This is just a copy/paste from another ticket that was closed to merge with this one. I have experience driving small boats (<3 ton) as well as large warships (~ 4000 and ~ 9500 ton). My experience with vessels in between is limited, but the concepts remain the same.

<copy>

1. When traveling at forward speed, if throttle is reduced to zero, the water resistance should cause the vessel to slow much faster than it currently does. This is not to say that it should come to a stop very quickly, but it should at least lose the top 1/3 of it's speed particularly rapidly. This is especially true of boats traveling on plane (like the speedboat) in which as soon as speed decreases, more of the hull descends into the water, causing even more drag to slow the vessel. If you couple this with reverse throttle, it should decrease speed even faster.

2. Whenever throttle is reduced to zero, the turning capabilities of the vessels should reduce to almost zero. These are not ruddered vessels, in which the wash of water over a rudder creates lift to cause a change in direction. The direction is changed by vectoring the power of the engines off of centerline. Currently, if you let off throttle, if forward or backward speed is maintained, there is absolutely no loss in handling capability, much as if the boat were on wheels.

3. The above issue also creates problems when trying to do skillful boathandling maneuvers. If I am backing down a boat and hold the key to turn left, then apply heavy forward throttle, the stern of the boat should move quickly to the right while also decreasing rearward speed. The inertia should keep the boat traveling somewhat close to its original vector (backwards) even as the heading changes. As it currently stands, what would happen now is that the stern would continue traveling left (changing heading to the right), stopping when rearward motion stops, then moving right (changing heading to the left) when I achieve forward motion.

</copy>

I have no doubt that these physics could be modeled in the current engine, as the helicopter flight model is fairly robust, and many of the concepts are quite similar, just with the axes rotated.

edit: my ~ caused numbers to get referenced to another comment in another issue.

I completely agree here. Turning at low speeds should cause a decrease in turn radius, but also turn rate. As it stands now, turning a boat is more akin to turning a wheeled vehicle, where it must follow the direction the wheels are turned and have forward motion to do so. This causes other issues that I will place in a follow-on post.

edit: spelling/grammar error

I would be happy if they just removed the beeping/burying sounds when placing explosives that carry about as far as the report from a rifle.

This thread may be dead, but here is what I think your issue is. I believe what you are experiencing here is a product of the ArmA flight model's "semi-auto collective" (as I like to call it) to simplify flying for M+K pilots.

When using keybinds for the digital collective raise and collective lower (simply listed as "collective raise" and "collective lower" in the key bindings), the flight model will attempt to keep you in level flight (relative to sea level) when changing helicopter attitude, even if you don't touch the keys to change collective. To replicate this, you can take off, sit in a hover, and then slowly pitch forward. If you do it slow enough, you will achieve forward flight without losing any altitude.

You will only experience altitude change when adjusting attitude if you pitch or roll faster than the collective can catch up (it takes time for collective to change) or if you exceed an attitude in which there is sufficient vertical lift vector for the helicopter's rotors to keep you level.

If you press and hold the collective raise or collective lower keys, it will adjust the collective all the way to maximum or minimum, respectively. When you release it will go back to the collective level for level flight. Tapping the key will cause it to bounce back and forth between these values, depending on how long it is held.

If you seek the results that you desire, use the analogue collective key binds. I'm not sure how they would work with a keyboard since you have no way of knowing what level your collective is at, but it works just fine with an analog throttle peripheral.