Friday, April 17, 2015

N+1 Problem Redux

Listening to the latest High Drag podcast episode I was struck by a misconception some of the hosts had that I heard before. Specifically, that FozzieSov will break up large fleets into several smaller fleets during an invasion or fight over a particular station or structure as they move to cover all the command nodes exposed, thus alleviating the N+1 problem that plagued the fights in Dominion Sov.

To reiterate from a post I wrote last October when discussing the Phoebe capital ship jump changes, the N+1 problem can be stated as thus:
All else being equal, the side with n resources is at a disadvantage to the side with n+1 resources.
There is nothing in Fozzie Sov that will prevent this axiom from continuing to hold sway.

Take for example a fight in Faction Warfare over a system. Two opposing fleets are clashing to run the plexes and either keep control or take control of the system. All else being equal, the side with more pilots in ships will win the battles, drive the opponent out, and have more time to run the plexes. There are a lot of variables in play, of course, such as what fleet hull size can fit in which sized complexes, like if the defender is flying a fleet of 10 frigates in a novice plex and the mixed 15 ship attacking fleet can only get 7 of their frigate ships into the novice plex, and these situations create terrain and conditions that requires planning and proper execution to overcome. But in the end the side that has the bigger fleet and can cover all the plexes has the advantage and will win the system.

In Fozzie Sov, the scenario is different yet surprisingly similar. As far as I know, their won't be any gates preventing ships of certain sizes from helping take over command nodes and the command nodes being attacked are spread over a constellation instead of a single system. But in the end the side with more ships and superior coordination and discipline will win the battles and complete the command node hacking. There is no where for a small fleet to hide and at the same time prevent the command nodes from being capped.

Further to this, I am not sure that fleets will even break into smaller fleets to attack. In Faction Warfare, a fleet may spread its forces around to multiple plexes, but a single commander still directs the fleet and any threat see the disparate pilots coalesce on a point to attack or retreat. I suspect fleets in Fozzie Sov will react the same way, although the large distances involved may prove me wrong.

All of this to say, Fozzie Sov does not solve the N+1 problem.

What it does do, however, is introduce more variables into the decision making and actions of the fleets involved. Where Dominion Sov was characterized by massive fleets smashing into each other at one timer or another, Fozzie Sov will hopefully be characterized by more dynamic fights over multiple systems where the battlefield terrain and distributed decisions will hopefully allow for more things to be unequal over a fight and give more chances to a smaller but more coordinated and capable fleet to win the fight against the larger but less professional fleet.

In other words, Fozzie Sov wants to turn this:

Into this:

And everyone in null sec and beyond is eagerly waiting to see if it is going to happen.


  1. the one big difference from FW to Fozziesov is that instead of one system in FW Fozzie sov will spread the nodes out over multi-systems. So backing each squad up is not a warp to but jump to system.

  2. You're forgetting bubbles. Suppose I have 70 guys and the defender has 100. If he puts 95 in one system and 1 in each of the others I'd scout the blob and pick off the lone dudes and entosis those sites. If he spreads I'll find a nice group of 20 and hit them with 60 people, using dictors to fend off reinforcements.

  3. "All else being equal, the side with n resources is at a disadvantage to the side with n+1 resources." is obvious and can't be changed.

    The problem with Dominion Sov was "all else doesn't matter, the side with n pilots is at disadvantage to the side with n+1 pilots".

    "smaller but more coordinated and capable fleet to win the fight against the larger but less professional fleet." was the goal, therefore problem solved.

  4. Nothing changes. The bigger blob still wins in 99.9% of cases.

  5. I believe the answer lies with a dynamic system. Occupying or flipping a contestation point should spawn more of them, assuming others are occupied as well. To address a red queen problem you need a red queen solution.

    Beyond that, we really need to be giving shape to the battlefield. Different systems need different global effects, thus restricting the flexibility with which we deploy forces. Timers aren't enough. If anything, timers should be the object of fighting, not the gating function.

    Sovereignty levels themselves should be conferring defensive advantages, but every defense needs some sort of exposure to weakness in the periphery. There should always be a set of systems that are better to attack than another set. A highly developed, highly protected system should need to be surrounded by slightly less developed systems, and so on. Keep entosis out of core systems, and the devs are free to make it an effective process in more vulnerable, but still necessary systems. Players desire to have a sense of where the front line is located.