Turn 3 is a pivotal turn. The first two turns you can analyze fairly well by looking at starting decks; turn 3 is where the game explodes with possibilities. It’s the first turn you get to play with cards from your codex – and the same for your opponent. It’s also a very asymmetrical turn: Player 1 gets to assault (or at least threaten) 2’s Tech 0 defenses with advanced spells or even units (unless they ‘whiff’, or draw neither of their techs). If they don’t force Player 2 to spend enough gold, Player 2 can jump to Tech II a turn early and turn the tables (sort of, if they manage to draw their Tech IIs)
The critical question is: what plays do you make turns 1 and 2, and what cards do you tech, to set yourself up for success?
First, how much gold does everyone have to spend?
Player 1
0 0 1 1 2 2 3 (tech level)
-----v-----------
4 5 6 7 8 9 10 (workers at start of turn)
3 3 5 2 7 *3 10 (gold after worker & tech)
Player 2
-----v-----------
5 6 7 8 9 10 10
4 4 *2 7 *3 10 10
The above shows how much gold each player will have to spend on each turn, assuming they worker up to 10 and build every tech building at the earliest opportunity. Asterisks mark optional techs; Player 2 will often delay Tech II until turn 4, giving them 6g turn 3 and 3g turn 4, and either or both players might delay or skip Tech III. Multi-color decks will have 1g less on turn 2, for 2g and 3g, respectively. Scavenger gold isn’t represented here; just consider it a discount on any units you know are going to die in that slot.
(Yes, P2 can technically build Tech I turn 1, but they shouldn’t, so we’ll ignore that.)
A few things jump out:
-
If P1 wants to play a hero and a 2g unit turn 2, they need to save a gold turn 1.
-
Assuming P1 spends all of their gold turns 1 and 2, and they want to be able to play both of their teched cards in case they draw them turn 3, they better not cost more than 5. Martial Mastery + Rambasa Twin works, Huntress + Centaur works, 2x Stewardess of the Undone does not.
-
A split draw allows up to 7g, but this may not be advisable.
-
Teching two very cheap cards (2x Sparring Partner, Overeager Cadet + Boot Camp) is essentially committing to spending a big chunk of gold on your hero turn 3 for Player 1.
-
P2 needs one heck of a gold efficiency advantage if they’re going to be able to keep a Tech II building alive with only 2g + savings after P1’s 5g turn. If P1, on the offensive, trades two equal-cost units with P2 in Scavenger, P2 will have 4g vs P1’s 5g on turn 3 – but their card cycling to play any Tech II stuff on turn 4 may be in trouble.
-
On the other hand, if they’d rather delay, P2 can absolutely afford to wall up behind two Argonauts.
But: how often are you going to draw both techs on turn 3? More than you might think, if you play your cards right.
Beginning of turn one, you have 5 cards in hand, 5 in your deck, and no cards in play. (We don’t care about tokens or heros – they don’t affect card cycling.) We’ll represent that as 5🖐️ 5📚 0♻️ 0🕴️.
Standard Line – play 1 worker and 1 unit/upgrade/building card every turn, stay at 5 cards, played card stays out of discard.
Turn 1: Play 1, worker, discard 3, draw 5: 5🖐️ 0📚 3♻️ 1🕴️
Turn 2: Tech 2, play 1, worker, discard 3, shuffle 8, draw 5: 5🖐️ 3📚 0♻️ 2🕴️
Turn 3 hand is 5 cards uniformly picked from your 2 techs + 3 turn 1 discards + 3 turn 2 discards. There’s 8!/5!3! = 6x7x8 / 2x3 = 56 ways to split 8 cards into piles of 5 and 3. Of those 56, only 6 put both teched cards in the bottom 3 – one way for each of the remaining 6 cards to occupy the 3rd slot when your teched cards are holding two. Getting both cards means splitting the remaining 6 evenly between hand and deck = 6!/3!3! = 4x5x6 / 2x3 = 20 ways to get both; we’ll call that a double. The remaining 30 are drawing exactly one tech, a split.
Note that this is equivalent to playing 2 things on either or both turns, with 1 of the things being a unit that your opponent helpfully kills in Technician.
Stay at 5 cards, keep 2 cards in play:
6/56 = 10.7% whiff
30/56 = 53.5% split
20/56 = 35.7% double
A double isn’t necessarily better than a split – you probably have to go down a card to play both, and there’s a chance you will only have starter cards turn 4 – but slapping down two Bone Collectors or Nimble Fencer + Two Step turn 3 can give a lot of gas to Player 1’s beatdown. A whiff is probably bad; if you didn’t want those cards in your hand, you wouldn’t have teched them!
Unit + Spell / Dead Unit – as above, but one card played ends up in your discard.
Turn 1: Play 1, worker, discard 3, draw 5: 5🖐️ 0📚 3♻️ 1🕴️
Turn 2: Tech 2, cast=discard 1, worker, discard 3, shuffle 9, draw 5: 5🖐️ 4📚 0♻️ 1🕴️
9!/5!4! = 6x7x8x9 / 2x3x4 = 7x2x9 = 126 to divide 9 cards into 5 and 4 cards. Putting both techs in the bottom four means splitting remaining 7 into 5 and 2 = 7!/5!2! = 6x7/2 = 21 ways. Both on top is 7!/3!4! = 5x6x7 / 2x3 = 35 ways, leaving 126 - 65 = 70 ways to split. (note: 126 is 7x18, possibilities are all multiples of 7)
Stay at 5 cards, keep 1 card in play:
3/18 = 16.7% whiff
10/18 = 55.6% split
5/18 = 27.7% double
That’s noticeably different. If you were deciding whether to kill a unit to make your opponent draw from 9 instead of 8 cards, and you did that over 100 games, there’d be 8 fewer games they draw both, 6 more they get none, and 2 more where they split evenly; same if you’re deciding between casting Summon Skeletons or Jandra.
What if you can’t keep any cards in play? Say, turn 1 Garth + Pestering Haunt + Skeleton, they kill your Haunt, turn 2 Summon Skeletons. Or turn 1 Fox Primus, turn 2 Grappling Hook to pull their hero into Lookout, Fox Primus trade with their hero.
Dead Unit + Spell – all played cards go to discard before end of turn 2.
Turn 1: Turn 1: play 1, worker, discard 3, draw 5, unit dies: 5🖐️ 0📚 4♻️ 0🕴️
Turn 2: Tech 2, cast=discard 1, worker, discard 3, shuffle 10, draw 5: 5🖐️ 5📚 0♻️ 0🕴️
10!/5!5! = 6x7x8x9x10 / [2x3]x4x5 = 7x2x9x2 = 9x28.
8!/3!5! = 6x7x8 / 2x3 = 2x28 (same for whiff and double)
8!/4!4! = 5x6x7x8 / 2x3x4 = 5x14 ways to have Card A and then Card B, 5x14 ways to get them in the other order.
Stay at five cards, only heros/tokens/add-ons in play
2/9 = 22.2% whiff
5/9 = 55.6% split
2/9 = 22.2% double
Keeping nothing in play vs. keeping 2 units gives about +120% chance of whiffing (~22% vs ~10%), at the expense of hands that would have been doubles.
Thinking about going down to 4 cards for turn 3? It makes a big difference whether you keep two or three things in play.
Unit + Two Units – aka Hardened Mox -> Fargo + Battle Suits.
Turn 1: play 1, worker, discard 3, draw 5: 5🖐️ 0📚 3♻️ 1🕴️
Turn 2: Tech 2, play 2, worker, discard 2, shuffle 7, draw 4: 4🖐️ 3📚 0♻️ 3🕴️
7!/4!3! = 35 possibilities, 5 ways to leave both in bottom 3, double is 5!/2!3! = 10, split is 21.
Drop to 4 on T2, three things in play
1/7 = 14.3% whiff
4/7 ~ 57.1% split
2/7 = 28.6% double
So, a little better than staying at 5 with one thing in play, but a little worse than 5 cards with 2 things. If a thing you teched draws a card, or they kill a turn-2 unit in technician before your turn 3, you come out about the same as the standard line (exactly the same for the technician draw) in terms of hitting your techs on turn 3.
Going down to 4 cards on the first turn is really bad, even if you keep everything in play (your opponent can do you a big favor by trading with a unit in Technician in this case):
Turn 1: Play 2, worker, discard 2, draw 4: 4🖐️ 1📚 2♻️ 2🕴️
Turn 2: Tech 2, Play 1, worker, discard 2, draw 1, shuffle 6, draw 3. 4🖐️ 3📚 0♻️ 3🕴️
Even though you’re still drawing 4 cards, only 3 of them come from the shuffle that includes your techs. So you’re as likely to whiff as double – 4 ways each, out of 6!/3!3! = 4x5x6/2x3 = 20.
Drop to 4 on T1, three things in play
1/5 = 20% whiff
3/5 = 60% split
1/5 = 20% double
If you lose one unit, you get the inverse of dropping to 4 on T2 – 2/7 whiff, 1/7 double.
If you lose both after turn 1, you’re drawing 3 out of 8 and you get the inverse of the standard line – 10.3% double, 35% whiff.
Conclusion: turn 3 whiffs happen, but if you play conservatively and keep your critters in play, you can avoid it 85-90% of the time.
Special Cases
There are a few special cases worth mentioning:
P1 T1 Thieving Imp – starting the game down a card.
You’re playing second, and your opponent drops a Thieving Imp and ends the turn. What do you do? If you play a unit, you go down to 4 cards on turn 1, and one card’s already in your discards. Here’s what happens if you play “the standard line” but with only four cards:
T1: (Discard 1) Play 1, worker, discard 2, draw 4. 4🖐️ 1📚 3♻️ 1🕴️
T2: Tech 2, play 1, worker, discard 2, draw 1, shuffle 7, draw 3. 4🖐️ 4📚 0♻️ 2🕴️
Drawing 3 from 7 is… not great. It’s the inverse of making 2 plays T2 – 2/7 to whiff, 4/7 to split, 1/7 to double – and you’ve only got 2 things in play, rather than 3 – and you only have 4 cards in your hand, so there’s a 5/7 chance you need to skip playing one of your early techs to get back up to 5. Losing your a unit in Technician on P1 T3 gets you back to 5 cards and drawing 1 + (4 of 7), but then of course you’re down a unit and however much gold you spent on it.
Assuming you still have 2 things in play. Fortunately, Player 1 can’t afford a hero + Sacrifice the Weak on T2 (unless you play StW on their Imp in Scavenger, which puts your T1 play in the discard anyway). But they can still Deteriorate a 2/1 or 1/1, or trade Imp with a unit.
If your unit gets killed not in Technician, or you open hero + spell, you get
T1: (Discard 1), Play 1, worker, discard 2, draw 4, bury 1: 4🖐️ 1📚 4♻️ 0🕴️
T2: Tech 2, play 1, worker, discard 2, draw 1, shuffle 8, draw 3: 4🖐️ 5📚 0♻️ 1🕴️
-or-
T2: Tech 2, play 0, worker, discard 3, draw 1, shuffle 9, draw 4: 5🖐️ 5📚 0♻️ 0🕴️
drawing 3 from 8, inverse of standard line – 35% whiff, 55% split, 10% double – unless you skip your T2 play, drawing 4 from 9 is 28% whiff, 55% split, 17% double, which is better except you enter T3 with only 1 hero and 1 add-on in play. So if you do choose to play from hand T1, don’t let it go to the discard!
If you lose your unit in Technician, all is well; it’s the same as being at 5 cards and playing a spell T1 or losing your unit in Squad Leader: 1/6 whiff, 3.5/6 split, 1.5/6 double, except that you’re out some gold you spent on it. Skipping a hand play either T1 or T2 (doesn’t matter which) and keeping your other play alive gives you the same odds.
T2 Young Treant / Lawful Search – when you absolutely gotta get your tech turn 3
Here’s a neat thing: If you play normally on turn 1, then draw an extra card during turn 2 – not before – you cycle your deck early, before you put your discards from turn 2 into it.
During turn 2, your board looks like this: 5🖐️ 0📚 3+2♻️ 1🕴️, where 3+2 shows the 3 t1 discards + your first 2 techs.
If you then play Young Treant/Lawful Search, it looks like this: 5🖐️ 4📚 0-1♻️ 1-2🕴️ (note that Lawful Search goes to discard after resolving, so it isn’t in the shuffle to itself.)
You just shuffled 5 cards and took 1. There’s an 80% chance you still have both your techs in your deck for next turn, and 100.0% chance you have at least one. Even if you drew a tech, you still might get it back when you shuffle this turn’s discards to draw your 5th card! If you play a another card and discard 3 cards, you have 80% + [1/3 to 1/4] * 20% = 85~87% chance of a double depending on if you have 3 or 4 cards in your discard. Or, a minimum of 83% if you play Lawful Search, Manufactured Truth, and then trade your unit for 6 in discard – or 100.0% if Bigby Hayes is around to stash the card you drew.
Purple can do some more complicated shenanigans with Temporal Research or Geiger’s midband, offset by the complications that it’s 2g that requires a hero, but does not protect it, and you’ll probably draw more cards than you can play. Unless you really like Vortoss Emblems, which, maybe you do?