lavaspout, why?

As long as the CD of Rapid Growth creates a scenario where it can be used 2 times or more before the champ dies, it'll be potentially infinite. If can only be used once, then it'll never grow beyond 2 or 3, depending on exact timing of death.

 

I think the best way to think about Rapid Growth champions is to calculate the value per rune over time. Every time it replicates, you are reducing the cost of the units by the value of an addition champion.

So at 70 nora (for easy calc), the Lavaspout on 1st division is 35 nora each. On the next division (which will spawn 2 if they both are used) it becomes 17.5 nora. Here we see that it rapidly becomes extremely cost efficiently which is the real problem.

Now imagine a scenario where the champion itself only costs 50 nora, but each activation costs 25 nora. This means that on the first division it will be 37.5 nora each (i.e., (50 + 25) / 2). The next time it will be (50 + 25 + 25 + 25) / 4 or 31.25 cost each. This is a much gentler efficiency curve.

Note that abilities in general SHOULD generate some kind of efficiency curve for you. More DMG, more HP, more VALUE.

Another thing with the Lavaspout in particular that the Crystalis doesn't have as much is that the value of Lavaspout outside of Rapid Growth increases with additional Lavaspouts. The per unit value of a Lavaspout when there are 10 out is vastly greater than the per unit value when there is only 1 out.

 

A nora cost tied to Rapid Growth would also mean that it doesn't matter how many champs with the ability you have out, your spam is tied to your nora generation, rather than the number, effectively capping your potential growth at Nora Generation / Activation Cost, while also introducing an opportunity cost each time you use it, rather than front loading all the cost to the champion's Initial Cost.

~

Another one of the interesting things about a nora cost based Rapid Growth is that the average cost trends towards the cost per activation.

This is because the weight of the initial cost of the unit goes down over time, while the weight of the Cost of Activations increases. The function will slowly approach the Cost of Activation, but never quite reach it (assuming the Initial Cost and the Activation Cost aren't equal).

Cost per Unit Created = ( Initial Cost / ( 1 + # of Activations ) ) + ( Activation Cost * # of Activations / ( 1 + # of Activations ) )

What this means, interestingly, is that you can also create a scenario where the Cost per Unit Created goes UP over time, instead of down, simply by making the initial cost lower than the activation cost.

For example, if the unit costs 20 nora initially, and 40 nora to Rapid Growth.

One Unit = 20 Nora / Unit
Two Units = 30 Nora / Unit
Three Units = 33.3 Nora / Unit
Four Units = 35 Nora / Unit
And so on.

 

This is an awkward thing with these kinds of units, as it means none of the "basic" functions can be part of the upgrade lines. One approach would be to have the upgrade lines be things like Stat Bonus HP so that the "head" unit is stronger in some way but not mechanically different.

It might also be interesting to have U2 be Lumbering, which would mean that the copies don't have Lumbering. It could also be done with Immobile as well, creating a situation where the copies behave differently than the head.

Lumbering base + Immobile on upgrades would also have a similar effect.

 

Yea Rapid could be an exception but I kind of like the idea of the main having immobile and sending out mobile ones now that cost Nora.

All of those are possibly except for the CD one.