Following the discussions in thread: https://robertsspaceindustries.com/spectrum/community/SC/forum/4/thread/3-10-ifcs-coupled-mode-prioritizing-vector-stabili/220510 I kept thinking about how to implement a single coupled mode that combines the best of the "max thrust" (3.9) coupled mode with the best of the "stable" coupled mode we have now in 3.10. Especially as @YogiKlatt-CIG has mentioned that they would prefer not to add another toggle to switch between different coupled modes. We have talked internally about having more flight modes in the past, but the general consensus is that we want to avoid adding more complexity if it can avoided. I've come up with a solution. I've even created an implementation to check, that it really does what it's supposed to and not just works for a few special cases. The implementation including more detailed explanations can be found at github: https://github.com/deng0/SimpleFlightComputer Basic Goal of the new "anti-drift stable" coupled mode Try to minimize drift as fast as possible with minimal impact for the acceleration in the direction of the desired velocity vector. Drift is defined as any motion that is not in the direction of the desired velocity vector. In the first iteration of my anti-drift mode, the flight computer only made use of it's anti-drift behavior when this would not increase the total time needed to achieve the desired velocity vector. This restriction has since been dropped to make the anti-drift behavior work better in all cases and make it more predictable for the pilot. This has also greatly simplified the algorithm. The increase in total time is negligible. Comparison The following pictures have been created from data generated by the algorithms in the repository. Although these are just 2D representations, the algorithm uses 3D space. They show a view from above with the spaceship's forward axis being up. They illustrate the benefits and weaknesses of the different modes. The left side shows the velocity vectors. The dark blue arrow is the current velocity vector and the light blue arrow is the desired velocity vector. The blue circle shows the speed limit that is currently set by the pilot (maximum ship speed when speed limit is disabled). The red line shows the path of how the velocity vector is transformed when the "max thrust" mode is used. The green line is the "stable" mode and the dashed black line is the "anti-drift stable" mode. While the IFCS is performing the maneuver the direction and speed of the ship can vary greatly between the three modes. This can obviously have large impacts on the flight path which is shown on the right. The blue point shows the location of the ship when the maneuver is started and the lines show the different flight paths. The end of the lines represent the position of the ship when the desired velocity vector is finally achieved. In theses examples the ship's thrust values and mass are set to those of an Avenger. (All these parameters including any possible combination of current and desired velocity vector can be fully customized.) In the first image the case is as follows. The ship is traveling sideways with having reached its currently set speed limit. The pilot now determines that he wants to go forward instead. Because of the fact that the aft thrusters of the ship are stronger in the "max thrust" mode halfway through the maneuver they will have finished their work and only the side thrusters will still be firing resulting in a change of acceleration direction, but also that the ship will carry a lot more forward speed through most of the maneuver. This will also have the effect that for a part of the maneuver the ship will exceed the speed limit set by the pilot (which can have negative impacts on gimbal performance). The two "stable" modes will throttle the aft thrusters to achieve stable acceleration throughout the maneuver. Typically the aft thrusters will overheat faster than side thrusters and because the aft thrusters are not going full power, this should make it possible to use the afterburner for a longer duration in the beginning of the maneuver than would be possible with the "max thrust" mode, which will slightly improve the forward speed but more importantly will make the "stable" modes outperform the "max thrust" mode when it comes to eliminating drift in cases like these. The second image show a similar case but this time the pilot has no speed limit set or is currently traveling way below the speed limit. This shows the main problem of the normal "stable" coupled mode which is used in Star Citizen 3.10, namely that it is very bad at countering the drift. The improved "anti-drift stable" coupled mode is capable of determining that in cases like this it's more important to counter the drift as fast as possible than to have a purely constant acceleration. These cases show that the "anti-drift stable" does not just work in special cases like the one above. Additionally once it has successfully countered the drift it will perfectly align with the desired acceleration direction resulting in a smoother flight path. The second image shows that the "anti-drift" mode even changes thrust direction along the lateral thrust axis during the maneuver to achieve a better flight path. In this case the ship is currently in standstill and the pilot sets a new desired velocity vector. In "max thrust" mode the ship will accelerate in a direction that does not match the direction of the desired velocity vector resulting in flight path that is not straight. In both "stable" modes the flight path perfectly matches the desired direction. This will be perfectly suited when trying to perform precise docking/landing maneuvers. In this case the pilot wants to slow down, but also adjust flight direction. The anti-drift feature allows the pilot to more effectively steer while slowing down. Conclusion In all cases, I can think of, the "anti-drift stable" mode described here would be better suited than the simple "stable" mode currently implemented in Star Citizen 3.10. Only in some very specific situations the basic "max thrust" mode might provide better results, but that might be mitigated by the fact that in these same situations "anti-drift stable" mode might allow for longer afterburner durations. So I would suggest that CIG updates their coupled mode to something similar to this "anti-drift stable" mode. Then they could still add a toggle to switch between "anti-drift stable" and "max thrust" coupled mode, but I don't think that would be necessary. It would be great if CIG could take a closer look at it. Update: I've made significant improvements to the algorithm. The results are even better now across the board. Additionally the source code and the underlying mathematics are much simpler and straight forward now, making it easier to understand.