Offroad Travel Revisited

[MatGB]

I definitely like the idea of ship part as core slot idea—I haven't added any core slots to designs in the patch I'm working on but it's definitely within the specs, there are currently one ship only researchable and buildable with a core slot, and it's, um, not very good as is. With only one researchable part, and that being hideously overpriced, Core hasn't had much love.

But I like the idea of a slot that only specific ships or high end high tech ships have. Hmm, adding a core slot to the Transpatial Hull would give it an actual use if we do this...

[AndrewW]

However, allowing offroading on a map that's basically a graph of interconnected nodes is like allowing units in a dungeon to pass through walls - it's an extremely powerful, disruptive feature that effectively voids a basic (strategic) feature of the game.

Passwall, stoneshape... Not exactly unheard of. Though I agree and the only way I think it maybe should be allowed would be in a very limited way.

[Dilvish]

*Sigh* This De Facto Underling can't help feeling my FO project time is being seemingly squandered in this fashion, though I suppose part of the blame is mine for needing to take so much time in trying to write a response convincing enough to not be idly dismissed.

I'm not sure that my ideas for improving AI choke-point handling will necessarily work even for the current planar starlane map, so it's hard to make a very strong argument based on those ideas. On top of that, wormholes and starlane modifications would indeed, if not accounted for, make flaws/weaknesses in the resulting analysis (though even then the starlane mods would only create a temporary flaw). However, I think I have given some logical reasoning about why I think those things could be adequately dealt with, in ways that wouldn't likely work for offroading. It is my very strong opinion that dealing with wormholes and starlane-mod complications would be tremendously easier than dealing with the offroading proposal, which seems to rampantly break planarity and explodes connectivity/density.

You appear to be saying that because my plan has some admitted uncertainties you find my opinion here vaguely unconvincing, but I don't see you making any effort to present a convincing counter-argument, or any specific counter-argument at all (like, 'your plan is doomed from the start because of X', or 'your plan will still work because of Y'). It doesn't even look to me like you tried addressing the compromise limited offroading I suggested to give the player some of this flavor joy-of-rulebreaking. You're just waving your hand and saying 'Meh.'

In case you still don't give any real weight to my opinion on AI tactical planning, l guess I'll try addressing the entire idea of (currently) facilitating the bypassing of blockades (which I understood to be the primary motivator for the current proposal, along with a minor sprinkling of joy-of-rulebreaking). Some of this I'll acknowledge as clearly 'opinion' rather then all being 'reasoning', but I think there is a fair bit of reasoning here and to the extent you disagree with any points I think it deserves a more specific response than simply 'unconvincing'.

A1) Geoff, as you recently told someone, multiplayer is not a significant focus of the game currently. Tactical possibilities that might be fun-and-challenging in a multiplayer context are not necessarily good in an AI-opponent scenario. I think you might be unconsciously blurring these scenarios here, with your focus on wanting to bypass blockades. Are you really thinking the game will be more fun if a player can more easily bypass an AI blockade, or vice-versa? BTW, if an AI does currently manage to blockade a choke-point, it is purely a matter of luck, not planning (more below).

A2) It is my view that the AI's current strengths are in non-military development and expansion, and perhaps in assessment of military strength of two respective fleets (at least, I've put a lot of time into that, but I'm not entirely convinced it's good enough to be a real 'strength' yet). AI current actual military planning yields mediocre tactics. If the AI wins, it's more because of out-production and luck rather than military tactics. It's only the fallout of randomly colliding priorities that sometimes gives the appearance the AI is being wily

A3) More specifically: AI current military planning/tactics are entirely dependent on assessment of local and near-local threats based on starlane connectivity. The AI currently makes no attempt to do enough graph analysis to identify and choke-points; any appearance of such is currently arising fortuitously from assessment of local threat.

A4) Even a limited ability to skip around a blockade, unless limited to unarmed ships as I proposed above but don't seem to have received a response about, can pretty much break the entire blockade since if an armed ship enters the blockaded system from a 'hind' side of the blockade, it allows all the blockaded ships to pass through. It would seem to put blockades back to the much more minor pass-through-after-a-delay-and-battle role they previously had, which we discussed greatly and put a fair bit of effort into strengthening.

A5) The great preponderance of players have requested that the AI take greater, not lesser advantage of choke-point blockades; in fact I am unaware of any player making the latter request. I also fail to see players clamoring for greater joy-of-rulebreaking. Thus, facilitating the bypassing of blockades appears, at the least, excessively premature to me.

[Geoff the Medio]

You appear to be saying that because my plan has some admitted uncertainties you find my opinion here vaguely unconvincing...

I was unconvinced by the above-quoted reasons Aquitaine posted, since seemingly nobody actively coding the AI was involved in them, and they happened ~10 years ago before current AI existed.

...a response convincing enough to not be idly dismissed...
In case you still don't give any real weight to my opinion on AI tactical planning...

I specifically asked for your opinion; I give your opinion, as the main AI coder, about AI coding quite a bit of weight.

Continuing to brainstorm related ideas doesn't mean your post or suggestions were ignored or dismissed. Rather, after your initial response, the discussion quickly moved to ideas for much more limited forms of off-roading, rather than the initial suggestion / discussion about more widespread / easy lane-avoiding.

Or at least that's how I interpreted the thread...

Are you really thinking the game will be more fun if a player can more easily bypass an AI blockade, or vice-versa?

I think there could be various fun scenarios where a player has to react to an opponent (AI or otherwise) establishing a large fleet in a system that blocks of access to a large number of systems, where the player can do something other than making their own big fleet to break through. Having just a single choke point isn't necessarily the most interesting tactical situation. And I have thought about having alternative means of moving around the galaxy being somewhat like ground vs. naval vs. air forces, with their different movement restrictions and plenty of interesting strategic and tactical situations that result.

Also, I'm not thinking of implementing all this immediately... I just saw a related comment in the thread and wanted to discuss the possibilities / issues. Brainstorming, essentially... And hoping for some informed clarification about the AI issues.

[mileser]

I know I'm fairly new here and I'm not doing any programming, just a very interested user. But, I personally think adding offroading is not a good idea. I don't think it adds enough to the game, makes AI programming more difficult and thus consumes more processing power, takes programming time from potentially more interesting game features.

IF offroading were to be allowed, it should be somewhat of an extreme measure. My idea would: Very expensive special ship hull with NO external slots and requiring a special building (with its own requisite tech)+Neutronium to build. Further, speed should be slow (maybe even random speeds between 10 and 20/turn) with a slight possibility of a disaster that destroys the ship.

OK, I guess I've made it pretty clear I don't like the idea of offroading.