Sorry for grammar and short description, but English is not my native language.
We were playing using poker chips, with only two persons (Me as the GM and Filip as a player).
So, he created ex-member of a Kanissian special unit, who escaped from army and hid in Ghostfield. He was created by the army in a medical experiment and he lost his memory.
The main problem during character creation was the lack on any helpful structures or guidelines.  There was no list of seeds or backgrounds/links. There was only a vacuum, so we created his character based on Final Fantasy 7. So, here is my point:
Make a big list of everything: characters seeds, croup seeds, traits, links and backgrounds. Maybe even a random generator if you want. Also, there is a need of brief description of many existing characters from Final Fantasy and some anime.

For example:
Cloud Strife is a ex-member of SOLDIER. However currently he works for an eco-terrorist group called AVALANCHE.  First, he was only fighting for money, but after some events, he started to fighting Shinra corporation for his own reasons. He has big sword, gaps in memory, and a cool motorbike.

We had only three scenes, one character scene and 2 conflicts scene.
The Story Seed was the invasion of an entire Kanissian expedition corpus
The characters seed was might opponent: Entire Kanissian nation.
First, he was stuck in a refugees camp, where everyone were trying to escape before the arrival of the expedition corpus (consisting of three airships and many flying scouts similar to Vanships from Last Exile). Some mother asked Avshum to help her find her daughter.

Second Scene: A group of Kessen skirmishers has landed near Fillip’s character, and attacked
Third Scene: A Battle with one of the Airship (I used stats of a steamboat).

OK, so we found many bugs:
1: A Power from the Ghostfield setting (spontaneous gating) has wrong cost: how am I supposed to know how many kilometers is character from core of the ghostfield when the system doesn’t use this kind of metric?
2:  Gift Dice from the GM are not so good,
3: Passions are not balanced. Glory is the best, because when you fight with a swarm type of opponent, you have a bucket of charge dice.
4: Wounding is too complicated: I need to overcome defense+1 to deal one wound, 2x(defense+1) to make two wounds and so on.
5: What happenes with an unaccomplished goal after the conflict? If it is something like “survival of a young girl”, can I change the difficulty of it in a fallow-up conflict?
6: When a GM ‘s monster or a player characters fails a strike, who and how describe outcome and how?
My suggestion are:
Make a big list of character concepts with lists of traits, marks and so on. Final Fantasy should be your main sources, of course. Create questions for the players that focus their characters concepts.
Design a fanmail system: every player could have his own pool of fanmail which he could award anyone for cool things, stunts and the like. Awarded fanmail are could turn into charge dice. After being spent, it could be transferred to his own pool of fanmail.
Backgrounds and Links: I’m still not sure what purpose backgrounds have in play. For links: big list of example things that can be in this section of character sheet.
After a failed strike, attacked character in his maneuvering could do something like an “active block” in his maneuvering.
Instead of gift dice and the rising group competence, character competence could increase from 1 to n for every wound circle checked. So, the character with two wounds would have 3 dice from his competence. Because of that, the GM could throw any monster against the players without being afraid of making the opposition too strong.

Hi there,

Thanks for the feedback!  Let me address your points first, and then I have a couple of questions out of curiosity.  I'd definitely like to help make this game more enjoyable for you, even though one-on-one play is not really its design principle.  For anyone else reading along, Anime Prime is a Final Fantasy and anime-inspired game focusing on action-packed conflicts and PC-to-PC roleplaying interaction, it's Creative Commons licensed, and available completely free at AnimePrimeRPG.com.  Currently in public beta testing :)

Gate Sense

I know what you mean about it being hard to apply.  I figured it would be obvious from the in-game circumstances, but you can always substitute a roll to figure it out.  I'm thinking I might replace it with an even more vague "in the heart of/somewhere inside/at the edge" thing, which would be more apparent from the fictional circumstances.

Gift Dice from the GM

What about them don't you like?  Part of the issue when playing one-on-one, for which AP wasn't designed (it's very much based on intra-group dynamics), is that GM dice are the only way of raising Group Competence.  I suggest using John Harper's alternative skill system instead, which alleviates the issue.  Here's how it works.  There's no Group Competence; instead each PC gets four skills.  Each skill has four checkboxes rated 1, 2, 3, and 4. When you want to use a skill, you check off a box, and add dice to your action roll equal to the rating. You can gain a maximum of four dice from skills at once (so you can use one level 4, or two 2s, or a 3 and a 1). After a character scene, you can choose to refresh all your skill boxes as a scene benefit.  I'll add this to the optional mechanics section soon.

Passions

I think you may have misread Glory and swarms.  You only get a charge die when you completely defeat the swarm, that is, you do damage to its last damage box and kick it out of the conflict.  It doesn't matter that you're taking out parts of the swarm left and right, the charge die is only earned for defeating the whole thing.  And again, balancing is a group rather than one-on-one issue.  Check out Compassion with 5 PCs, where you get charge dice every time your friends get hurt.  Compassion does nothing in a one-on-one game.

Wounding

It seems like you understand how it works after just playing through two conflicts.  I'm willing to accept that kind of learning curve :)  It really becomes less of an issue with practice.  But I appreciate you bringing it up, I might just have to explain it better than the text currently does.

Unresolved Goals

Unresolved goals at the end of a conflict are, well, unresolved.  Apparently no one cared enough to spend actions and dice on them, so most of them can be abandoned.  If characters still care, the rules say that you can actually fulfill a remaining goal as a follow-up character scene benefit.  Unless that would lead to a new conflict :)  So yeah, you could have the same goal in a new conflict, and make it harder or easier, depending on circumstances.

Describing Failed Strikes

The authority remains as usual: the player describes what happens with the PC, with some added input if the enemies are squads and even more if they're swarms, the GM describes the rest.  It doesn't need to be total failure; it could be that the PC succeeds, but it happens to have a different effect than expected, for example.  (Think of striking a hydra and cutting off a head, but it just grows right back!)

Fanmail

That's what gift dice for maneuvers from other players are for.  I don't see the benefit of adding another pool; by taking it out of their Action Pool, they express that they are sacrificing some of their own resources to support a team mate's cool maneuver.  I want all such choices to cost something (see also Achievements).

Background and Links

This is explained in the essay on Flags.  In short, background is for items that matter to how you roleplay your character but that you don't want used for plot purposes.  They are static.  Links are plot seeds for the GM, so they will be dynamic.

More Examples and Guidelines

I hear ya.  I'll put more in for the final version.

My Questions

1. Tell me more about your scenes!  How did the fight with the airship go?  Did you feel like you could do cool, Final Fantasy Cutscene maneuvers with the system?
2. Did you use player-suggested goals at all?  They are the heart of the system, and the more you play with them, the more you'll get out of it.

Thanks again!

My main conclusions after the two sessions I played in (note that I'm not all that familiar with the book, so I might be missing something):

1. I didn't like the Gift Dice from the GM. Basically, it felt like stunt awards in Exalted, only here, they were much more important for succeeding in a conflict, much more like in Wushu. The problem is, on the one hand, the GM was providing mechanical adversity for me - but on the other, I've been at his mercy when it came to my effectiveness in dealing with that adversity. It felt like a game of "please the GM" to me.

As much as I like the way Beast Hunters work, this system doesn't have it. At least, not in one on one play.

The Group Competency rule was the thing that actually made me interested in the game - but we didn't have an opportunity to see how it works in those one on one playtests. However, when we discussed it I came to a conclusion that most probably, in practice Group Competency would just increase by one every round, as the players would have no reason not to give each other the dice. It doesn't feel like a very elegant solution to me - but again, I didn't see it in actual play.

Now, the skill system you mention sounds neat.

Other than that, I had some thoughts on other possible alternatives for current Competency rules:

* Characters could have individual Competency based on their current condition - i.e. the more wounds you have, the more dice you roll. This would create a somewhat (though not perfectly) self-balancing mechanism, too, as character effectiveness would dynamically adjust to the adversity.

* Another option would be to have a public pool of tokens that could be spent once per round to increase Group Competency by one. Tokens could be added to the pool any time one of the characters gets wounded, and possibly as a potential benefit of a character scene. Basically, a group limit break mechanic. The cost of increasing Group Competency by one could be fixed or based on the current Competency, I'm not sure what would work better. Anyway, this could result in an effectively self-balancing system.

Note that, as you might recall from my posts on SG, I have some issues with the fact that the adversity is not controlled in this game, be it through a GM's budget or some other mechanism. These ideas are the result of my attempts to find some easy solution for that.

* Group Competency could increase along with Charge Dice or Awesome Token gains. Probably not on a one per one basis, though.

* There could be a list of a few triggers that increase Group Competency. Something along the lines of " increase GC by one the first time someone makes a dramatic reference to one of the previous character scenes; increase GC by one the first time someone goes over nine thousand; increase GC by one the first time the characters cooperate in a conflict; increase GC by one the first time a character uses his Charge Power; increase, GC by one the first time a character gets KO-ed etc." or whatever.

2. The wounding formula is needlessly confusing, indeed.

3. The variable cost of that funky gates power, frankly, sucks. Why not simply have a fixed cost, independent of the character's location?

4. The whole goals system felt awfully loose to me. I'm afraid we didn't play long enough to fully anaylse it, but something feels kind of clunky here.

5. I don't really like how Passions work. It's another thing that gives me a clunky impression. They're kind of like Keys in TSoY in some respects, but you can have only one at a time here, and they're quite varied in terms of possible gains. What if, instead of fixed Charge Dice rewards, hitting some more general triggers let the players apply for a Charge Dice award? Or, maybe draw Charge Dice from some public pool upon hitting a Passion, where they'd be returned after spending? If I were to run this game, I'd probably houserule it kind of like that, tying Charge Dice to that group limit break idea above.

But well, I'm not a fan of loose economies in general.

Now, as for your questions:

1. Tell me more about your scenes!  How did the fight with the airship go?  Did you feel like you could do cool, Final Fantasy Cutscene maneuvers with the system?

2. Did you use player-suggested goals at all?  They are the heart of the system, and the more you play with them, the more you'll get out of it.

1. Tell me more about your scenes!  How did the fight with the airship go?  Did you feel like you could do cool, Final Fantasy Cutscene maneuvers with the system?

Also from GM side, I have blank page.

Yes, Filip wrote down some links, but there is no solid structure that helps me create scenes. Yay, I can always push to conflicts, but in book you said it will make a Final Fantasy like game.

So, I have not so bad conflict system, but there is no system that supports things out of battle.

About goals: What there are for? If to make battle more interesting? But it is like begging for mercy from GM "Oh, common, you maked to strong opponent, can I at least make his defense smaller?"
There is no strategic options, because there is no limits for opponents.
Outside battle you have nothing solid, only "backgrounds and links". But how I can effective playing with this?
You created big game book with half of Beast Hunter conflict system and some random essays. Make a new, smaller and more focus game text. Don't write about "Something awesome that you have read on some website", write how to play you game.

Ps. My favorite is rule that suggested Filip is

Characters could have individual Competency based on their current condition - i.e. the more wounds you have, the more dice you roll. This would create a somewhat (though not perfectly) self-balancing mechanism, too, as character effectiveness would dynamically adjust to the adversity.

I'm somewhat ambivalent about the essays and play advice, too. They give me this impression of being, at this stage of development, a somewhat premature. Frankly, lots of the advice strikes me as sort of "a collected wisdom of Story Games" thing. I suspect that at least some of it might be, in the context of this particular game, mostly wishful thinking - and the presentation of some crucial issues might not be very optimal. For example, I wonder whether the discussion of discomfort is really needed, given that it's basically a "Final Fantasy game" - and for similar reasons I have some doubts about the way the game addresses the mood and content part. Wouldn't it be easier if the group just, dunno, picked specific source material as references for all such stuff, or something?

As for goals, I'd like to see some more specific rules for setting their difficulty and determining mechanical effects, if any. Probably a set list of effects with fixed difficulties, or something like that. Otherwise, maybe some sort of "I pick the outcome, you set the difficulty; you pick the outcome, I set the difficulty" dynamic?

Jacek,

This suggestion might well not work very well, exactly because of its simplicity.

Overall, I think I like the new skill system.

Now, if I were to use the skill system and still include some self-balancing mechanism into the game, I'd probably try to work with Passions and how they generate charge dice. Maybe a public pool of charge dice one could draw a number of them from when showcasing the Passion through dialogue and dramatic acts. A few dice could be added to the pool for each wound dealt to a PC. Maybe this could be tied to the goals, too, and a number of charge dice based on the goal's dificulty could be added to the pool whenever one is either introduced or achieved. Possibly, this could somewhat tighten goals rules and tie them to the overall economy of the game stronger.