Create local dropoff if the situation warrants

[dgets]

If the shipyard isn't within a certain # of squares (requiring 2x # of squares to return to), then we should probably just create a dropoff. Be aware that creating a dropoff requires 4,000 halite, so other factors should probably be taken into account as well (surrounding halite, # of ships in the vicinity, etc).

[dgets]

Okay, for what it's worth, creating a local dropoff if there isn't anything within 5 squares is going to be massive overkill. It takes 4k halite to create a dropoff point, after all. So we're going to have to do several things here...

• determine the amount of halite we have in relation to the 4k needed
• determine how long it's taken us to accrue this
• determine whether or not it is worthwhile to create another one (ie will other ships be using it, as well?)
• create the new dropoff point only iff the above criteria indicate that it is a good choice

[dgets]

This is going to be much easier after the MapChunk functionality is implemented, in order to better determine the local concentration of halite. Waiting on this until then... Er, well, at least waiting on that to handle the 3rd checkbox above...

[dgets]

As the above issue reference was creation of a duplicate issue, I'm simply pasting that information into this thread. NOTE: there are new factors taken into account...

New (and probably some duplicate) information & factors follow:

Now that we've got some bots out that are a bit better about spawning and mining, we're ending up with a lot of ships that have almost full halite that are unable to return to the shipyard, due to clogging of the main route & the pawn formation issue (#33). It would be very beneficial to pick those that are congested, and farther away from the shipyard than the others, and create a new dropoff in order to facilitate easier collection of halite.

This will necessitate, at least, the following:

• analysis of the ShipHistory for more than just the last turn, or at least proper turnstamp setting and analysis; however, being as more ShipHistory is going to be required at some point soon, anyway, it'd make sense to just Do It Right(tm) the first time here (or second time, as it were)
• analysis of the game_map in order to determine the proper [full] ship to create a dropoff with; this will be based upon ships surrounding it (at least those in pawn formation and/or full of halite) and distance from the shipyard or other dropoffs
• how long any pawn formation blockade has existed (potentially negating the above factor)
• (at some point) analysis of the closest local maxima to determine if it may be most efficient to create a dropoff anyway