flavor later

Deceive should be written as an activated ability, and definitely needs a cost associated with it. It's probably also too much on-board trick to have at common in any meaningful amount

To be clear, it is an activated ability. Its just reminder text.

As to the "on-board trick" of it, it is a really minor one. It is part of a class of abilities that are extremely normal at common. This ability was originally inspired by rootwallas for example.

Granted, I haven't done a ton of playtests, and I'm sure there are some board states that a few decieve effects makes too complicated, but in the playtests I've done, it hasn't felt any more complicated than first strike. It is a thing you have to take into account when making attacks and assigning blocks, but isn't too much of a burden.

This is an interesting mechanic, I'd love to know how it played during playtests.

I do feel it could cause decision paralysis with a bunch of deceiving dudes on the board. Maybe you could apply one of many possible simplifications:

1. Make it a triggered ability like bushido.

2. Remove the parameter and set it to a single value everywhere (e.g. +2/-2 always).

3. At least make all commons share the same conceptual effect, like every deceiving common creature becomes an X/1.

Anyway, please share some of your playtest results.

In terms of playtests, mostly it acted like "when ~ attacks and isn't blocked, ~ gets +K/+0 until end of turn." It incentivizes trading when possible, and just generally lets deceive creatures punch a bit outside their weight class, kind of like first strike does.

With modern design, it would definitely be a triggered ability just to make digital play better. Nailing the exact trigger condition is a bit awkward though. Options are...

1. Whenever ~ attacks or blocks, you may give ~ +K/-K until end of turn.
2. Whenever blockers are declared, you may give ~ +K/-K until end of turn.

...Option one makes the ability more defensive oriented, because making decisions after blocks is much better than after attacks, which I don't love. The second ability triggers after blockers but has really weird wording.

I'm concerned about a lack of design space unless it is parameterized. I suppose I'll try to do +2/-2 as much as possible and see if I can come up with enough interesting designs to remove the paramter, but I don't like my odds.

Making deceive commons all go to the same toughness seems reasonable.