Action Bar / Speed System

[Wolkendrache]

Alright, here’s what I think how the action bar works in Suikoden Tactics…

While playing the game you might have probably realized that a character with double the speed of another character will normally not have made double the actions of the slower character at any point. It’s more like by the time the fast character made 6 moves, the character with half the Spd made 4 or even 5 moves. There seems to be a balancing factor to avoid fast characters getting overpowered.

The game seems to balance it by adding a certain number to everyone’s Spd stat. The higher the number, the smaller the relative differences of the Spd stats. The action bar uses modified Spd stats (mod_Spd), it doesn’t use the plain Spd stats. The formula is: Spd + X = mod_Spd. Now what is X?

I’ve made some tests. I entered a battle with 6 characters and they were all moving once (without an action), and then always wait, so they had the exact same speed delay (once from moving). I wrote down the order of executions, then I looked up when exactly one character overlapped another one for every combination.

Example: Flare had Spd 81, Tal had Spd 59. After both had 7 turns, Flare overlapped Tal. What you have to do is (81+X) x 7) / 59+X. The value must be slightly below 1. Then you do (81+X) x 8 ) / 59+X. The value must be slightly over 1. This means that after both made 7 turns, Flare will move twice (thus overlap Tal). The order is …Flare 7 – Tal 7 – Flare 8 – Flare 9 – Tal 8. Now you need to find an X value that meets the condition, but not only for Flare and Tal, but for all the character combinations. In this first test X was 112 or 113. In a second test X was 110 or 111.

It is clear that X must be a developing stat. If it was that high at the beginning of the game then a character with 20 Spd would need ages to overlap a character with 10 Spd, although s/he’s “double as fast”. At first I thought X must be based on levels. One test showed that it can’t be the enemy levels, because X was the same against Lv22 enemies and against Lv38 enemies. In the above two tests, when X was 112 or 113, my characters were at Lv38; when X was 110 or 111, my characters were at Lv33. Both X were too close, so I don’t think X is based on character levels either. The only thing I can think of is the number of recruits. 112 divided by 2 is 56; excluding the mercenaries, this was exactly the number of recruits in that test. In the other test I didn’t have Lalacle, so I had 55 recruits (again excluding the mercenaries), and 55 multiplied by 2 equals 110.

I’m pretty sure the action bar calculates with modified Spd stats, and in my tests I’m pretty sure I got the exact modified stats. However, I’m a little less sure about the “recruits x 2” theory, but I can’t think of a better theory. It worked in my tests, but unfortunately I have no early game save with only few recruits to test further.

If you like to do some testing, don’t forget you must do similar actions with the characters you’re comparing, because different actions have different speed delays, and speed delays are very heavy. I can’t tell you much about it. I know that a higher level spell causes a bigger delay than a weaker level spell. I know that moving (doesn’t matter how far) causes a speed delay. I guess waiting (without moving) causes no delay. I think using an item causes no delay, but if it does, then I know that this delay is smaller than the one from moving. That’s all I know about speed delay.

[Wolkendrache]

What does this mean?

If you have fewer recruits, Spd would have more effect: The higher Spd, the more badass that character gets; even slower characters get more effective as long as they are faster than the enemy; characters slower than the enemy become even worse. In the first example, where all characters are recruited, Flare (Spd 81) overlaps Tal (Spd 59) after 7 turns. If you had recruited only the story characters plus Tal (impossible, but anyway), then Flare would overlap Tal already after 5 turns, although Tal isn’t too slow of a character himself. In short, since most of the characters are mostly faster than the enemy: the fewer recruits, the more effective your characters.

Btw, I’ve made another test with Lv20 characters. I already had 47 recruits, but no mercenaries. It seems the formula doesn’t exclude the mercenaries, instead it is always deducting 4 from your total recruits, so X = (recruits - 4) x 2. This formula works for all tests I've made so far, but I'm still not 100% sure. Some more tests wouldn't hurt.

[Antimatzist]

Which battle do you use for your tests? Did you also look on enemies and what they do?

Is the initial order of turns influenced by speed?

[Wolkendrache]

I mostly used Obel 1 for testing, but also Collanbal. In Obel 1 I was lucky to always get the version without flying lizards (although I don't think the sleep status changes anything if you're only waiting anyway. Getting attacked or hit doesn't inflict a speed delay.

In these first tests I was trying to get the basic formula for the action bar. This formula applies to everyone including enemies, why should enemies have a different formula? For the basic formula there's no need to look what the enemy is doing. I haven't made tests on speed delay of specific actions yet, but I think even speed delay works the same for enemies.

Since X is the same for all members of a battle, you can just watch the plain Spd stats of everyone. So yes, these determine the initial order. If you restart a battle, the order might slightly change, but that's just because the enemies have a slight randomization in (all of) their stats. Your characters' stats are fixed, so the relation between them will not change regarding the initial order.

[Antimatzist]

Using enemy data allows you to look at a larger batch of data at the same time, that's why I'm asking. I also wonder if enemies get the same or a different "x" than you. That'd be interesting to know because the same boni and mali for your party would be the same for your enemy.

I wonder about that #RECRUITS - 4. Because if that's true, your very first battles would result in a malus. An "easy" way to find out would be fighting the first Middleport battlle, once with recruiting Rita and once without. In the first case, your X would be -2, in the second case 0. It's hard to figure out, as the speed stats of all characters are nearly identical so early in the game, but maybe one can spot a difference.

[Wolkendrache]

The problem is your characters and the enemies must do the same actions (actions with the same speed delay), but since you can’t control the enemy, it's hard to use all enemies for testing. X should be the same for the enemy. If you want to test this, just go to Obel 1 (the version without lizards). Then check the Spd of one of the skeleton swords. Use a character with the same Spd, but preferably with more Mov. Always run circles away from the skeleton with the same Spd (it doesn’t matter if the other enemies attack or hit you). If the enemy’s X would be different, then that enemy should eventually overlap you, or you should overlap it. I believe this won’t happen, or in other words, I believe X is the same for friends and foes. I don’t know what you mean by boni, I guess there’s only mali (speed delay).

As for “recruits - 4”, for some reason there are lots of NPCs in these first battles (most of them actually unnecessary), maybe they count as “recruits” for that particular battle. But even if they do not count, in the first battles you have at least 4 characters that do count (playable characters), so the first battle where all battle participants would get a malus (-2, if you don’t recruit Rita) is Middleport Underground. But by that point your Spd is around 10, so this malus would be possible and not much of a problem. Moreover it could still be possible that the game sets 0 as the minimum for X (0 means that the Spd stats determine everything). It’s a good idea to test this as you’ve suggested.

I should add that I believe duel battles work differently. I think there’s no X in these kind of battles, only Spd is taken into account (or from another point of view: X = 0). The Rita recruitment is basically a duel battle, but I believe this battle works even differently. Rita joins at Lv2 or so. Kyril could already be at a much higher level by that time, so he should have a much higher Spd than Rita (like 7 vs 3). And yet, you’ll probably not see Kyril overlapping Rita in their duel, so I believe for that duel Rita’s Spd is adjusted to Kyril’s Spd.

[Antimatzist]

You're right about enemies. You could use Elementals as neutral enemies that only move and never do anything else. I think in Terana plains or so, there are even two Elementals, and they have different speed stats (iirc wind elementals have highest speed of all elementals).

With boni, I mean having an advantage in avarge, like mali is a disadvantage against the average.

There are only two fights with three characters at your disposal, the battle at pirate's nest also gives you two additional fighters. It's certainly an interesting idea to test out extremes, best way to verify a hypothesis.

Can't you see Rita's speed in the stat screen during the battle?

There are only four duels in the game, right? Rita, Mitsuba, Martin and the final one.

[Wolkendrache]

With boni, I mean having an advantage in avarge, like mali is a disadvantage against the average.

Ah, now I got it. Imagine X would be different for the enemies. This would mean a Spd 40 enemy would overlap a Spd 40 party character although doing the same action, and this wouldn't make sense at all. Then there would not really be a need for enemies to have a visible Spd stat at all.

Can't you see Rita's speed in the stat screen during the battle?

*facepalm* of course. Unfortunately I don't remember it.

There are only four duels in the game, right? Rita, Mitsuba, Martin and the final one.

I think so. I'm not sure how Cedric is handled. He doesn't seem to be much faster or faster at all, he only has high Mov, but it's difficult to make an estimation because this battle is so short.

[Antimatzist]

Ah, now I got it. Imagine X would be different for the enemies. This would mean a Spd 40 enemy would overlap a Spd 40 party character although doing the same action, and this wouldn't make sense at all. Then there would not really be a need for enemies to have a visible Spd stat at all.

On the other hand, what sense does it make to change the turn roder depending on the amount of recruited characters? ^^

I don't know if you could count Cedric as a duel since you have several characters. I'll pick up the game soon again, I'll keep an eye on it

[Wolkendrache]

On the other hand, what sense does it make to change the turn roder depending on the amount of recruited characters? ^^

Exactly wrong. The amount of recruited characters does not change the turn order, BECAUSE X is the same for characters and enemies. See, if a character has Spd 50 and an enemy has Spd 49, and X is 100, then they have 150 and 149, so the turn order is thereby guaranteed to not change. So you should say "moreover" instead of "on the other hand".

I don't know if you could count Cedric as a duel since you have several characters. I'll pick up the game soon again, I'll keep an eye on it

I will check the early battles until Middleport Underground and Rita's Spd, but I'm not gonna play further for now. It would be cool if you would check the duels and the Cedric battle once you play the game again. You should make a save after every battle and not overwrite them, just for testing purpose. While testing I wished I had done this.

[Antimatzist]

Exactly wrong. The amount of recruited characters does not change the turn order, BECAUSE X is the same for characters and enemies. See, if a character has Spd 50 and an enemy has Spd 49, and X is 100, then they have 150 and 149, so the turn order is thereby guaranteed to not change. So you should say "moreover" instead of "on the other hand".

It does change the overall turnorder, otherwise you wouldn't notice an effect. I'm just saying it's a nonsensical concept, so why not have a similarly nonsensical concept for enemies? YOu could not differentiate betwee a x(party9=100 and x(enemy)=95, that's too close to each other. Or how big is your error margin?

[Wolkendrache]

Error margin = 0 so far. I’m calculating the X margin for every character Spd relation in a battle. The results point to 1 or 2 possible numbers for X. The mod_Spd = Spd + X is anything but nonsensical, but you’ll see once you test yourself.

As for the definition of X, I’m still not sure, it could still be coincidence that X was always (recruits - 4) x 2. It could also be a fixed stat (e.g. 50) + a developing stat (whatever it is based on). Actually I believe there should be a fixed stat to start with, otherwise in the first battles Spd would have too much effect. Anyway, X seems to develop throughout the game: In a battle with Lv20 characters (after Haruna; 47 recruits) X was 86; in a battle with Lv34 characters (before S.P.Facility; 59 recruits) X was 110 or 111; in a battle with Lv38 characters (before S.P.Facility; 60 recruits), X was 112 or 113.

The overall turn order within a battle is primarily changed by actions with speed delay (that's why I said speed delay is heavy), and secondarily by Spd. X is simply a balancing factor to decrease the effect of Spd in a battle. Remember, the action bar works with Spd + X, not just X.

I really don’t get what you find nonsensical about all this.

[Antimatzist]

I do't find your findings nonsensical, but from a gameplay point of view. Why not use only speed? Ok, you want a modifier so double the speed != double the movements. Why make it dependant on the number of your recruits? Especially when the same modifier applies to your enemy as well. It's just a strange way to implement such a feature, and if #recruits is really the determing factor in this, I think it's nonsensical.

This has nothing to do with your findings or your tests, just with the concept itself.

Did you write down the stats of all of your characters? Maybe it's something like Speed of the highest character times x, or average speed of all your characters times y, or something like that, and coincidentally, it is the same as your #recruits method. I'm not saying it's wrong, though, and I still wonder what accounts as a turn.

[Wolkendrache]

First, about “turn”: In an action bar system, “turn” is equal to “action” or “point of execution”. It is always relative. The opposite of an action bar system is a so-called turn-based system, that’s why using “turn” in an action bar system is a little confusing. So let’s either use an alternative like “action” or “p.o.e.”, or use “turn” while keeping in mind that it is Spd-based.

The modifier is something that decreases the Spd gaps between characters. Spd x 2 would increase the gaps, so this is impossible. It must definitely be Spd + X, and X must of course be the same for all. In one test, the Spd stats were: Ameria 75, Nalkul 60, Busk 55, Red M. 45, Jeremy 37, Lino 30. They overlapped each other in a way as if their Spd stats were: Ameria 185, Nalkul 170, Busk 165, Red M. 155, Jeremy 147, Lino 140. I’ve made several tests like this, and they lead me to the assumption that the modified speed is Spd + X, and in the above test X is 110.

Ok, you know what, here’s the data, maybe you find a different interpretation…

Backstreets of Razril
Kyril Lv3, average Lv3, Kyril natural Spd 4, highest natural Spd: 4 (both Kyril and Andarc), 3 recruits
Spd stats: Kyril 14 (due to equipment), lightning elemental 11, fire elemental 8, Andarc 4
Modified Spd (according to action bar development): Kyril 84-86, lightning elemental 81-83, fire elemental 78-80, Andarc 74-76
-> mod Spd = Spd + 70-72
-> these results also prove that X is the same for characters and enemies.

Middleport Underground
Kyril Lv6, average Lv3.4, Kyril natural Spd 7, highest natural Spd: 9 (Andarc), 5 recruits
Spd stats: Andarc 14, Kyril 10, Seneca 6, Lalacle 2
Modified Spd: Andarc 80-84, Kyril 76-80, Seneca 72-76, Lalacle 68-72
-> Mod Spd = Spd + 66-70
-> X is smaller than in the previous battle

Obel 1 (before Small Border Village)
Kyril Lv20, average Lv18, Kyril natural Spd 27, highest natural Spd: 36 (Mizuki), 47 recruits
Spd stats: Akaghi 36, Nalkul 33, Kyril 30, Hervey 26, Reinbach 23
Modified Spd: Akaghi 122-123, Nalkul 119-120, Kyril 116-117, Hervey 112-113, Reinbach 109-110
-> Mod Spd = Spd + 86-87

Obel 1 (before S.P.Facility)
Kyril Lv33, average Lv26, Kyril natural Spd 35, highest natural Spd: 64 (Noah), 59 recruits
Spd stats: Ameria 75, Nalkul 60, Busk 55, Red M. 45, Jeremy 37, Lino 30
Modified Spd: Ameria 185-190, Nalkul 170-185, Busk 165-170, Red M. 155-160, Jeremy 147-152, Lino 140-145
-> Mod Spd = Spd + 110-115

Obel 1 (before S.P.Facility)
Kyril Lv37, average Lv38, Kyril natural Spd 41, highest natural Spd: 72 (Wendel), 60 recruits
Spd stats: Flare 81, Akaghi 72, Tal 59, Red M. 50, Lino 35, Axel 28
Modified Spd: Flare 193-194, Akaghi 184-185, Tal 171-172, Red M. 162-163, Lino 147-148, Axel 140-141
-> Mod Spd = Spd + 112-113

Analysis:
- The action bar modifies natural Spd in a way so that the intervals of the natural Spd is kept, but the relative differences are decreased (e.g. in the above example: Flare is still 9 points faster than Akaghi, but Akaghi has 95% of Flare’s Spd, not 89% as per natural Spd).
- Mod Spd = Spd + (recruits – 4) x 2 is wrong, it was indeed coincidence (see tests)
- X can’t be based on enemy levels or enemy stats: with the Lv38 party I went to Collanbal and fought about Lv24 enemies (Obel 1 enemies were Lv38), and the results were identical.
- X can’t be based on Kyril’s level, character with highest level, or party average level (see tests)
- X can’t be based on some average party stat (see last two tests)
- even if in the second test X is higher than in the first one, there’s still a tendency that X is increasing throughout the game, with some fluctuation though.

I really wonder what X is based on…

[Wolkendrache]

One more thing: Rita’s Game

This “battle” can’t be used to calculate X, because it ends too early or you can’t overlap Rita. To make calculations, a character must overlap another one, and these two characters must do the same actions. If you do the same as Rita (moving and changing terrain), you’ll not overlap her. So just by the way…

If you wait 3 rounds, you’ll overlap her, but only in a NG+ with Godspeed Anklet, Merchant’s Cap and Imperial Ring (Kyril’s Spd is around 14, Rita’s Spd around 4). The battle will end either once Kyril has done 5 actions or once Rita has done 5 actions, it depends. If Kyril has made 5 actions but fire tiles are still fewer than lightning tiles, the battle will end prematurely, even if Rita has made only 4 or even only 3 moves. But if there are not too few fire tiles, then Kyril can even make 6 moves (he must overlap her for this). Then the battle ends after Rita has made her 5th move. Unlike Kyril, Rita seems to always be allowed to make her last move, no matter if she has too few lightning tiles to win through her last move.

And btw, using an item seems to have a very small speed delay (smaller than moving). Extra Move seems to have no extra speed delay.