Decline in Pathfinding Abilities?

    This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

    • Decline in Pathfinding Abilities?

      Lately my units seem to be yearning for a journey over the ocean rather than a rapid land advance. Case in point is from my alliance match on 1939 historical world map where a light tank would rather take a 9hr and 10min sea trip from Odessa to Kiliya instead of the 2hr and 6min journey across land, and the case is the same for the entire Romanian front along the Black Sea.
      Images
      • Capture.PNG

        296.3 kB, 463×540, viewed 63 times
      • Capture2.PNG

        302.22 kB, 448×575, viewed 31 times
      :00000441: Forum Gang Commissar :00000441:

      Black Lives Matter!!!!! All Lives Matter!!!!! :thumbsup: :thumbsup: :thumbsup: :thumbsup:



    • Well I think an airplane convoy would travel faster by water than land, so that doesn't seem far fetched :S
      War is a game that is played with a smile. If you can't smile, grin. If you can't grin keep out of the way til you can. - Winston Churchill



      VorlonFCW
      Retired from Bytro staff as of November 30, 2020.

      >>> Click Here to submit a bug report or support ticket <<<
    • The fact is that the system tries to avoid recently-conquered "friendly" provinces to avoid potential future resistance. But it takes a long time for that effect to wear off. Also, whenever there is a possible path to an enemy province that avoids passing through other enemy provinces, then the pathfinder will route away from the other enemy provinces to avoid premature or unscheduled conflict. Perhaps your objective is assumed to be skipping one province in favor of another. The system assumes that as default.

      It'd be nice if there was an option to turn on pathfinding settings...that is, a high-danger pathfinder for the most direct and fastest route, regardless of potential threats, a mid-danger pathfinder for the most direct and fastest route, but avoiding known opponent armies that are not the target, and a low-danger pathfinder for the most direct and fastest route, but avoiding all possible threats (the current pathfinding seems to be like this third "setting"). Maybe such an ability to change pathfinding settings would be added as a premium account feature for High Command as yet another incentive to pay for the special account services.

      Along those lines (pardon the pun), the pathfinding system often makes mistakes even through high-morale, friendly (or even owned) territory. Surely the reasons are arbitrary. For example, sometimes you have to make a unit move in one direction for a moment before it will allow you to switch to a different direction without running back to the nearest intersection in the wrong direction before proceeding forward (doubling back in a sliver-thin path back to its starting point before continuing on to the destination).

      Well, certainly, the pathfinding system must be pretty complex. But fixing it will require a plethora of use-cases and, who knows how thoroughly the dev team at Bytro tests out all those possibilities in examples such as this, the pathfinding capabilities of the game?
      It seemed like such a waste to destroy an entire battle station just to eliminate one man. But Charlie knew that it was the only way to ensure the absolute and total destruction of Quasi-duck, once and for all.

      The saying, "beating them into submission until payday", is just golden...pun intended.

      R.I.P. Snickers <3
    • T-3PO wrote:

      about a 5 hour difference
      So the sea route is the faster route by 2 hours when embarking and disembarking are not accounted for. This has always been the case and so this isn't really a new bug, as it functions the way it is designed. We might suggest that a value for embark and disembark be added to the path finding code, but that might be something that has been tried before and not worked well. Planning an invasion next to water has always required liberal use of the "add target" button.




      At some point you still have to think to play this game, you can't just expect the computer to win for you. :P :D


      Quasi-duck wrote:

      take a 9hr and 10min sea trip from Odessa to Kiliya instead of the 2hr and 6min journey across land
      9hours 10 minutes - 3 hours embark - 4.5 hours disembark = 1 hour 40 minutes, which is less than the 2 hour 6 minute trip by land. Again nothing new.



      Ciladen W wrote:

      it should account for territory after a province is taken as "friendly."
      Are you talking about pathfinding or are you talking about unit speed?
      War is a game that is played with a smile. If you can't smile, grin. If you can't grin keep out of the way til you can. - Winston Churchill



      VorlonFCW
      Retired from Bytro staff as of November 30, 2020.

      >>> Click Here to submit a bug report or support ticket <<<
    • VorlonFCW wrote:

      So the sea route is the faster route by 2 hours when embarking and disembarking are not accounted for. This has always been the case and so this isn't really a new bug, as it functions the way it is designed. We might suggest that a value for embark and disembark be added to the path finding code, but that might be something that has been tried before and not worked well. Planning an invasion next to water has always required liberal use of the "add target" button.




      At some point you still have to think to play this game, you can't just expect the computer to win for you.
      Lol, the old meme of "It's not a bug it's a feature" :P

      I made this post because I had to use add target all along the Romanian coast from Artzys to Bazargic, to cut off the sea. If I had ordered my generals in real life to do this, and they propositioned a naval landing in Bazargic and to avoid all the other provinces, I would have them summarily shot. It isn't so much the computer doing the work, it is that the computer just makes it more time consuming. If the devs tried to implement it and failed, they need to hire better code monkeys or purchase some stout walking sticks....

      Diabolical wrote:

      The fact is that the system tries to avoid recently-conquered "friendly" provinces to avoid potential future resistance.
      You have grabbed the complete wrong end of the stick lol.
      :00000441: Forum Gang Commissar :00000441:

      Black Lives Matter!!!!! All Lives Matter!!!!! :thumbsup: :thumbsup: :thumbsup: :thumbsup:



    • VorlonFCW wrote:

      Planning an invasion next to water has always required liberal use of the "add target" button.




      At some point you still have to think to play this game, you can't just expect the computer to win for you.

      I feel like I may have requested this before, but at the very least, couldn't the devs implement a chaining "Add" path feature? Instead of having to go and hit "Add" so many times for highly-specified routes, why not let you be able to click once on an "Add" toggle switch. That way, after you've picked an army(s), you can flip the switch and, until you hit the toggle a second time, all mouse clicks on the map are treated as way-points along the most immediate possible path. Also, along these lines, there needs to be a "Remove" button that lets you remove the last waypoint or segment of travel for a selected unit(s).

      Obviously, the "Add" toggle would work efficiently as a means of adding route segments or waypoints while it's activated, but the "Remove" button need not act as a toggle since you would simply be erasing the last currently-existing portion of a unit path and so you could click the button as many times as needed to erase the latter portion(s) of a long route without having to redo the earlier portions....which is particularly tedious if you have to manually re-make a longer winding path just because of one mistake near the end of your chosen course.

      I'm concerned that, since the game engine is quite old (based on Supremacy 1914), it's possible that the pathfinding calculations for multiple segments might not be very efficient and that they might not have relied on a simple linked list of vectors for unit movement which could make coding for a "Remove" button overly-complex. Hopefully, this isn't the case. But if the list of waypoints isn't a linked list or similar programming feature, then adding a "Remove" button might require lots of additional math to retrace a set of vectors while not using the last segment. If that is the case, I really hope they'll adopt the linked list method of tracking waypoint vectors which would require a fairly extensive update to the code.


      Quasi-duck wrote:

      You have grabbed the complete wrong end of the stick lol.
      Why?
      It seemed like such a waste to destroy an entire battle station just to eliminate one man. But Charlie knew that it was the only way to ensure the absolute and total destruction of Quasi-duck, once and for all.

      The saying, "beating them into submission until payday", is just golden...pun intended.

      R.I.P. Snickers <3