Advanced strategy guide : Gambits

= Advanced strategy guide : Gambits = This article will deal with gambits. Gambits are both funny and risky : they can win you a game through the accumulation of little benefits, or lose it on the spot if you make a miscalculation. That's a high risk, low reward strategy, so you should think twice before using it.

What is a gambit ?
A gambit is the decision to deliberately not defend fully when you have the ability to do so, and when you are not going breachproof as well. Basically, you decide that your opponent will not attack with all his power on the next turn. This is different from not defending fake threat, as all the threat you are not defending is real, and you leave your opponent with the choice to breach you or not.

Why the risk ?
The answer is simple : greed. Gambiting is greedy by essence. You will gambit when you don't want to spend too many resources on defense, and you prefer to spend it on something else. A good example of that is when you could defend perfectly by holding Drones, but you decide to build a Wall that defends for exact instead, because you want to keep your Drones for later turns. Or you have still a gold in bank at the end of the turn and could hold back a Drone, but you prefer to keep the gold. Often, the player feeling behind will be the one to gambit, in order to regain some advantage.

Why not taking the breach ?
Here's a scenario to illustrate :

''So, you're opponent has offered you a gambit : by attacking with all your resources, you can breach him for exact. Cool, isn't it ? He even had spare gold and he could hold a Drone ! What a mistake, now you get to kill all his defenses ! Or can you ? Problem : to do this, you need to absorb on a Rhino instead of a Wall, then click the Rhino. Now, you're left with no more defenses up. And he's going to attack with a lot on the next turn, too much for you to defend actually. If only you still had a Wall... OK, too bad for the breach, you rewind your defense and absorb on the Wall and let the Rhino die.''

What makes a gambit successful is that in order to accept it, your opponent will have to commit a lot of his defense resources on offense, and will not be able to defend against your next attack. To take a gambit, your opponent will often have to absorb on an attacker, on attack with their absorber (like Omega Splitter), so he is suddenly faced with a lack of soak and absorb he has to rebuild in a single turn in order to make his defense. On the other hand, the immediate gain is not much from his side : you offer a gambit for exact, so he will not kill much of your own threat.

Easy or risky gambits
Some gambits are pretty easy to take as you can safely guess that your opponent will never gain anything from taking it. On the other hand, some as so risky that you can lose the game on the spot if your opponent takes it.

1-lifespan attackers
1-lifespan attackers (Grimbotch, Chieftain and Doomed Mech) are pretty easy to gambit against. These units get the maximum value by attacking with them during the first turns, and blocking with them on their last turn. If you gambit against them, then it will require your opponent to click them and not take advantage of their life, which equals taking damage.

Try to think in his place : if you click a Grimboth on 1-lifespan instead of soaking on it, you lose 2 health in order to deal that one damage, meaning that to make it count, you need to at least breach a Wall by doing that (if you breach a Wall, you deny the 2 absorb, which roughly equals the 2 health you lose). However, you also lose the 2 health defense tempo that Grimbotch provides, meaning that you have to pride those 2 health by buying defense. Usually, it is not worth it and you'd better keep you Grimbotch on defense.

Against 2 Grimbotches on 1 lifespan, the loss of the double click is 4 health, so you can usually gambit for at least one Grimbotch click pretty safely. Against a Chieftain, that's 7 health your opponent is going to lose, and 7 defense tempo. Your opponent really has to kill something valuable (like Centurion, or breaching some of your damage) to make it count. This is also a safe option. Against a Doomed Mech, that's 5 health that you opponent will want to soak on, so once again it's pretty safe. Not to mention that often Doomed Mech can be used as absorber too : the one on 1-lifespan is held as soak, and another one is held as absorber. Clicking both would not only result in 5 more damage from the dead Mech, but also 9 total defense tempo (from both Mechs). This is also an easy gambit.

Always greedier
You can gambit for more than exact. For instance, if you believe your opponent has to keep his Omega Splitter no matter what, then you can defend for 3 less than necessary, because you have evaluated that threat as fake threat. Of course, the more you offer, the higher the risk if he does take it, but the higher the reward if he does not.

One other thing to consider is the degree of breachproof form both parties : if all your attackers are breachproof, then it will not be that bad if he takes it. On the other hand, if he can kill Tarsiers or Shadowfangs while breaching, you will lose a lot of pressure and he might be able to defend that critical turn. The same goes for him : if he is breachproof, then he may easily take your breach by offering one of his own but will end up ahead when both your defenses will be down. Offering a gambit for more than exact is a seriously risky business in this game.

Don't gambit a Centurion !
Well, Prismata is a deep game, so maybe there are cases where it is legitimate to gambit a Centurion, but... The idea is to avoid gambiting using a threatening absorber (Centurion, Omega Splitter, Arka Sodara ...) because by taking the gambit, your opponent will take out some threatening pressure at the same time as getting a breach. It can work, but it will cost you a lot, so make sure than it costs more for your opponent.

Should I take the gambit ?
Gambiting is interesting, because if the decision to make a gambit or not is a hard and risky one, the decision to accept it is no less hard and risky.

When your opponent is offering you a gambit, should you take it ? The answer depends on time controls and your time bank. Considering a gambit takes time, so if you low on it, you should give it little thought. If not, however, play your defense as if you are accepting the gambit. Then, try to defend against your opponent's next attack. If you end up in a worse position than what you gain from the gambit, then restart your turn from your defense and refuse it.

A good gambit
No matter how risky, a good gambit is a gambit your opponent has to think carefully before taking it. Whenever you decide to gambit, after weighting the gains and losses on both side, you should once again step in your opponent's shoes and try to imagine how you would defend in his stead if you took the breach. That's right : play his next turn in your head, just like when you're offered a gambit yourself ! If you end up with an easy defense, then that is a bad gambit. If you have a hard time defending, and you end up with more losses than the gambit offered, then it is a good gambit and you can take it.

The gambit chaos
Sometimes, to respond to a gambit, whether you take it or not, you can offer your own gambit. When one is taken, it can quickly turn into a chaos situation where both players offer a breach on each turn. Usually, that is because the defense becomes quite desperate, so, if you have the time, you should try to take it on each turn it is proposed. Soak supplies of the overuse of Forcefields can make the defense harder, and the once good gambit will end up turning a bad one.

Gambits are for losers (but not only)
If you believe you're losing the game, you will end up gambiting more aggressively in order to regain some footing, because if you don't, you will slowly lose the game anyway. So by gambiting, not only can you gain some little advantages here and there, but you can also pull the game into an uncertain state if your opponent takes it, which is good for you because uncertain is better than lost. Let's be real, though : if you're behind, gambiting will usually accelerate the end of the game, because your opponent will be in a better position to take the gambits anyway. But it's good : you'll still lose, but you'll lose faster !

If you believe you're ahead, however, stay away from the risky business. Be it making risky gambits or taking risky gambits, if you go one step too far, your opponent will be delighted to transform a losing position into an uncertain, or winning one. Stick to the safer gambits, or do not gambit at all.

If you don't know whether you are ahead or not, or the game is still close, it's a good time for making some gambits for exact that your opponent will have a hard time taking in order to gain the upper hand.

Gambiting units
A gambit is not always offering a breach : you can also gambit units. The most gambited one is the Rhino : basically, everytime you buy a Rhino to defend an attack when your opponent has a threatening absorber, you actually gambit the Rhino against the absorber click. The reward for winning the gambit is to be able to get the full value of your Rhino by dealing its two charges as damage and then using it as soak, the risk is to lose that if your opponent takes the gambit. Gambiting a Rhino is less risky than gambiting for a breach, that's why Rhino is a good defender against absorber threat. In the same way, you can gambit any attacker, like holding a Steelsplitter, and Odin... You can also gambit valuable defensive units like Xaetron : by not overdefending your Xaetron completely, you can offer your opponent to kill it (or kill your other main absorber) if he takes your gambit.

In a nutshell
In a nutshell, gambiting correctly is hard and the consequences of a mistake are usually a straight loss. That's why you should carefully weight the pros and cons of your opponent taking the gambit (or you taking your opponent's) and consider the consequences by playing in your head the following turns. All in all, it is still a powerful resource to win games.

< prev : granularity and abuse | next : freeze >