If you’ve watched my stream (shameless plug: https://www.twitch.tv/child__roland ) you’ve heard me mention a spreadsheet after almost every draft and draft game. I’ve been tracking my deck lists, influence breakdowns and the influences of my opponents at the end of games. And, of course, I track the results of games and drafts overall. Here is a link to a snapshot of that spreadsheet, as well as basic draft probability calculator that can take a win percentage and tell you how likely you are to go 7-x, 4-3, etc. Feel free to copy these sheets and use them as you see fit. A big shout-out goes to my friend Matt, for helping me write the formulas that made all the data digestible. Another shout-out goes to my brother Phil, for joining me via Google Hangouts for a couple of the 7-x drafts.
The above linked sheet show the results of 34 drafts, every single Omens of the Past draft I have done since last Friday, when the set launched. There are a couple spots where I forgot to track a piece of data, but for the most part, each draft is fully recorded with the above mentioned data. Under the data for the first draft are calculations on the entire sheet. 34 is not really a very large sample size, especially when you start breaking it down into factions played, the data is probably not very statistically significant yet. However, looking at my opponents’ data gives us larger numbers, as I had between 3 (thankfully only once so far) and 9 opponents each draft, for a total of 223 decks I played against. Obviously, I do not have complete data on these decks. The only information I have recorded is what their influence was at the end of the game (for example: 1 Fire, 5 Time, 2 Shadow) and whether I won or lost the game. From this small amount of data though we can do some interesting calculations. For example this section shows my opponents win-rates against me with each faction, weighted by the ratio of that faction’s influence to their total influence at the end of a game:
Opponents’ performance based on factions played:
So for example, if an opponent beat me with an influence breakdown of 5 Fire, 2 Time and 3 Justice, the Fire faction would get .5 credit for the win, with Time and Justice getting .2 and .3 respectively. If I beat an opponent on 3 Primal 2 Shadow, it would count against those factions with respective weightings of .6 and .4.
What the above data at least starts to show is that if you want to beat me, Shadow is your best bet by a decent margin, and stay away from Justice. I can’t say that a deck that beats me is also likely do do well against the player base at large, as I obviously have my own tendencies and weaknesses. For example, maybe I tend to build decks weak against quickdraw or unconditional removal, and that is why shadow does better against me. All I can say is that I’ve drafted a fairly wide range of decks, and while the data shows I tend to favor Time (even though it provides my lowest win-rate, go figure), I have drafted decks of each faction quite a few times now, so hopefully I’m not skewing the results too much.
The other interesting observation of my opponents’ decks I can make is the number of factions they showed me. Because half on faction strangers and the like are somewhat prevalent I don’t count a player as being on a faction unless it accounts for at least 10% of their total influence. Here is the data I have on performance based on number of factions played:
Opponents’ performance based on # of factions:
|# of colors||1||2||3||4||5|
As you can see, the 2 times a player only showed me one faction, I beat them. These were likely influence screws, not mono faction decks. I have played against 4 faction decks 8 times and 5 faction decks 7 times. Obviously these are not statistically significant numbers. The real meat here is in the 2 and 3 faction numbers. If a player brings 2 factions (or 2 plus a very light splash that falls below the 10% threshold) they are favored. This could be because I tend to go off the deep end a fair bit and 2 faction decks tend to lean aggressive and consistent, which is tougher for my brews to handle. It is worth noting, however, that 2 faction decks account for less than half the total of my opponents. So maybe a lot of people are going 3+ factions in this format. And maybe, as my data shows, drafting a solidly consistent 2 faction deck is the way to take us dreamers down. The very poor performance of opposing 3 faction decks makes me think that maybe there are a lot of drafters out there who heard this was a 3 faction format and just went to town without really knowing how to draft such a deck, as they were relatively rare in quad Empty Throne draft. If that is indeed correct, maybe I can help, as the data shows no such trend for the performance of my 3-5 faction brews:
My performance based on # of factions:
|# of colors||1||2||3||4||5|
|sum of win ratios||0||3.302777778||7.001190476||1.696428571||1.677777778|
The data above is harvested a little differently than the data for my opponents. Since I have full information of my decks, I don’t need the 10% threshold here, I know whether I was really playing a faction. Also, as I said before there is a much smaller amount of data points for my 34 decks vs my opponents’ 223. But as you can see, not only do I do better against 3+ faction decks, I do better with them as well.
Tip 1: Pick fixing highly. Banners and strangers are your most common sources, but Seek Power and Amber Acolyte are still commons in 2 of the packs, and they are still great. Amaran Archaeologist sucks, 5 is just too much power to spend on fixing. Maybe he is OK in the sort of deck that wants to consistently hit 7-8 power, but that’s not what we are looking for here. Strangers are the most powerful source of fixing, but including too many 2 power 2/2’s in your deck is a real liability, unless you can make them better than that, which leads to my next tip.
Tip 2: Stranger lords are great. The best 4 (non-rares, I’ll be trying to focus mostly on commons and uncommons, as those are the vast majority of what you will see) are Battle-Tested Stranger, Soaring Stranger, Savage Stranger and Determined Stranger, in best to worst order, although they are all great. Ruthless Stranger gets an honorable mention. The common theme here is that these all add stats, plain and simple, with the exception of Soaring Stranger which makes them all fly. The other keyword strangers are mostly decent, but not what you need to really tie a deck together. If you want to build a true “Strangers” deck you want at least 2, preferably 3 of the above lords. You also want about 6+ 2 drop fixing and/or minor lords, but as I said above, you should be taking the fixing strangers fairly highly already. One thing to keep in mind is that this strategy is far from a secret, so the people passing to you could easily be trying this, so if you aren’t seeing what you need, you should jump back to a safer draft plan before it’s too late to get enough playables. If you’re deep in pack 2 by the time you realize this, it may already be too late, so good luck.
Tip 3: Especially early on, take bombs and removal of all factions. If you can’t end up playing them, that’s too bad, but you only wasted a single pick, not a huge deal in the long run. If you can fit them all in, you have a ton of bombs and removal, which is great. Be careful with double influence cards, but if you’re diligent with taking fixing, something like a total influence requirement of should be doable.
If you follow the above tips and stick to a reasonable curve (mostly cards that cost 3 or less, very few that cost 6+) you should do alright with 3, 4 or even 5 faction decks. The power-level ceiling is so much higher that occasional loss to influence screw really is worth it.
While we’re on the topic of leveling up, I wanted to highlight just how powerful a few points added to your win percentage can be. The draft probabilities portion of the linked spreadsheet is defaulted to showing what you can expect from a 50% win-rate. Cell J8 shows that someone with that win-rate can expect to 7-x (7-0, 7-1 or 7-2) in just under 9% of their drafts. If we up that to 57% (you would need to copy the sheet in order to edit it) we see that number jump to just under 18% of drafts being 7-x’s, just about double! So you don’t need to get that much better for your results to see a very measurable improvement. Just shore up a couple weak points and you will be pulling down diamond chests in no time!
Some other less organized thoughts about the format and some individual cards:
Granite Ring is an amazing card. Yes, it doesn’t do much when you are behind, which is a liability, and not the type of card you want a ton of in your deck. But, it gives all your units (not just one, read it again, it’s units, plural) +1 attack and overwhelm. This is a somewhat slow grindy format, and this ring will break a boardstall right open. Consider holding it to play and activate as a surprise all in one turn, but if you can easily fit into your curve earlier, you probably should, since it’s unlikely your opponent can do much about it anyways. Great with deadly as you will only spend one point of attack to kill each blocker and the rest hits the opponent. Great with cards that make multiple units such as Grenadin Drone, Assembly Line and Humbug Swarm. Even Amber Ring is a decent combo with this, but that takes a long time to set up, so it shouldn’t be plan A.
Cloudsnake Breeder is bad. Just too much of a setup and not enough effect. The real nail in the coffin is the double primal influence. If you are on an aggressive deck with at least 10 primal sources, it becomes playable, but that is the only place I would run it.
Talon of Nostrix is pretty good. It can kill any 2 health unit by itself, which includes the vast majority of flyers. You probably don’t want more than 2 in a deck, but those first 2 will really pull their weight in the early game and even be OK later. Keep in mind you can run a small unit into a big problem unit and finish it of with the Talon, which is sub-optimal, but sometimes the best play. It can sometimes even kill something like an Initiate of the Sands or a Sand Viper and survive to take down another small thing on the next turn. This won’t happen often, but when it does, or when your opponent spends a card to stop it from happening, you will feel great.
Purify is amazing. Fast speed, very tricky removal is always great and this fits that bill. You can strip flying, deadly, quickdraw and any other combat relevant text mid-combat. If you have one in hand, whenever you opponent attacks, just consider what would happen if you used it on each attacker, and you will often find a lot of value.
I left out a lot of great cards, but I think most people can identify that stuff like Dragonbreath, Mortar and Slay are cards they should take early and try to play. I wanted to go over some of the less obvious cases. There are plenty more though, hopefully I’ll find time to go over them soon!
To close out, I’ll give you the best moment of the format so far. I had drafted a pretty awkward 5 faction deck, you can find it under column BU in the linked sheet. I included Cat Burglar, because even though the Summon effect was pretty likely to whiff, it was at least a deadly unit, so it would trade with most attackers. In one game, my opponent and I were deep into a board stall, probably about 8 units on each side. I top-decked the Cat Burglar and was reasonably happy, since it at least made my opponent’s attacks worse. My opponent had a ton of power and influence, so I was pretty sure the summon was going to get nothing. However, when I played it, the window showed a card. That card was The Last Word. I guess my opponent was waiting till the board was just a little more favorable, but whatever the reason, that still felt amazing. I’m pretty sure I went on to lose that game, I finished 2-3 in that draft, but that one moment more than made up for it.