[Freeciv-Dev] Re: (PR#4387) The Return of the Rand() Moves

["Christian Knoke" <chrisk@xxxxxxxxx>]

<URL: http://bugs.freeciv.org/Ticket/Display.html?id=4387 >

On Mon, Mar 07, 2005 at 04:11:42AM -0800, Christian Knoke wrote:
> 
> I add another suggestion for an alternative rule:
> 
>     "reach target if you have more than half of the movepoints required to
>     enter terrain type"
> 
> Now we have 4 possible rules:
> 
> Subject: When does a unit reach a tile if act. MP < terrain request?
> 
> 
> Rule A (current rule): reach target tile with random chance (act.MP/tile
>         request) percent OR act. MP = full MP
> 
> Rule B: reach target only if act. MP >= tile request OR act.MP = full MP
> 
> Rule C: reach target only if act. MP >= tile request OR act. MP >= 1 MP
> 
> Rule D: reach target if act. MP > (tile request / 2) OR act. MP = full MP

> Design board once suggested to have the 50% threshold in Rule D as an
> option.

On Mon, Mar 07, 2005 at 05:09:22AM -0800, Per I. Mathisen wrote:
> 

> Rule E: reach target if MP > 0

This is equivalent to 0% threshold in D, above.

> IIRC civ3 uses this, and upon playtesting, I find this the most enjoyable
> to play with. I don't need to fuss around with units with left over MP
> that can't move. Units get where they should slightly faster, and it makes
> using roads and rivers more strategically important.

Maybe this is an improvement, but as you say, it has not a small effect on
the movement and thus on the tactical effects of units, and on balancing
between unit types. We should be a bit careful about this.

> Rules B-D above suffer from the 'press space to make unit wait until next
> turn' problem. We could solve this by auto-'space' it whenever you try to
> move onto something that has higher movecost than your MP, but this may
> not be to everyone's taste.

Is this a big problem? I try moving the unit in different directions and see
whether the unit makes it or not. If it doesn't, I take another choice or
press space. I think this is good - the user has the choice, and its
transparent, it is even possibly to know (compute) the success of the move,
before you try it, which let me think that this should be client implemented
(at least some pre-decision), so that is is does not need network interaction
(in all cases) and gives faster play in online games.

Christian

-- 
Christian Knoke            * * *            http://cknoke.de
* * * * * * * * *  Ceterum censeo Microsoft esse dividendum.