Results for a manabase without fetchlands are either 2/3 in or 1/2 in depending on your perspective. I'm about halfway done with the second of the three decks, but the info for the second deck is pretty clear, making the rest of it a formality for that deck.
Unlike the previous tests that showed the deck speeding up when going from 19 to 22 lands with fetches, the fastest results I had without fetches were at 16 lands every time (my starting point) and the deck slowed down from there with each additional land added.
This is pretty interesting because it contradicts all previous work on the subject and my only explanation is because the curve is so low. Interestingly at 16 lands the Magma Jet deck was almost equal to the max Bolt deck (well within variance) but as the number of lands increased so did the gap. I was expecting the results to be the exact opposite with Jet improving as land counts went up because it would be able to scry away lands.
There's still the creature deck to run to see if anything changes, before I do any indepth analysis here such as looking at mulligan rates but the initial feedback is very puzzling. My initial reaction is that it's due to a small sample size, but it's very consistent across both decks as the land count increases. Eventually when it's all done I'll present things as a graph, I think that's probably the most friendly way to do things.
Does anyone have suggestions for fetchlands? Unless I limit things to one deck, it's too many results to test every combination of fetch/land at every land count that seems interesting.
--------------------------------------------
Edit: The creature deck without fetchlands was optimal at 17 lands.
Given that previous data sets have shown that a 13/9 manabase is superior to an 11/8 manabase either deck thinning is having a much bigger impact here than previously believed, or I'm being too conservative in not using fetchlands to keep known good cards on top of the library. In order to eliminate the latter possibility the next test will fetch as soon as possible in all circumstances. It makes Magma Jet slightly worse, but I don't think that's a big deal since Magma Jet isn't looking like a hidden gem.
Also worth pointing out, the winner of this round by a full half turn was the creature deck at 4.281 turns. The decklist for that configuration would basically be:
17 taiga
4 eidolon
4 hellspark
4 guide
4 marauder
4 vexing devil
4 price of progress
4 fireblast
4 atarka's command
11 bolt
Considering the fact that 17 Taiga's aren't possible, I'm curious how it will look with fetchlands.
This deck had a very high mulligan rate, out of 1000 games the keep information was
7 - 626
6 - 249
5 - 71
4 - 37
3 - 11
2 - 3
1 - 3
After fetchlands, would there be any interest in me looking at the impact of 0 mana cantrips like Bauble/Probe on the deck in terms of what it does to deck speed? It would involve me adding a new card type which I was trying to avoid, but a 0 mana cantrip is a pretty simple one to add.
Last edited by Brael; 07-04-2016 at 11:56 PM.
Made the fastest/easiest optimization last night in between running these. It went from running 1000 games in 18 minutes to 1000 games in 58 seconds. That speed increase let me drastically increase the number of games being run in my overnight trials. Last night involved running 20 decks at 20,000 games each for 400,000 games total in order to determine manabases. What I was trying to see is how fetchlands affected the optimal land count, and what the right number of fetches were. It was all the same version of a single deck (the Jet one), just with different land counts.
I tested 18, 19, 20, and 21 lands at fetchland counts of 16, 13, 11, 9, and 6 with other lands filling in the rest for a land count. This means everything from a 16 fetch/3 land deck to a 6 fetch/16 land deck was tested with 20,000 games in each configuration.
Yet again the 19 land deck proved the fastest, but because 19 was the minimum tested the optimal amount could be even less. I think the curve is too low and that's why the optimal land count is appearing so low. The data however consistently showed that in all cases the medium fetchland value which was 11 fetch/8 to 11 land gave the best performance by 0.02 turns. I'm currently running some followup tests at 10, 11, and 12 fetchlands in order to determine which is best and should have those results in 1.5 hours or so.
Next up after that will probably be some cantrips tonight, that or adding Exquisite Firecraft. Given the optimal land counts looking to be so low, I want to see how Firecraft changes the numbers.
----------------------------------------
Edit: More detailed fetch manabase results are in (for this specific list)
10 fetch 9 land had an average turn of 4.842 turns
11 fetch/8 land was 4.826 turns
12 fetch/7 land was 4.839 turns
Each of those was the average over 20,000 games.
It's marginal, but the sample size is high enough combined with the other data that 11 fetch/8 lands looks optimal at 19. Though there's still issues that 19 doesn't look optimal in a list without 3 drops.
For now, I'm going to move forward with the idea that the optimal land count is around 60% fetch, 40% land and make some lists that include cantrips to see if they're any better.
Also, one additional bug was found. Fireblast was never firing. Kind of a big deal, as a result I'm redoing several land counts. It doesn't impact previous tests, but it does impact being able to fire these things off.
-----------------------------------------------------
Ran a larger test on mana tonight, found a flaw in my previous one (that's what I get for writing code without testing it). I threw the data in a chart so that those who don't like numbers can easily trace how the average win turn drops as you add lands. For reference with this build (used the Magma Jet build, it didn't really matter which I picked) the optimal land count was 19.
http://imgur.com/MLvMQUF
What I find most interesting is how quickly the average turn drops as you add land until you get to the optimal point, but how slowly it rises if you go past that point. So the lesson to take away from this is that if you're not sure, go a little high on lands, you won't regret it. Also, keep in mind that the higher you go on lands, the better the cards are that you're cutting. The 28, 29, and 30 land decks had cut the bolt count to the bone going all the way down to 8. I suspect that if I had cut everything else first the curve up would be even more gradual, but I wanted to leave Guides in and I felt taking out Jet for some of them would skew things so there wasn't much else to remove.
Ran a set with Gitaxian Probe too. The end result was that at all land counts (and thus burn counts) the deck was slower with Probe or other 0 mana cantrips than without it, but I didn't record the numbers because mulligans weren't being taken properly.
The next test for probably tomorrow is going to include Exquisite Firecraft because it's a very common sideboard option, and given how little the deck appears to punish having extra lands, I want to figure out what the optimal number is with them.
-------------------------------------
Edit vs spamming posts I guess.
Here's the chart with Exquisite Firecraft from 2 to 30 lands, same decklist as the previous chart except Magma Jets were turned into Firecraft
http://imgur.com/jkmJGX1
You'll notice that the fastest turn happens at 21 lands, but the curve back up from that minimum point happens much slower. The gap between 18 and 21 land is bigger than the gap between 21 and 28 land.
So again, this leads towards the idea that if you're not sure, go a little higher on lands.
Also worth noting is that the lowest time with Magma Jet was 4.94 turns while with Firecraft it was 4.89 turns, so Firecraft beats out Jet.
Last edited by Brael; 07-06-2016 at 04:06 PM.
Just completed Legacy Burn for a tournament this Sunday! I have one question though. Should I play any copies of Sensei's Divining Top? If so, how many would you recommend? Thanks!
P.S. If you want to give me any tips, please do so!
I run a very different deck than those here and I would not run top.Originally Posted by RPS
For me any spell that costs 1(which is 24, 8 of which are creatures), I'd rather it be damage potential.
Top smooths out draws at the cost of slowing down a deck that wants to be fast.
For me every 1 open mana I better be doing damage, I am looking into some land changes, 21 is way to many and even with barbarian ring it's counterproductive with glm.
My list for reference
http://tappedout.net/mtg-decks/legacy-burn-12-09-12-1/
The theory is that SDT speeds you up because it finds damage spells, it also operates as a mana sink which lets you go higher on lands more safely. I would say the jury is still out on it (my simulator doesn't play it perfectly, but my spreadsheet really likes it). In practice I've found myself really liking SDT but the data I've gathered so far shows that it and Magma Jet slow things down. Of course, library manipulation to get your sideboard cards is a bit of an intangible that can't really be measured.
The card seems defensible, but I'm starting to come around that it may not be optimal.
-----------------------------------
Ran another simulation on a Probe vs non Probe deck.
The base list was
11 fetch
9 land
4 eidolon
4 swiftspear
4 goblin guide
4 price
4 fireblast
4 atarka's command
8 bolts
4 exquisite firecraft
Then it ran either 4 additional bolts or 4 probes
Each ran for 5,000 games and the probe deck came out ahead by about 0.03 turns. I then repeated it and it came out at about the same.
Next I decided to use the data to compare Swiftspear again. With the Probe deck Swiftspear dealt 23,192 damage over 11,262 turns it was on the board making it 2.059 power. With the Bolt deck Swiftspear dealt 21,593 damage over 11,108 turns making it 1.944 damage.
So neither Probe or Swiftspear appear to be good enough on their own, but the combination is slightly superior to neither. So I guess the question is how much damage you want to deal to yourself. It could be an option for mono red decks that aren't taking the fetch damage. There's also the possibility of a different 0 mana cantrip being used instead like Mishra's Bauble (effectively the same thing in the simulation).
That's probably the last simulation I'll run short of an interesting suggestion until I make a few changes to the program. Originally I didn't build in turn tracking because I figured I could just take the rowID in a database query, but I've found a few places where it's useful so I'll be adding in some turn tracking. For the non technical explanation, this will let me link data such as hands on turn 1 to games that ended on turn 3 or 4 and that's something I think could be pretty useful to evaluate opening hands.
Are there any sideboard/MB cards people would like to see added that I'm not currently using? Given that I've been doing more focused testing I'm not as concerned with limiting the number of card types that a more broad approach testing all combinations requires because of math. Because of my approach not every card can be implemented (this program isn't DotP or MTGO) but if the card sounds useful enough or easy enough I'm willing to include it.
Right now Sulfuric Vortex sounds good. I would like to include Grim Lavamancer but because I'm not tracking the GY it's a bit bigger of an undertaking, though not completely impossible to implement, I would just have to add a GY. If I add a GY I could try out Barbarian Ring as well. If I do add Barbarian Ring I'll also add a pain index tracking how much damage we deal to ourselves per game, though the Price of Progress/Flame Rift category won't accurately add damage.
Would there be any more cards people are curious about?
Last edited by Brael; 07-06-2016 at 11:41 PM.
Hi,
This new creature has been recently spoiled :
Thoughts ?
Getting 6 instants/sorcery into the yard is actually pretty difficult. Getting them into the graveyard before your opponent is dead is impossible.
What's your realistic outlook with this card? If you cast all 1 drops and hit all your lands the earliest you can cast this is turn 4 (and you'll have done 18). It doesn't have haste. At this point, when it swings assuming it lives (you'll have had no other creatures) it's 3 damage, so that's an incinerate.
While it's powerful, the outlook isn't so hot. It can represent a lot of damage, but it's also quite slow. You want your creatures to get in under blockers, not try to go over them.
Honestly, I question whether this card would even be good enough if it had haste, I think Incinerate would be stronger.
Knocked a few points off of my to do list. Gitaxian Probe, Sulfuric Vortex, Grim Lavamancer, and Barbarian Ring were all added and a bug with Fireblast was fixed.
My approach has some flaws though since I'm starting to get some slowdown in the program. Run time has doubled since I added the extra logic for those cards so I probably can't add anymore until I find some ways to optimize things (which I've been tinkering with).
The decks are also winning a little slowly, not quite sure what's up with that (it's probably just my card choices) but it's fine to just say one deck is faster than the other, so it's not really all that important.
I recently ran a few thousand games with some varying decklists. They were all some variation on this. Everything from 1-4 GLM, then some 2 GLM's with different land counts filling in more Burn as stuff came out.
11 Fetch
9 Fetchable land
2 Barbarian Ring
4 Eidolon of the Great Revel
4 Goblin Guide
3 Vexing Devil
2 Grim Lavamancer
4 Price of Progress
4 Fireblast
9 Lightning Bolt (or Lava Spike if you would prefer)
4 Gitaxian Probe
4 Exquisite Firecraft
Not the best list, I admit but I was looking at some specific things that those cards would show rather than going for raw speed.
As I said, it's slowing down so over the 7 decks I played 1000 with each deck for 7000 games total.
With the above deck the fastest wins were on turn 4, 13/1000 games had a turn 4 win.
Oddly enough, there wasn't a single 7 card hand that won that early, they were all mulligans. The average hand size for a turn 4 win was 5.92 and the average hand was
1.42 Fetch
0.50 Land
0.00 Barbarian Ring
0.50 Eidolon of the Great Revel
0.42 Vexing Devil
0.75 Price of Progress
0.08 Fireblast
1.42 Lightning Bolt
0.58 Exquisite Firecraft
0.25 Grim Lavamancer
So roughly 1.9 lands, .8 creatures, 2.9 burn spells, and anything else. So without the decimals that would imply the opening ideal hand on 6 cards is 2 land, 1 creature, 3 burn, and 1 of anything.
Opening the dataset up to all 7000 games there were 80 games which had a turn 4 win, and again there were no 7 card keeps and almost no 5 card keeps. It's interesting enough that I'll have to look into the why of this further as there were plenty of turn 5 wins that kept 7 cards. The hands were all on the play so I don't think it has anything to do with going 8 cards deep with a mulligan scry on the draw. Because of the differing decklists I don't think average cards are all that useful but there are a few trends that I see as peculiar. There were no Goblin Guides in any of the fastest hands and no Swiftspears. It was almost always looking for 2+ bolts, an additional 3 damage burn spell, and a 4 damage. Exquisite Firecraft featured rather prominently in the fastest hands and had a significantly higher presence than Fireblast. Lavamancer showed up from time to time, and when he was around Barbarian Ring never fired but I find it very odd that Lavamancer, who should in this situation be strictly worse than Goblin Guide would show up when the Guide made 0 appearances in thousands of games.
Some data on Lavamancer: He appeared in 11,295 turns and activated 2571 times for a grand total of 13,866 damage or 1.23 per turn.
While the ceiling for Bedlam Reveler is incredibly high, I don't think that Burn is the most suitable deck to utilize it.Hi,
This new creature has been recently spoiled :
Thoughts ?
The best-case scenario for Burn is being able to cast this for 2-3 mana, and drawing 3 Burn spells, and then surviving long enough to either dome them that turn, or the turn after that. This requires 5-6 spells in the graveyard (theoretically 15+ damage dealt at this stage), and 2-3 lands in play (less demanding, but sometimes 3 mana just doesn't happen). The worst-case scenario is that it gets stuck in your hand, either by not hitting enough mana+spells to get there, or something like Deathrite Shaman/Rest in Peace thwarting our plans.
The real question is, what sort of Burn deck do you think is most suitable for Bedlam Reveler to fit in, while also not supplanting better cards in the process?
@Brael,
I'm not quite sure how your simulation is currently programmed--ie is it a set of rules to govern a burn deck to count to 20 damage or is there another deck/set of rules that will interact with it. However, in either case, I think it may be appropriate to look at the standard deviation in turns (some histogram information might also be good with regards to how deck performance is distributed) along with the average kill turn. I feel that the current metric you use completely forgoes the concept of stability for speed, leading you to want to test more aggressive probe/bauble builds. However, as deck builders/players we need to consider both, especially if the deck doesn't try to win on t0-1 (where you can almost just reduce opponent interaction to a percentage of them having disruption).
There currently isn't another deck to interact which skews things towards creatures. It was written in such a way that it can be added, and is in fact coming up on my to do list (I have to add player life tracking first). I've been thinking about the implementation for a week now, mostly I've been stuck on the combat zone because attacks/blocks are extremely complex, but I think I have a way to do it quickly but at less than optimal efficiency if I treat the combat zone as an implementation of the stack. Though that means implementing a stack in the first place (I don't think that's too hard though, it's literally just a list that acts as a holding zone).
I could include some graph information similar to what I did for the land counts and mana curve a couple posts back but because most of my posts have involved decks that aren't all that different I've been holding off as the curves would look extremely similar.
All of my output gets written to a database (though I've been discarding old data as I make changes to the logic), and from there I can pull anything up with a combination of SQL and occasionally writing the tables to a spreadsheet so providing graphs of win turns at different turns is pretty easy to do in order to create something of a curve. I've been using the average though because a single number is a little easier to express than pictures and I've been trying to not drown people in data, and just provide the interesting stuff.
Do note, that I'm only posting the averages, but I'm already looking at all of this stuff so it's somewhat filtered before I post. Despite my past few tests all being done with 0 mana cantrips, the data has been quite clear that it's a bad choice. I've just been playing with the Probe data because I'm trying to answer several questions at once such as if it's possible to use these cards as a budget replacement for fetchlands when it comes to things like filling the graveyard for Grim Lavamancer and Barbarian Ring, or if it could ever be relevant for Spell Mastery in a build that leans closer to sligh than burn.
I see. Combat would definitely increase the program's complexity, but I don't think it can simply be an extension of a stack. It can probably be boiled down to a pretty straightforward algorithm by using maximum damage output of all iterations of attackers/blockers as the choice moving forward.
However, a simple solution may be to add a decaying factor to the creatures' damage output that will drop creature damage to 0 based on some randomly generated noise, turn #, and number of creatures. This may also account for why swiftspear is giving you less average dmg in your simulation--goblin guide's dmg output is a static 2/turn, whereas swiftspear's is far more frontloaded and falls off.
Similarly, how is the PoP and eidolon damage output calculated without an opponent to play them off of?
I am curious what a typical histogram (kill turn) of a simulation would look like though, if you have one available.
Regardless, good job in creating this simulation. It seems like a sizable project to have undertaken.
Lots of info this time, only tested 2 decks but there's lots to say because of the turn tracking implementation. I used a burn deck and a sligh deck. I can give decklists on request, the lack of any sort of spoiler tag on these boards makes me not want to post them since I can't hide them. It would be ok in this case since it's just 2 decklists, but gets to be really problematic with 30.
What I really want to talk about here is the average opening hand, starting with the Burn deck. I ran 10,000 games with each deck. The simulator had no games that ended on turn 3 but several that ended on turn 4. Out of the turn 4 games the average number of each card showing up in the opening hand was
2.35 Fetch
0.57 Land
0.10 Barbarian Ring
0.24 Vexing Devil
0.91 Flame Rift
0.20 Fireblast
2.06 Lightning Bolt
0.36 Exquisite Firecraft
So the best performing hand was 3 Land, 2 Bolts, and a 4 mana burn, and then a 1 of 3 or 4 mana burn. Or to put that another way, 3 lands 13 points of Burn (more specifically: 13.42).
With the Sligh deck there were fewer games. The hand information for them was
2.36 Fetch
0.57 Land
1.14 Flame Rift
0.07 Fireblast
0.29 Atarka's Command
1.64 Lightning Bolt
0.07 Gitaxian Probe
0.78 Exquisite Firecraft
0.07 Grim Lavamancer
So what you're looking for here is again 3 Land, and some 3 and 4 mana bolt spells. Once again coming to 13 points of burn (13.75 average in this case) I find it extremely telling that the best performing hands in the creature deck involved no creatures at all and that even though my card choices were clearly pushing Eidolons, there were no Eidolons involved in the fastest decks.
Taking the suggestion of posting a standard deviation on turns and other statistical data, I'm going to one up that (partly because it's easier for me) and I'll provide a graph of the turns.
Do note, I cap the games at turn 12 so there will always be a spike on that turn.
First, here's the burn deck, the average was 6.02
http://imgur.com/lGbRbWf
Second, here's the sligh deck, the average was 7.57
http://imgur.com/vd97MKx
Based on this information I'm going to try putting a list together that's just the best performing cards.
I'm going to put off adding in self damage for now as well as interaction. Instead I'm looking at a new way to make my card decisions. I have a general system mapped out in my head but implementing it will take some time as I've never had a class on AI or ever really attempted it before. It doesn't help that I don't know the name of the type I'm trying to add so Googling examples has been less than helpful but I do think it will work.
Until it's done, any further analysis will just be with what I have right now. There's a lot things to check out with turn tracking and opening turns. Most notably I would like to make a comparison on the amount of burn in your opening hand with various decks, and the chances of winning on each turn with that much. I think that 13-14 points of burn information is a solid piece of data. At a minimum, it's not something I had ever read about in a metric for evaluating an opening hand. There's a lot of information to be gleaned from how things are now, most notably everything I've done has been on the play but with a couple keystrokes I can change it to be on the draw and build that info.
Something I'll probably do tonight after our Saturday night tournament, is take the above hand information and look at the frequency of each card compared to how many are in the deck in order to try and determine what is under/over performing. From the looks of it right now Bolt/Firecraft overperform and everything else underperforms.
One last thought, I find it very interesting that the fetch/land distribution was near identical in the fastest hands though it could just be coincidence.
The reason I say I can use it stack like, I can make each creature attack, put a blocker if it exists on the biggest one, and then assume a trade happens. Alternatively, I can make a couple categories of blockers like big/small, the small ones trade and the big ones never go away. It's not perfect but it's good enough without calculating clocks, counter attacks, and so on. Which is probably beyond my ability, and also increases computational complexity beyond the point where it's viable. It's important to keep speed up in order to generate enough data.
I included something like that in my spreadsheet and it worked well enough. I'm actually running into the situation with the simulator though that creatures are under performing compared to burn spells even without any removal for them. It's a bit of an interesting problem and I'm not sure why it's happening yet because it defies all logic. I've read and reread the program a good 10-15 times now, and poured through over 100 games worth of log information looking for some indication that creature damage isn't being applied, but I haven't found anything yet. My best guess though is that I broke something when adding Lavamancer because I had to tweak the combat function.However, a simple solution may be to add a decaying factor to the creatures' damage output that will drop creature damage to 0 based on some randomly generated noise, turn #, and number of creatures. This may also account for why swiftspear is giving you less average dmg in your simulation--goblin guide's dmg output is a static 2/turn, whereas swiftspear's is far more frontloaded and falls off.
Generating a set of data with Swiftspear and comparing how much damage it does each turn is on the list of things to do. It should perform better as the average turn goes down. If you notice, most of my simulations run for 6-7 games and by that point Swiftspear peters off but still, the data hasn't been kind to it which I find very odd because it runs completely counter to both the theory behind it and my experiences.
I'll throw it onto the list of stuff to do.
Just estimates based on experience. PoP deals 4 on the assumption the opponent figures out our deck and only gets one non basic (this is why it's lumped in with Flame Rift right now, they're both effectively 4 damage for 2 mana). Eidolon swings once per turn and triggers once per turn.Similarly, how is the PoP and eidolon damage output calculated without an opponent to play them off of?
Part of this stems from my original approach which I've since abandoned due to run time, which involved giving a range of every single card and testing every permutation that made a valid deck. Then looking at that massive database for results. This approach required grouping similar cards, as time has gone on though I've been moving away from that and adding more cards. Adding new cards still has some cost though because of how I'm choosing which spells to cast when. I need to develop a new, more dynamic AI. I have an idea of how to implement it, but it's going to take me some time.
I actually posted some for you last night, but it looks like my post got eaten in the forum database errors. See above for some simulations I ran today. I think it's what you were asking for, it's another piece of data that becomes easily available now that turn tracking is in (technically it was available before, but hard to get).I am curious what a typical histogram (kill turn) of a simulation would look like though, if you have one available.
----------------------------------------------
Edit: Ok, after some searching I found the bug. When I added Lavamancer I accidentally disabled every creature other than Lavamancer from dealing damage. That explains a lot.
Last edited by Brael; 07-09-2016 at 07:28 PM.
I can see it(him) with gitaxian probe... And who plays Rip against burn? And well DRS is a strong card so you probably bolt him anyway. So if they counter/discard/chalice for 1/DRS -> in other words if the opponent plays magic it can be a really nice topdeck.
He looks bad if you only goldfish; but might be turn out very solid in a real game. But I kinda doubt it, he is really bad when you flood lands.
“Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.
Taking Burn to my LGS's first Legacy event today. I've been playing some matches on MTGO to get the feel of the deck (I come from playing Burn and Aggro decks in Modern) and have done pretty well. My question is what sideboard do I bring to a brand new meta? This is what I'm thinking at the moment.
4 Exquisite Firecraft
2 Ensnaring Bridge
4 Leyline of the Void
3 Smash to Smithereens
2 Pyroblast
Here are my questions.
Is this a decent sideboard for an open meta?
What would you change?
(Or) What sideboard would you bring?
I take it you're playing mono red? I find I prefer Faerie Macabre to Leyline of the Void simply because it doesn't force as many mulligans. 4 seems like too many though, we have a reasonable enough game against Dredge as it is, and Reanimator is almost a never win. Your graveyard hate is usually at it's best against decks like Storm who rely on bouncing/counters and random combo decks (especially the budget decks like No Spells). In those a 1 shot that's virtually impossible to interact with short of a Stifle is good enough.
Instead of Smash I prefer Destructive Revelry, it hits slightly more targets and the green splash is solid because it also enables Atarka's Command in the MB.
I like Lavamancers in the side for the Delvers of the world, never cared much for them MB but I think they strike a good balance between board control and aggression. Another one I like is Volcanic Fallout. It's strong against most aggressive decks, and it pulls double duty against Miracles. My sideboard (keep in mind that despite what my computer toy says, I'm still using top) looks like this:
2 Volcanic Fallout
2 Faerie Macabre
4 Exquisite Firecraft
2 Destructive Revelry
2 Red Elemental Blast
1 Sulfuric Vortex
2 Grim Lavamancer
Creature bug found and fixed in the program, it's amazing what 1 line being bumped slightly out of place can do.
Reran the previous two tests, same lists, same number of turns, just looking at a Burn deck vs a Sligh one
In the case of the burn deck it came out to 14 points of burn in hand being the average in the fastest winning decklists (vs the 13.5 last time). The best average hand was
1.30 Fetch
1.51 Land
0.15 Barbarian Ring
0.90 Eidolon
0.15 Vexing Devil
0.68 Flame Rift
0.32 Fireblast
1.49 Lightning Bolt
0.36 Firecraft
So an optimal speed hand is 3 lands, 1 creature, 1 3 damage burn, 1 4 damage burn, 1 misc burn
The average kill turn was 5.42 turns. There were still no turn 3 kills, though it should be theoretically possible. In chart form the turn curve looked like this:
http://imgur.com/8XcqvvP
With the Sligh deck there were a handful of turn 3 kills, but they were very rare. The average hand for turn 3's was
1.50 Fetch
1.75 Land
0.25 Swiftspear
1.25 Guide
0.25 Fireblast
0.75 Atarka
1.00 Lightning Bolt
0.25 Gitaxian Probe
So again, 3 lands, 2 1 drop creatures, 1 4 damage spell, 1 3 damage spell
The average hand for a 4 turn game was
1.34 Fetch
1.69 Land
0.73 Eidolon
0.12 Swiftspear
0.40 Guide
0.53 Flame Rift
0.28 Fireblast
0.28 Atarka's Command
0.85 Lightning Bolt
0.06 Probe
0.42 Firecraft
0.11 Lavamancer
So 3 land, 1.5 creatures, 1 4 damage burn, 1 3 damage burn, 1 misc
The average kill turn for this deck was 4.85 turns. The curve looks like
http://imgur.com/VIjfN2f
The data here between both decks is pretty clear, the ideal opening hand looks like
3 Land
1 4 damage spell
1 3 damage spell
2 creatures (or 4/3 damage spells)
I'm doing a test tonight, 100,000 games each on the play vs draw with the same deck that's basically just a pile of cards that have performed well so far. I'll be busy tomorrow though, so it will probably be rather late when I get that data posted.
Who runs Gitaxian Probe in Burn? The card is incredibly sub-optimal in it. Also, I've seen quite a few Miracles players bring it in for Game 2, to counteract both Grim Lavamancer and Exquisite Firecraft, both of which are typically hard for Miracles to deal with.
Bedlam Reveler just looks bad in general in Burn. Even with your ideal situation, which is tapping out on Turn 3 (getting 5 Burn spells in the GY by then requires an act of God, even with them all costing 1 mana) and drawing into 3 more bolts. It doesn't speed up the fundamental turn, it doesn't interact with your opponent (who will have larger creatures out by then), and forces Burn to play a more goldfish-heavy game of Magic. To get this out reasonably reliably by then in Burn, you'll need to run cards that are bad in Burn. Your only 'realisitic' cantrip is Gitaxian Probe (as the other 0-cost ones are artifacts), and cards like Gut Shot just aren't what you should be playing. In spite of this, there's still no realistic room for it. For decks running 19-21 lands, 10-13 creatures, and 3-4 Sulfuric Vortex+SDT, that leaves you 22-25 spells to play around with, which is nowhere near enough to consistently cast it on Turns 3-4.
Bedlam Reveler looks even worse if you actually plan on interacting with your opponent. It is antagonistic to how the interactive builds play out, as you're forced to run more Burn spells in lieu of creatures. This also means you can't run Grim Lavamancer if you want to reliably cast Bedlam Reveler, as those two cards cannibalize each other; sacrificing a decent percentage of Miracles/DRS/Delver matches to run an at-best win-more card isn't the strategy Burn wants to be facing. IFF you give zero fucks about interaction with your opponent, and just want to windmill slam every single Burn spell, then Bedlam Reveler gets better.
What cards are you removing from Burn to put Bedlam Reveler in?
More data, and an update.
I have an idea to move away from the current spell casting system which is basically just trying to cast spells in a given order if possible to one that can better make situational decisions such as being more likely to cast a creature if it will speed up the clock, but a spell if not when there's a choice between the two or being able to recognize that it's better to play 2 2 drops at 4 mana than a single 4 or a 3 and a 1, and prioritize that line. AI programming isn't really my thing though, so it may or may not work out, what it has going for it though is I think I figured out how to implement it.
Now for an update, a poster PM'ed me their decklist and asked me to run an analysis on it. It was essentially 21 land, 4 guide/lavamancer, and the rest burn. So I took this and tested 11 decklists using everything from the base list -5 lands to the base list +5 lands. When removing land I first added a Firecraft, then I added 4 Swiftspears. When adding land I removed Lavamancers followed by a Lightning Bolt.
The data indicated that in this case, more lands were not necessarily better
http://imgur.com/66dNeNO
That's the graph at various land counts. The lower the land count got, the faster the deck got. I suspect part of this is due to Swiftspear being more powerful than a land
The base list had the following turn graph
http://imgur.com/Qgj5Neo
The fastest list looked like this
http://imgur.com/nacN4Wp
Across all 11 decks the average opening hand that won on turn 4 was
1.38 Fetch
1.47 Land
0.02 Swiftspear
0.34 Goblin Guide
1.02 Flame Rift/Price
0.34 Fireblast
1.95 Lightning Bolt
0.04 Exquisite Firecraft
0.26 Grim Lavamancer
0.17 Sulfuric Vortex
So rounding stuff off that's 3 Land, 1 Creature, 11 Burn as your ideal opening hand at 7.
The fastest hands on a mulligan to 6 were
0.96 Fetch
1.50 Land
0.01 Eidolon
0.03 Swiftspear
0.36 Guide
0.85 Rift
0.21 Fireblast
1.63 Bolt
0.05 Firecraft
0.25 Lavamancer
0.14 Vortex
So that's 2-3 Land, 1 Recurring threat (creature/vortex), and 10 points of Burn.
Basically the same thing as the 7 land.
Going down to 5 the data starts getting a little thin but the average optimal 5 was
0.58 Fetch
1.31 Land
0.02 Swiftspear
0.42 Guide
0.69 Rift
0.14 Fireblast
1.51 Bolt
0.05 Firecraft
0.22 Lavamancer
0.06 Vortex
Here you want 2 land, 1 creature, and 8 points of burn with no 3's.
So if you want to apply this information to mulligan decisions, a 6 with less than that would be better off as a 5. Interesting to note, 5's that won on the earliest turn possible were about half as common as 6's that won on the earliest turn possible, so that would suggest you want to mulligan pretty aggressively on 6.
I'll probably start trying to incorporate some of this mulligan information into my mulligan decision process both IRL and in this program once I have a bit more data.
I recently got back into Magic after a 13 year hiatus, but have been playing since Legends. I've always loved playing Burn(and Stasis/control decks), and have been playing Modern Naya Burn recently with success. The power of Legacy has been calling my name and I'm building Legacy burn deck, but a lot has changed since I stopped playing. This is an excellent thread with invaluable information, but there seems to be some variance in what the most efficient build is.
Here is my current build:
Lands
4x Scalding Tarn
4x Bloodstained Mire
2x Arid Mesa
9x Mountain
Creatures
4x Eidolon
4x Goblin Guide
4x Swiftspear
Instant
4x Lightning Bolt
4x Price of Progress
3x Fireblast
2x Searing Blaze
1x Thunderous Wrath(I'm aware this is janky as fuuuuu)
Sorcery
4x Chain Lightning
4x Lava Spike
4x Rift Bolt
Enchantment
2x Sulfuric Vortex
1x Molten Vortex
_____________________
Sideboard
3x Pyroblast
4x Exquisite Firecraft
1x Grim Lavamancer
2x Smash to Smithereens
1x Shattering Spree
3x Ensnaring Bridge
1x Pithing Needle
Now let me give you my reasoning behind my particular build/go on the defensive. My meta consists of a good deal of Eldrazi, Merfolk, Delver, etc. So essentially a critter heavy meta, but not without some Miracles, Dredge, Lands. Chalice sees a LOT of play in my meta as well. While I love me some GLM, he always seems to be killed off immediately and I often feel like playing him is slowing me down, which is why he remains in the SB. In his place I've put in 1x Molten Vortex. It's a weird(and probably wrong) substitution, but I've found that I really only need 3 mana to win, so I'd rather pitch the excess lands and deal some damage instead. It also saves me from taking Eidolon damage while speeding up the clock. I only run 3x Fireblast because I draw 2 in my starting hand or draw them back to back way too often, and that just feelsbadman.jpg. Can I justify cutting one, or am I doing it wrong? The Thunderous Wrath is so dumb, and so bad when you're not topdecking it, but I've had a really great experience with it as a one-of in Modern and want to see if it has a place in Legacy. I'll cut it with great haste if it proves to be the garbage everyone thinks it is. This brings me to a couple of questions about card choices.
- Since I've read numerous articles about fetchlands not being all that helpful in mono-colored builds, would it behoove me to run straight mountains(and maybe a Ring or two) to avoid life sacrifice for a negligible benefit? Mind you, I'm not in love with GLM, so I don't really care about filling my graveyard.
- SDT is a card I'm interested in running as a 1 or 2-of, and I would think it would be rad with Thunderous Wrath as well. Is SDT considered an added benefit to Burn(outside of the obvious pros), or should I focus more on spells that actually play towards our objective?
- I noticed that our old staple of Flame Rift has been cut. While I know it doesn't feel great using it with Eidolon, wouldn't it be a good addition to Burn, even as a 2-of?
- Do you feel that I need to even keep GLM at all?
- I'd like to run 1x Volcanic Fallout, but don't see many sweepers in SB's across the internet. Is there an argument against boarding a sweeper?
Since this is my first post and an extremely dense one, I'd like to thank you all for the great discussions and help.
There are currently 1 users browsing this thread. (0 members and 1 guests)
Reply With Quote