This basic strategy guide was adapted from a Reddit post by 307th. This guide is mostly targeted at beginner or intermediate players.
Attack is an Investment
Attacking in Prismata works differently from attacking in other games. In most games, when you attack, there is some risk involved (the attacking units may die). In Prismata, there is no risk in attacking - when you attack, you deal damage to your opponent without receiving any yourself. This guaranteed, risk-free attack means that in Prismata, attacking units can be thought of as investments that generate long-term value, rather than as short-term attempts to kill your opponent or prevent your opponent from killing you.
If you build a Tarsier and your opponent wants to defend its damage with Engineers, they have to build one Engineer a turn. This means that the Tarsier locks down 2 enemy Drones for the rest of the game while costing much less than 2 Drones! In Prismata, attack that does not get absorbed is a better investment than economy.
If attackers are a better investment than economy, why ever make Drones? The reason is that attack can be absorbed. In Prismata when a unit blocks damage, but isn't killed, it regenerates all of its health instantly. We refer to a unit taking damage then regenerating itself as "absorbing" that damage. A blocker can absorb 1 less damage than its health per turn. Absorb is made to be very strong. For example, a Wall can absorb 2 damage per turn. This is like saving you from making 2 engineers per turn. So a Wall that absorbs is freeing up 4 of your drones! Not bad given that it costs 5B (5 gold and a Blue), and 4 drones costs 12EEEE (12 gold and 4 energy). In terms of return on investment, Absorb is better than Attack, which is better than Economy.
If Attack is balanced by Absorb, how is Absorb balanced? Due to the way damage is assigned in Prismata, you can only absorb on 1 unit per turn. So the maximum amount you can absorb per turn in a game is determined by the blocker with the highest health. In the base set, Wall is the biggest blocker, so the maximum absorb is 2 per turn. If Energy Matrix (which has 5 health) is in the random set, then the maximum absorb is 4 per turn.
Since a Wall costs less than 2 Tarsiers, the first 2 Tarsiers we make are bad investments by themselves. The first 2 Tarsiers cost 8RR (8 gold and 2 red) and force your opponent to spend 5B to negate it. So the first 2 Tarsiers aren't a complete waste, but they are definitely inefficient. Investing in attack is only good when you can build up damage fast enough that the initial attack that gets absorbed is compensated for by the later attack after that that does damage. So if the biggest absorber is a Wall you can think of the first 2 Tarsiers as costing 8RR to force out a wall (5B) from your opponent, and also to give you the option to make highly efficient investments in the future (aka Tarsiers that do damage that can't be absorbed).
Once we do have more damage than our opponent can absorb, we want to avoid making Drones, because Attack is a better investment.
The bigger the possible absorb, the bigger the hurdle before attack can be an efficient investment. This means you need to make more Drones, so that you can make enough attack to overcome the absorb quickly. How many Drones you should make in a game is in large part determined by what the biggest absorber in the set is. If it's Wall, you usually want 12-15 drones. If it's bigger than that, you'll want more. In some unusual cases where there are a lot of good aggressive red units and not many good Blue or Green units, neither player will want to make a Blastforge, and so the biggest absorber will be Rhino, meaning that absorb will only be 1! In these cases you want fewer than 12 Drones.
The biggest absorber still matters even if it hasn't been bought yet - so in a game with Wall and Centurion, if your opponent has a Wall out and is still a few turns away from making Centurion, you don't want to rush out attack as soon as you hit 2 damage, because the damage you make now might hit for 1 or 2 turns, but after that it will be negated by Centurion.
Phases of Prismata
Here is a rough outline of how most Prismata games play out. Please note that this is a general model, it is possible that in a small number of games some of these stages are skipped.
1. Economy phase
This is where you're building up your economy. Look for the biggest absorber to determine how greedy to get (aka, how many drones to make) in this phase. As you get more advanced, you can look at other factors: how good are the attackers? Can the biggest absorber get bought? Sometimes the biggest absorber is countered by another unit in the set (for example Defense Grid is countered by Tatsu Nullifier). Sometimes, the biggest absorber is also an attacking unit (for example Omega Splitter or Mahar Rectifier). These units are good to absorb on, but if there's another good absorber in the set like Infusion Grid, you might want to defend with Infusion Grid and attack with your Omega Splitters. If you want to make Zemora or Savior, this is the phase to do it in.
2. Transition phase
This is where you're still building economy, but you're setting up the tech you'll need for the later phases of the game. During this phase it's also common to get a little bit of attack, to force out a defender from your opponent and work towards overcoming their absorb. You also might build smaller absorbers to handle your opponent's preliminary attack efficiently. For example if your opponent only has 1 damage, you might build a Rhino instead of a Wall to defend it. Also, it's OK to just take 1 or 2 damage without absorbing it in this phase if it helps move your build along. Remember not to rush out your biggest absorber in this phase. You don't want to get absorbers until they can actually absorb damage! If you're going to make a super-long-term attacker like Gauss Fabricator or Resophore this is the phase to do it in.
3. Attack and absorb phase
This is where you build your biggest absorber, and start pumping out as much attack as you can. You don't want to build your biggest absorber until it is absorbing at or near its maximum. For example you don't want to build Centurion until it's absorbing at least 4 damage. If it would only absorb 3 damage, you might as well absorb onto a Wall instead, and spend the resources on attackers. This is the phase where you want to focus on getting long-term attackers. A long-term attacker is any unit that can keep attacking until the end of the game, like Tarsier or Tantalum Ray.
4. Attack and defense phase
Once your opponent has more damage than you can absorb, you need to start building some defense to stop from getting breached. You will still have resources left over after building defense at this point, and attack is a better investment than drones, so you should still be building attack. Long-term attackers are still good early on in this phase. Later in this phase you should aim to get short- and medium-term attackers. These are attackers that deal most or all of their damage over a limited period of time. Rhino is a great example of a short-term attacker. Some examples of other short- and medium-term attackers: Tia Thurnax, Plasmafier, Grimbotch, Sentinel, and Grenade Mech. Also, if you're on red-blue tech, this phase is a good time to make a conduit for forcefields.
5. Defense phase
This is when your opponent's attack has reached such a high amount that you have to devote all your resources to not getting breached. If you prepared for this stage in phase 4 by making attackers that turn into defense, or by making a conduit for forcefields, you'll thank yourself in this phase. Hang on and hope your opponent breaks first!
6. Breach(proof) phase
In every game one or both sides eventually get breached. While most games end here after one player wipes out all the economy or damage of the opponent, some sets allow the players to transition to a breach proof position and keep playing. It means that although they can no longer get any absorb and they lose some of their damage every turn because of getting breached, they have units with high health on board which allow them to stay in the game. Example of such units are thorium dynamo, iso kronus, asteri cannon, militia etc. In some endgames both players mutually breach each other and the side with the more breach resistant units win. There are some sets which skip stages 2-5 and jump straight to stage 6 which means that they don't build any defence from the start, only alternate drones with high health and beefy attackers which also have a lot of health. While this kind of scenario is rare, it can be the optimal play in a small number of sets.
Using the Random Set
In Prismata, the Random Set is balanced to be slightly stronger than the Base Set. This is to encourage players to use the Random Set and to make sure each Prismata game is different from the last. Although I use Tarsiers for a lot of my examples here, in actual Prismata games you want to use the Random Set instead of the Base Set where possible. If there's nothing from the Random Set that fills the right role, then it can still be appropriate to build units from the Base Set.
Similarly, the very big units are balanced to be stronger than the smaller units. If Lucina Spinos was no stronger than 4 Grimbotches, then there'd be no reason to ever make Lucina since it's so difficult to set up and buy. So, when you see big units in the set, remember that even though they are difficult to get, they are often still worth buying.
A lot of players make mistakes that are good strategies in other games, but don't work well in Prismata. For example, making too many drones, because they're not used to attackers being efficient invesments, or making too many short-term units like Grimbotches in the early game, because they're used to using attacking units to kill the opponent rather than to secure a long-term advantage. Hopefully this guide gave you a good idea of how Prismata strategy differs from other games.