new
Last post

I tracked my games over the weekend and found some pretty interesting results.  I played 123 games with a random deck (with a few extra games played as GG, PI, ED after the beer kicked in) and put into a spread sheet every deck's wins and losses.  

The only deck that came out WAY above or below average was Goblins.  Everything else won or lost within 10% points of 50%.  I got this figure by applying a won or loss to each deck of each game, so there is a total of 246 data points.  For example, if I play BoF and defeat PK, then BoF gets +1 win and PK gets +1 loss.  I know this is still small-sample-size territory, but the initial results are encouraging from a balance perspective.  

I actually tracked all the games according to matchup, but here are just the final totals.

Edit...updated with info from other players!  Thanks!!!  Keep sending me info and I'll keep updating.  

Now it is a close match between Goblins/OD/BoF with everything else falling into a 2nd Tier and DP being the worst.





























































































































UsThemTotal
WLWLWL
Ancient Wilds632781044.44%
Born of Flame238612292059.18%
Celestial Light82310111247.83%
Crosswinds8439111345.83%
Dream Puppets6111271335.00%
Exalted Darkness105518152339.47%
Goblin Gangland172915261760.47%
Obiedient Dead295316322160.38%
Pack Instinct108615162341.03%
Peacekeepers75510121544.44%



While I admire you taking the time to test this, here is the problem. First off not everyone plays with the exact same builds as you do. Not only some players are more experienced than others and do not use the same strategies. Unfortunately, there results are only going to apply to you.

While I admire you taking the time to test this, here is the problem. First off not everyone plays with the exact same builds as you do. Not only some players are more experienced than others and do not use the same strategies. Unfortunately, there results are only going to apply to you.


I know all that.  Not really arguing otherwise. It is anecdotal.  Just thought I would post my results for discussion purposes.  


While I admire you taking the time to test this, here is the problem. First off not everyone plays with the exact same builds as you do. Not only some players are more experienced than others and do not use the same strategies. Unfortunately, there results are only going to apply to you.





This.

Unfortunately statistics never really prove anything.  They also tend to avoid explaining in detail how they were formed, and people often infer their own random ideas as to why the results are what they are.

They are also usually biased and based on a sample that would differ from another sample, and don't include all the variables (in this case, deck builds and level of skill in each opponent).


My results over the last 4 days would be completely different.  I'm 38-6, my opponents 6-38.  Goblins would only have 1 victory in 6 games, whereas Pack Instinct would have 1 in 4 games. 

Skill and build matter way more than the deck balance itself in these type of scenerios.


My results over the last 4 days would be completely different.  I'm 38-6, my opponents 6-38.  Goblins would only have 1 victory in 6 games, whereas Pack Instinct would have 1 in 4 games. 

Skill and build matter way more than the deck balance itself in these type of scenerios.





Post your results and I'll mix them with my results and thus slightly dilute the bias of the sample size.  If we get enough people that post their own results we can get a more accutrate sample of the decks with a smaller margin of error to account for player skill and customization.  

 

Deck                  Me                          Opp
Goblins              0-0                           1-7
Odric                  0-0                           1-4 
Exalted               0-0                           0-5
Obedient Dead   21-3                         1-3
Mill                      0-0                           0-5
Ancient Wilds      0-0                           1-3
Crosswinds         0-0                           0-4
Born Of Flame     17-3                         1-2
Pack Instinct       0-0                           1-3
Ajani                   0-0                            0-2 

Totals
OD - 22-6
BoF - 18-5
Goblins - 1-7
Odric - 1-4
AW - 1-3
PI - 1-3
Exalted - 0-5
Mill - 0-5 
Crosswins - 0-4
Ajani - 0-2 
Interesting. The two decks most famous for not allowing your opponent to play spells/effects on your turn.
Yeah, I don't think it's news that after the promo unlocks, Obedient Dead is the best deck overall (though not necesarily by a large margin).

Try as they might, DOTP is still very much a creature dominanted format.  OD simply has more and better answers to that fact than any other deck, and with the promos it now also has a very good selection of its own game ending creature threats (Phyrexian Obliterator, Dread, Massacre Wurm, Grave Titan, etc.).
Try as they might, DOTP is still very much a creature dominanted format.


Just as an aside, considering 60+% of Magic: The Gathering's 10,000+ cards are creatures, this is insanely hard to avoid.
ILl gladly add my numbers when i have some got a day off this wed so ill take some down and post them here.
I am Black/Green
I am Black/Green
Take The Magic Dual Colour Test - Beta today!
Created with Rum and Monkey's Personality Test Generator.
I am both selfish and instinctive. I value growth and community, as long as they favour my own objectives; I enjoy nature, and I particularly enjoy watching parts of nature die. At best
I don't really find it surprising that the decks packed with the most kill stuff cards are (for lack of a better term...so dont freak out) the best when it comes to winning.
I don't really find it surprising that the decks packed with the most kill stuff cards are (for lack of a better term...so dont freak out) the best when it comes to winning.




And yet, if you ask the general forum user "What is the best deck?" or "What is the most overpowered deck?", I bet the vast majority say Gobbos.

The decks I face the least?  Born of Flame, Obedient Dead and Ajani's Lifegain.  So either people are playing ranked games and PURPOSELY not choosing 2 of the best decks, or the average player hasn't actually figured out how to win with the control versions of those 2 decks.

Having played magic, I am about 99% sure it's not that people are not playing them because they are too good.  The overabundance of Knights last iteration makes that abundantly clear. 
I think Goblins is a general favorite because (depending on build) you are very unlikely to be mana short.  Which only leaves being mana flooded for a 'worst-case' draw.  I would think people on ranked (or who take thier rank seriously) would try to remove as much luck from the equation as possible.  Goblins just does that by the nature of the deck.
The fact that it's even hard to come up with tier lists - means they've done a great job with balanced... really impressed.
Try as they might, DOTP is still very much a creature dominanted format.


Just as an aside, considering 60+% of Magic: The Gathering's 10,000+ cards are creatures, this is insanely hard to avoid.




I understand that, but my point was that DOTP has thus far been more creature dominated than the game of MTG in general.  I suspect this is due to DOTP's intorductory nature and the relative ease of understanging combat interaction between creatures as oppose to more difficult concepts of card interaction (this is evidenced by each iteration of DOTP including more enchanments, artifcats and more complex creatures and abilities).

I didn't mean it as a fault to Stainless or WOTC in their design choices (not sure who exactly has ultimate creative control over DOTP), I simply meant that comment as an agreement to and an expansion of your comment about the Obedient Dead and Born of Flame decks not allowing your opponent to play spells and abilities.

The consequence of crafting decks with an abundance of cards that allow a lot of creature control (including both specific and mass removal) is that those decks will tend to be more effective in a creature-centric format such as DOTP.  I readily admit that this also applies, although in my opinion to a lesser extent, to MTG in general.  I haven't followed the actual card game since 2005 so I may be off base, but that is my understanding based on my previous experience with MTG.
Is this all  1on1?
Mine are.  The skew that would result from some random, possibly 90 card deck tossed into the mix on my side, would make any tracking worthless.

All of my games are ranked 360 1v1 
Is this all  1on1?

Yes.
I think it is difficult to create an accurate tier list this way without thousands of games.  Right now one player contributed about half of the data and has an 80 or 90% win ratio with born of flame and OD.  That really skews the data.  I wish someone here with a masters in statistics could help us come up with a way of arriving at better results. 

Players submitting data are going to be experts far better than the rest of the community.  Thus the decks that data-submitters play will look a lot better.  Also not only do builds vary, but the quality of decks varies tremendously depending on whether people have the promo unlocks.

I don't keep track of wins and losses directly, but I would rate the decks as follows:

Goblin tier: Krenko
Tier 2: Liliana
Tier 3: Yeva, Ajani, Odric
Tier 3.5: Chandra, Garruk, Talrand
Tier 4 : Nefarox, Jace
..."window.parent.tinyMCE.get('post_content').onLoad.dispatch();" contenteditable="true" />I don't think Chandra deserves all the hype.  Its an effective deck but very lacking in flexibility and late-game card advantage compared to OD.  I would say flame is better vs Mill and Talrand than OD, but I think OD is better at every other matchup.  Chandra also suffers in this game due to the fact that so much burn is specialized, aside from searing spear, most of her cheap spells are EITHER burn to the face OR burn target creature. 

The balance in this game is better by far than in previous versions, but goblins vs. jace is still ridiculous.  The best deck in the game gets two cards that just completely hose Jace:  Goblin piledriver
I submitted about 800 games worth of results. That bigger sample size should help.
For me the tiers are;

1- Krenko, Liliana, Chandra
2- Talrand, Odric, Garruk
3- Yeva, Ajani, Nefarox, Jace 

The balance in this game is better by far than in previous versions, but goblins vs. jace is still ridiculous.  The best deck in the game gets two cards that just completely hose Jace:  Goblin piledriver



I don't enjoy playing goblins at all, but if others playing it causes people to play less Jace I am all for it.   S Jace. S him in his Q.  It is a glorified burn deck.  
It's virtually impossible to have a perfect balance in any game. I think that Duels 13 is more balanced than previous versions, all decks at least are competitive. My worry is expansions, in the past they have overpowered the new decks to the detriment of the starting decks.
You are Red/Blue!
You are Red/Blue!
Take The Magic Dual Colour Test - Beta today!
Created with Rum and Monkey's Personality Test Generator.
You are both rational and emotional. You value creation and discovery, and feel strongly about what I create. At best, you're innovative and intuitive. At worst, you're scattered and unpredictable.
All I see from this data is that the people submitting it are far better players than the randoms they are playing. I think any games against an untrimmed deck should be removed from the data set, or at least make a separate tracker for trimmed decks only.
While it's true that many of the games will be against unskilled randoms with bad deck builds, if your sample size is large enough then it should all wash out as each deck should have an equal probability of matching up against a bad deck/player.

Problems will only arise if a particular deck has a specific advantage against untrimmed decks/unskilled players that other decks don't have.   And even then, whether you consider that a problem is a matter of interpretation.  The statistics are still telling you which decks fare best in this particular metagame which just happens to include some poor players and decks.

 
If those games took place in the unranked section of XBL, I'd maybe agree.  However, each and every one of those games are ranked.  If someone steps into the ranked area, one has to assume they know what they are doing.

That being said, the vast majority of those games were vs 70 or fewer card decks.   
I guess none of the Steam players are reading this forum, because I see Liliana far more than I see Krenko

In fact, I don't think a single person played Goblins in the Steam tourney, now that I think about it.. 
I said that because on all of the decks listed, US has more wins than losses and THEM has more losses than wins. Also the 3 decks with the highest percentage were played the most by US allowing the percentages to be more consistant showing that the players submitting just outskilled the opponents. It can still be interesting to see, just not a very accurate reflection of any kind of deck strength.
Sadly, with Duels being such an entry-level experience, and attracting new/curious types, and not to simply dismiss all the hard work being done compiling data...but unless some kind of playtest group of similarly skilled players is formed, overall, the data isn't gonna help much.
Yes and no.  Just tossing all the data together in a puddle and calling it good doesn't say much, but with some processing you can tell quite a lot from a mixed data set like these (particularly if it was broken up deck vs. deck rather than pooled by deck).

I don't have data to submit, but if someone wanted to send data (or if it continues to be posted) I'd be happy to take a crack at removing some of the bias from the skill level disparity between Us and Them and the deck distributions.

I've gotten a few submission with additioanl data.  Haven't had a chance to incorporate them yet, but will post resuslts as soon as I do.  

Thanks to all those that submitted. And it really isn't that much hard work, just a little bit of time punching some # into the comp. I track all my games anyway, just because I'm interested.   
Sadly, with Duels being such an entry-level experience, and attracting new/curious types, and not to simply dismiss all the hard work being done compiling data...but unless some kind of playtest group of similarly skilled players is formed, overall, the data isn't gonna help much.



This.

Also consider that different players have different playstyles.  Want to figure out what decks are best for you? Refine your decks and play them out.
Some of those decks you've only played 10 or less times.  None of them have you played 100+.  There's not enough information here to extrapolate anything substantial.

You're a lose cannon.

 

 

"I played 70 card decks before it was cool to play 70 card decks." -Random M:tG hipster

My results over the last 4 days would be completely different.  I'm 38-6, my opponents 6-38.  Goblins would only have 1 victory in 6 games, whereas Pack Instinct would have 1 in 4 games. 

Skill and build matter way more than the deck balance itself in these type of scenerios.





I'd say that's false. Goblins is virtually brainless to play. It's better than 2009 vamps/2012 knights, but not by a whole heckuva lot. The only decks that can usually really beat up on goblins are born of flame and obedient dead because they can usually hang on long enough til they get a board wiper and then punish goblins.

It's no wonder that people's power listings show BoF and OD rising. It's not that they're particularly dominant decks themselves, it's just that they're as close to a rock as goblin's scissors gets. And given how often ppl play goblins you're going to see them rise. Both are good decks in their own right but they generally don't simply roll nearly 1/2 the decks the way goblins do.

Skill/deck balance help, but there's not much several decks can do vs goblins. Particularly the blue decks. I'd say I'm skilled and I routinely get beat by goblin decks with 80-100 cards in them when I play either blue deck, and I can't imagine either green or exalted doing much better unless that get just the right cards or goblins gets screwed some how. Odric probably does okay, but that's a toss up, and Ajani probably needs some good early card combos to weather the storm. And this is all assuming the goblin guy is poorly skilled.

Skill and deck editing certainly help, and in more evenly matched decks help a lot. But usually the only way I get over goblins is with luck and luck alone. And some how I seem to lose on that front too as despite 100 cards in library vs my 61 I can't tell you how many times a goblin has draw  two turns in a row when I'm at 10> health but in control of board, or a second Goblin Piledriver just in time to haste through my Fog Bank barricade or some other BS, only possible out left.  
Pack Instinct has Baloths and Companions, both early game ways of slowing/stopping an avg goblin draw.
Odric has tokens and Doomed Travelers, as well as early game first strike AND buff enchantments.
Exalted has about the most removal (Doomblade, Unmake, Vindicate, Mortify) and can pick off the major pieces, but it is rather slow.  If allowed to begin to churn, it can handle Gobbos 

Ancient Wilds and Blue have issues, as they aren't fast enough, but it's by no means impossible.   If Ancient wilds gets a good draw, Naturla order into some of the nastyness early can slow down the rush.

Also, I'm up to 80 games now, and gobbos aren't even 25% of the games, not even close.  While they may be the easiest and most prevalant, it's NOTHING like RoI or Knights in 2012.  I played one of those 2 in about 60% of the games. 
Pack Instinct has Baloths and Companions, both early game ways of slowing/stopping an avg goblin draw.
Odric has tokens and Doomed Travelers, as well as early game first strike AND buff enchantments.
Exalted has about the most removal (Doomblade, Unmake, Vindicate, Mortify) and can pick off the major pieces, but it is rather slow.  If allowed to begin to churn, it can handle Gobbos 

Ancient Wilds and Blue have issues, as they aren't fast enough, but it's by no means impossible.   If Ancient wilds gets a good draw, Naturla order into some of the nastyness early can slow down the rush.

Also, I'm up to 80 games now, and gobbos aren't even 25% of the games, not even close.  While they may be the easiest and most prevalant, it's NOTHING like RoI or Knights in 2012.  I played one of those 2 in about 60% of the games. 



I don't know many ppl running all 4 companions and/or 2(?) leatherbacks. Even still, companion usually trades/gets shocked. You lose 1 creature they lose 1 creature. You'll replace it with one bigger, they'll replace with 2 more OR one that pumps the rest of his creatures. A baloth slows them, but they're still pushing damage through. You have to assume they've got a grenade because, Hell, I don't think I've played goblins where they haven't hit me with a grenade at least once even though they are 2 of. I still think the clock is too fast for PI. PI needs the right cards, just like every other deck, to line up to fill curve perfectly. Goblins just need to draw cards. If not grenade then a warstrike.

Maybe your argument will be "It takes skill to know all goblins have to do is push through damage, even if they're losing creatures in process" - but I think even the most casual of player knows that.  

I can see Odric having a slightly better chance but it's still in favor of goblins I think. Goblins get out creatures faster usually and then they a) have haste often between bushwackers, chieftans, and what is the last dude? captain? whatever, b) often get pumped in the process between aforementioned cards and things like warcaller, and c) have a nifty little suite of low cost direct damage finishers.

As for exalted, with the exception of doomblade all of his removal is one target for 3 mana. Goblins don't need specific goblins to work. They all synergize. Some better than others but even then they're mostly 3 drops anyway. At most you're trading 3 for 3 and they still scale faster. You're already on an accelerated clock because chances are they get you to 5> life and here comes the KO. And as improbable as it should be, I seem to get hit by double grenade fairly freaking often for 10 by like turn 4.  I can see exalted being good against it, esp. control focused and you get a doomblade and a 3 many kill out to delay then drop turn 4 No Mercy. Or if you survive to get Silent Arbiter but I mean that's a lot of "ifs".

And that's why goblin is so darn oppresive. It doesn't really need any "ifs" other than "If I get 3 land..." or "If I don't draw land for 4+ turns in a row". It is consistant vs many decks even with garbage filler like Goblin Balloon Brigade in it bringing library counts over 85. I'm sure in a closer call matchup like Odric garbage filled decks prob lose to better constructed rivals, but the odds are heavily stacked against me with at least both blues as soon as they pick Krenko's deck from the list.

If I ran goblins, which to now I've refused to do, I bet I'd win. A lot. It's not even because I'm that good or anything, it's mostly just because of the deck and its consistency. My skill/editing will just give me the edge vs mirror match and closer calls but I'd say I'd have a competitive edge vs most except maybe Chandra and Liliana and that's assuming Liliana gets Massacre Wurm (and can hold off til it drops), Infest or Mutilate.
Hell yeah I run all 4 Garruk's Companions and all 3 Leatherback Baloths! I'm also the worst player in the history of Duels.

Don't forget Garruk also has Dungrove Elder at the 3.

I'm not saying it's an easy matchup, and Goblins is still favored, but it's not quite the blowout some folks are making it out to be.

(and I would say Beknighted was more of a braindead deck to run than Goblins...at least against a skilled player, a skilled Goblins player is gonna have to do some good combat math if he doesn't get a godhand...Beknighted...you just crapped out a bunch of "protection from" dudes against most decks and that was it...you were on straight-up autopilot with Beknighted most of the time if you weren't playing against Illusions or Dark Heavens, and maybe Chandra)
2nd worst player, apparently killa is worse.

I agree about the Baloths and Companion, I run at least 3 Companions, might be 4.  I play my PI deck faster and more aggressive, with much less mana ramping.