More card distribution woes.

How about a two stage game. In the first stage players play some kind of card game in which they try to capture a hand incorporating the cards they need, followed by robot construction from their hand and then the battle.

For example:

1) Players draw an initial hand of say 3 cards, displayed face up.
2) Then they get dealt a number of cards face down
3) Some kind of game ensues based on the information on the face up cards and players individual knowledge of their own private hand.
4) Players then construct their bot from their hand at the end of (3). Since all players have numerous concealed cards, exact bot constructions will remain uncertain.
5) Reveal bots and fight battle. This could be augumented by cards collected and saved during phase 3 which can be played during the battle.

Alternately, if powerplants and weapons have different power levels, give each player X points to "purchase" a powerplant and a weapon before the game, and let them choose a crappy weapon and a great powerplant or a great weapon and a crappy powerplant or a middlin' good one of each kind. Alternately, they might buy a middlin' good powerplant and two weak weapons, or something.

What about players starting with different "starter bots" which would basically be play mats, that they would build onto as the game goes. You could make the starter bots different enough to your taste.

Alternatively, and I don't know how random your game is, or if players draw a hand of cards. But you could let players play three rounds before the action starts, basically a build phase. And one of the actions that you could take during those turns would be to draw a fresh set of cards, hopefully finding what they need.

And if you don't have all your bits after that! good luck!

Freshly excell(-ed) for you:

<br />
X   No wp OR pw    No weapon AND no powerplant<br />
1	85,00%			70,00%<br />
2	72,12%			48,79%<br />
3	61,08%			33,85%<br />
4	51,64%			23,38%<br />
5	43,57%			16,08%<br />
6	36,69%			11,00%<br />
7	30,83%			7,49%<br />
8	25,86%			5,07%<br />
9	21,64%			3,42%<br />
10	18,08%			2,29%<br />
11	15,06%			1,53%<br />
12	12,53%			1,01%<br />
13	10,39%			0,67%<br />
14	8,60%			0,44%<br />
15	7,10%			0,28%<br />
16	5,85%			0,18%<br />
17	4,80%			0,12%<br />
18	3,93%			0,08%<br />
19	3,21%			0,05%<br />
20	2,62%			0,03%<br />

First one is the probability that you won't get either a powerplant or a weapon. Second columni is the probability that you'll get neither.

Say that they choose ten cards. They've got a 98 % chance of getting either a powerplant or a weapon but only a 82 %^2 = 72% chance of getting both. But if the player only get's one weapon and draws 9 new cards he has a 79 % chance of getting a powerplant amongst those cards.

(BTW the name of the first column is missleading. I mean that if you draw 1 card you have 85% risk of not getting a card from a set category, not a logical OR. Disclaimer: I suck at math ;))

Filwi's numbers seem strong enough for a ten-card deal. If the player doesn't have both required parts, they simple get a new deal of ten, which would happen about one out of five hands. That's not too bad, really.

On the other hand, a player may get nine power plants and one weapon, which might be equally rediculous, but that's what the game is about, it sounds like.

Why not just have the Power Plant and Weapons cards in thier own piles. Possibly using different card backs to distinguish between them.

Let the players draw one of each plus some from the main pile. Create new 'draw power plant' and 'draw weapon' cards for the main pile to keep the same odds as before.