List of unsolved mysteries(?)

[VincentASM]

Mainly for me to keep track of what needs checking.

However, I'll leave it here in case some other people are interested. Feel free to add to the list as well.

To be tested

• FE6: Ranking password in the Epilogue, how to view it?
  
• FE9: Lethality activation rate, Skill or Critical-based (evidence points to Critical)?
  
• FE9: Iron Wall's effect and activation rate. (thread)
  
• FE10: Hidden treasure Coins not found in the event data.
  
• FE10: Does Calill receive the Data Transfer bonus from Largo (the FEW site said so)? - Reopened
  
• FE10: Experience gain formula.
  
• FE10: Soren Epilogue criteria (do Soren and Pelleas need to be alive?).
  
• FE DS: Experience gain for deviant classes (eg. Priests, Ballisticians).
  
• FE3: Hidden item-finding rate.
  
• FE DS: The default support bonus with Marth- does it skip the A Rank support? (thread)
  

Still testing, but not much progress

• None
  

Given up for now

• FE6: Detailed Ranking data.
  
• FE7: Criteria to get more than 7 Tactician stars (9)? (thread) - Solved, see below
  
• FE DS: Dynamic growth rates- odd behaviour when reclassing. (thread) - Solved, see below
  

Edited February 1, 2010 by VincentASM

[CT075]

Callill does recieve the transfer boost from Largo. I remember proving it...

[VincentASM]

Thanks for mentioning it. I kind of neglected to update this page >___<

[Jaybee]

RD Exp calculations

HM: (Enemy level - Player unit level) / 2 (round it) + 5/20 (5 for weaken, 20 for kill)

NM: (Enemy level - Player unit level) / 2 (round it) + 10/25 (10 for weaken, 25 for kill)

EM: Unknown

I wonder how you found out the FE8 exp formula. That thing is sick.

[VincentASM]

I was actually hoping to figure out the complete FE10 exp formula, factoring in bosses, thieves, Laguz and whatnot. However, that's a good start I suppose. Although I *might* have figured out the basics a long time ago.

The FE8 exp formula is only slightly modified from FE6 and FE7's. Some Japanese fans figured out those and I simply just tested if they worked or not.

Edited June 22, 2009 by VincentASM

[Jaybee]

Interceptor figured out the total exp gain for HM mode, and add 5 for NM.

Fans figured out the entire exp gain? How many nights of sleep did they lose? (Referring to FE6, 7 and 8. FE10 is way simpler except EM)

Edited June 22, 2009 by Joshybear25

[CT075]

I believe that the EXP bonus for easy mode on FE10 is liek x2, if I'm not mistaken.

[Robert]

FEDS- Roger was given on level as an armor. He leveled +1 Hp, +1 Spd. Next, I reclassed him to Pirate. He leveled +1 Hp, +1 Luck, +1 Def.

Roger has no defense growth as Pirate, but a 30% as knight, so if dynamic growths are true, he would have had a 3% chance for def.

[VincentASM]

I know Roger is supposed to have 3% growths once you do that.

However, I don't know if it's exactly 3% (I do know it's at least 1%) and also what happens after multiple Level Ups as a Pirate. It's likely that after this Roger won't gain Defence again.

[Robert]

He didn't, and it the level up wasn't planned, it happened as I was playing the game for fun. It was unlikely that it would be less than 3 because it was the first level as a pirate.

[Nitrodon]

I should probably post what I know about these.

• Item-finding rate in deserts (for games with unknown finding rates).
  

FE6 and FE7: A straight 1/11 chance for non-thieves. This does not use the same RNG we all know and love, but a different RNG which handles things such as shaking the screen during a critical hit or making characters' mouths move when they talk. If abusing savestates, this RNG can be advanced by repeatedly opening the Item menu.

FE8: I think it's (luck+1)/101. This one uses the usual RNG.

• FE7: Criteria to get more than 7 Tactician stars (9)? (thread)
  

Tactician points are added at the end of each chapter (including Final-1 and Final-2) based on your rank as of the end of that chapter. You get 0 points for an E rank, 1 for a D rank, all the way up to 5 for an S rank. These points are the same as the number that appears to the right of the tactician's name in Battle History. Every 12 points gives you one tactician star.

If you wanted more detail, here are the ranking criteria:

Tactics rank: 40/80/120/160/200 points

Survival rank: 15/35/55/75/100 points (thresholds: infinity/5/3/1/0)

Funds rank: 0/20/40/60/80 points (thresholds: 0/20/40/60/80%)

Exp rank: 0/20/40/60/80 points

Combat rank: 10/30/50/70/90 points (thresholds: 0/15/25/35/40%)

Thresholds for D/C/B/A/S rank: 100/200/300/450/550

[VincentASM]

Thanks for the info ^^

I figured something was up with FE7, since manipulating the "regular" RNG seemed to have no effect in finding desert items.

Which reminds me, are you aware of the Devil Axe backfire rate formula for any of the games? It's approximately (Luck - 31)%, but I have noticed it deviates a little from the approximation in practice.

I'm a bit puzzled as to how the Tactician points work though. I understand where some of the numbers come from, but I can't seem to get my head around it.

Is it possible if you could show me a brief example to illustrate how the points work? For example, in one or two chapters?

[Nitrodon]

The big list of numbers is just how to calculate your overall rank based on the number of stars in each of the individual ranks. This overall rank exists even in the middle of the game, and is used to determine your tactician points (and thus tactician stars). I'm not good at examples, but if your overall rank is B as of the end of each chapter, you will gain 3 points per chapter, and thus gain a tactician star every 4 chapters. In a ranked playthrough, you'll end up with more stars than that.

Also, I should mention that the zero requirements (for tactics, exp, and funds) are indeed a programming error. There are separate numbers for Eliwood's and Hector's routes, but the programmers forgot to check which route you're on when determining rank, and so the game always uses the numbers for Eliwood's route. In Hector-only chapters, these numbers are zero.

As for the Devil Axe, it's precisely (31-luck)% in FE7. I can check the other games with such a weapon (except Shadow Dragon) later.

[VincentASM]

The big list of numbers is just how to calculate your overall rank based on the number of stars in each of the individual ranks. This overall rank exists even in the middle of the game, and is used to determine your tactician points (and thus tactician stars). I'm not good at examples, but if your overall rank is B as of the end of each chapter, you will gain 3 points per chapter, and thus gain a tactician star every 4 chapters. In a ranked playthrough, you'll end up with more stars than that.

Ah, thanks. I see what you mean with the overall rank. I had a suspicion that the type of playthrough might affect the number of tactician stars per chapter (since I kept getting one per 4 chapters), but for some reason I didn't imagine it had to do with rankings.

For the list of numbers, is that per chapter or does it depend on the number of chapters that you do? In other words, do you need 550 points to get a S Rank in one chapter or is that for the whole game?

Also, I should mention that the zero requirements (for tactics, exp, and funds) are indeed a programming error. There are separate numbers for Eliwood's and Hector's routes, but the programmers forgot to check which route you're on when determining rank, and so the game always uses the numbers for Eliwood's route. In Hector-only chapters, these numbers are zero.

I thought that might be the case, since I saw the apparent ranking data for Hector's routes. One thing I always wanted to do was make a patch that incorporated the intended Hector mode rankings, which might interest people who play for rankings. My hacking skills are relatively basic though : P

As for the Devil Axe, it's precisely (31-luck)% in FE7. I can check the other games with such a weapon (except Shadow Dragon) later.

Okay, thanks for that. I'm particularly interested in FE6's calculation, since even with 127 Luck, you can't manipulate the regular RNG to stop the backfire.

[Nitrodon]

For the list of numbers, is that per chapter or does it depend on the number of chapters that you do? In other words, do you need 550 points to get a S Rank in one chapter or is that for the whole game?

The list of numbers shows how to determine your overall rank. Your current overall rank is based on the current individual ranks based on that list of numbers. Your current individual ranks are in turn based on your entire playthrough up to this point.

I realize it may have been stupid of me to use the word "points" to refer to two different things in the same post, in case that confused you at all.

Also, I should mention that the zero requirements (for tactics, exp, and funds) are indeed a programming error. There are separate numbers for Eliwood's and Hector's routes, but the programmers forgot to check which route you're on when determining rank, and so the game always uses the numbers for Eliwood's route. In Hector-only chapters, these numbers are zero.

I thought that might be the case, since I saw the apparent ranking data for Hector's routes. One thing I always wanted to do was make a patch that incorporated the intended Hector mode rankings, which might interest people who play for rankings. My hacking skills are relatively basic though : P

That was my logic too, combined with the fact that chapters in only one route had zeros for all the numbers which seemed to correspond to the other route.

I'm particularly interested in FE6's calculation, since even with 127 Luck, you can't manipulate the regular RNG to stop the backfire.

Perhaps you should try this with precisely 31 luck, and then again with precisely 32 luck. There does not seem to be any underflow protection in FE7, so the same might be true here. I don't know of any way to get over 31 luck legitimately with a cursed weapon equipped, so I doubt this matters in practice.

[VincentASM]

The list of numbers shows how to determine your overall rank. Your current overall rank is based on the current individual ranks based on that list of numbers. Your current individual ranks are in turn based on your entire playthrough up to this point.

I realize it may have been stupid of me to use the word "points" to refer to two different things in the same post, in case that confused you at all.

Ah, I think I see now. I was indeed somehow directly connecting them with Tactician points for some reason >___<

Now that I think I understand them, is it alright if I use your Tactician points data on my site (with credits of course)?

On a related note, the prototype FE7 ROMs seem to display the Tactician points in one of the debug menus. I was actually hoping to make some kind of breakthrough using that knowledge.

Perhaps you should try this with precisely 31 luck, and then again with precisely 32 luck. There does not seem to be any underflow protection in FE7, so the same might be true here. I don't know of any way to get over 31 luck legitimately with a cursed weapon equipped, so I doubt this matters in practice.

For FE6's Devil Axe, I tried both 31 and 32 Luck and I still experience backfire. I haven't tested it for FE7 yet, as Hector's been through over 100 battles with just 30 Luck and still hasn't been backfired. However, I did try the RN = 0 code for FE7 and found no backfire for 31 Luck or over (it was a lot time ago that I tested it though).

While I'm here, is it possible if you could show me how you calculate the standard deviation for enemy stats? I think I've managed to calculate it for one run of Level Ups, but I'm stuck on doing it for three (for the displayed Level, promotion and HM boosts). I suspect people who want to see enemy stats as averages will probably want the standard deviation too ^^;;;

[Nitrodon]

Devil Axe backfire rate in FE6: (21-level)% (minimum=1 for obvious reasons)

There is no underflow prevention, so a hacked level 22+ character will always suffer backfire.

As for standard deviation, if you can calculate the variances of each of the three runs of levelup, and the stat gains are independent (which will happen if it's impossible to cap), then the variance of the total is just the sum of the variances. Take the square root of that as before to get the standard deviation.

[Vykan12]

Raising Largo alone for transfers doesn't give Calill any in FE10. I'm supposing it's a requirement for Calill to reach lv 20 in POR as well for Largo's transfers to take effect.

[Nitrodon]

While I can't do any assembly hacking for FE11, I can still watch and modify RAM. From this, I have some empirical findings about dynamic growths.

The game keeps a running total of each character's average stats, as well as the amounts actually gained. The 1/10 estimate seems to be empirically accurate, as 1000 gives one extra point in each stat. Based on this, I have one adjustment to make to the current understanding of this phenomenon. Namely, gaining a stat reduces the dynamic growth by (100-growth)/10 instead of growth/10. To illustrate, in the example from the dynamic growth page, Marth would have -2, -5, 0, +4, +5, -3, +2, +0.2 instead.

Other notes:

1) The maximum you can gain in any stat in one level is 2.

2) If a stat is capped before you gain a level, its dynamic growth is completely unaffected by that level.

[VincentASM]

Thanks ^^

I knew something was off with the reductions, but I couldn't imagine what it could be. That also explains there is a fall off when attempting to gain stats in a class with 0% natural growth.

I'm a little surprised you can't get a +3 during a Level Up, but then again I don't recall any character having more than 200% total growth without hacking.

Can you tell me a little bit more about how average stats affects things? I'm not sure if it's related, but I've noticed my (yeah, it's personal experience) Marth's Level 30 stats seem to match his simple!averages quite well.

[Nitrodon]

The running total of average stats is how dynamic growths are implemented in the game. The dynamic growth rate is not stored by itself, but instead calculated from the average stats and the total stats gained (which is also stored and used to calculate current stats).

[VincentASM]

Ah, okay, I see.

So if one of the stats is currently below the average, would the effective growth rate always be higher until the two evened out?

Also, if it's not too bothersome, could you describe a short example using numbers? I always seem to find it hard to get to grips with new concepts ^^;;;