Gameplay Decision

As most of you may not know, I am in production of a game called Demonic Roots. I've made the basic engine and am proceeding in actual production of the game. However, recent problems have brought into question a certain swimming method I have.

Ok, pretend this ~ and this - are water, and this | is the player...

~~~~~~~~~~~~~~~~~
--------------|----------------
-------------------------------
-------------------------------

The way my current method of swimming works is when the player presses the left or right key...

~~~~~~~~~~~~~~~~~
--------------\----------------
-------------------------------
-------------------------------

The player object tilts.
When they press up...

~~~~~~~~~~~~~~~~~
-------------------------------
-------------------\-----------
-------------------------------

The player object goes forward in the direction he is tilted towards.

Now this is all fine and good, but I recently have made a certain object which transports the player to the next room through contact. I noticed that 75% of the time, the player would not go into the next room, and I had to go back and forth until it worked.
This was a mystery to me why this happened until I happened to be screwing about in de-bug mode. I asked the program to record the player's x and y values, and found that when in the water, the x would go from, say 475 to 476.98330923587.
In short, the player was a fraction of a pixel off the regular grid, thus causing collision problems.
Knowing this would frusterate players during gameplay, and the fact that the current method of swimming already has problems, I have begun considering a new way to interact in water...

~~~~~~~~~~~~~~
-----------|--------------
--------------------------
--------------------------

Instead of tilting...

~~~~~~~~~~~~~~
--------------------------
------------|-------------
--------------------------

~~~~~~~~~~~~~~
--------------------------
--------------------------
-------------|------------

...Why not just decrease the gravity? That way, I won't have already existing problems with the current swimming method AND those object's will work!
Only thing stopping me, however, is that quite a few objects are only activated when in the water, and I'm afraid the new way might cause other, internal problems. I don't know what to do, SO...

YOU decide!
Should I keep with the tilting method, besides the quirks, or should I invest time into the new low-gravity method and take the risks?

Unless you are wearing gads of armor, I say no gravity. When in water, make it eight-way control. Except in the case of where not sinking feels out of place due to excess weight, it only frustrates the player to make the character always sink. It's a sort of fake difficulty that doesn't do anyone any good. An employment of friction is fine, but low-gravity doesn't well represent implied physics (the ones you think about, not necessarily real physics).

The tilting also seems to me like it is out of place. I don't know about you, but I swim with my body parallel to the direction I'm heading. It'll mean more sprites, but it should look a lot better.

Now, if you do decide to go with the tilting method, your problems should be fixed by rounding or truncating your x in every step. That is, try something like x=round(x).

dixee wrote:Unless you are wearing gads of armor, I say no gravity. When in water, make it eight-way control. Except in the case of where not sinking feels out of place due to excess weight, it only frustrates the player to make the character always sink. It's a sort of fake difficulty that doesn't do anyone any good. An employment of friction is fine, but low-gravity doesn't well represent implied physics (the ones you think about, not necessarily real physics).

I don't quite understand what you mean by eight-way control. Also, have you ever played an old Mario game? You know the way Mario swims through water? That's what I'm implying here.

dixee wrote:I don't know about you, but I swim with my body parallel to the direction I'm heading.

I wouldn't know a thing about that. I live on a cloud.

dixee wrote:Now, if you do decide to go with the tilting method, your problems should be fixed by rounding or truncating your x in every step. That is, try something like x=round(x).

Aha, that should work nicely! However, tilting has other problems to it...

Eight-way control is where you have eight possible directions to go in. I didn't make it clear, but I meant up, down, left, right, up-left, up-right, down-left, and down-right.

Yes, I have played Mario. I know exactly what it's like. It feels like you have infinite jumps and are on the moon. Not like swimming. If you like that method, by all means use it. I don't like it. So, I don't use it.

Um... gravity would be easier for me to play but the other one sounds unique so... I'd prefer unique but then I'd probably fail when playing it... ARG! I'm so torn!

You wouldn't fail if it's done right. If there's a proper learning stage, a well crafted learning curve, and it's implementation is good, you'll have nothing to fret over.