!! Some NEW tricks !!

[beowuuf]

The code doens't care what you are hitting really - if there is a monster close you gain more XP, and if you hit I beleive you gain more than if you miss although that second part I am going by memory and could be wrong.

So you could disrupt a trolin and probably gain equivalent XP.

Umm, CSBwin++ is still available (even if it is old) that allowed you to see the XP in real time. It would let you see if there is any difference between these two figures (and CSBbuild should be compatible with it - ie you can edit CSBwi++ dungeon with CSBuild stikl, jsut don't try to run any new CSNwin dungeons under CSBwin++)

[Sophia]

Yes, a 'mon' potion still has a variance range, I'm not sure how much depends on caster and how much depends on luck. But there is a power range of 0 - 255 associated with potions.

The formula is (Power)*40 + (Random number 0-15).
So, a LO potion can be from 40 to 65, etc. and a MON potion is from 240-255 in power. Nothing actually depends on the skill-- only whether you can pull the spell off.

There is definitely a variance with magic missiles too - you can see the difference between a newly castable mon fireball and the balls of death an archmaster can throw, even though both are mon level.

The missile formula does figure the skill into the power level, and the missile's power is quite dependent on skill level-- (2*Skill + 4) * (Power + 2). If it goes over 255, it's capped at 255.

In therory I would guess that somehow the 'base' skill is also taken into account somehow, otherwise you would not be able to gain access to fighter feats at the normal level if you hadn't used that skill at all - and yet you can.

This is correct. When you use a subskill, you get xp for the main skill, AND for the subskill. This would seem to be double xp except that when the game checks xp for a subskill, it adds subskill xp + main skill xp and divides by 2. So, the effective xp level of any subskill is effectively the average of that subskill and the main skill's xp.

But yes, the only limit to your experience gain would be the database figure...so if raw experience and not your actual rank is used in any calculations, then experinece would make a difference.

Right, the function to determine your level of mastery always works from the xp totals. Your "level" isn't stored anywhere.

Umm, CSBwin++ is still available (even if it is old) that allowed you to see the XP in real time.

Adding a couple of lines of Lua to xp.lua of DSB would also allow this.

[Octopuss]

Damn, even training with script (=uber fast clicking) it's slow as hell. Is there ANYONE who did archmaster in something the normal way? That's freaking impossible

Soph - so it's totally unimportant how high the hidden priest skills are as long as you can cast the spells? That's good, I;ll just keep spamming fireshields as usual.

Hoping to get an answer on the experience at archmaster level thingie.

[Seth]

Octopuss: What Archmaster are you trying to reach?
Wizard? Priest?

[Octopuss]

Ninja atm, but fighter is my ultimate goal. The whole deal is: is experience still added to hidden skills when you already have archmaster?

[Sophia]

is experience still added to hidden skills when you already have archmaster?

As stated above, yes it is.
I guess eventually you'll get to archmaster at the main skill and archmaster at all the hidden skills, at which point your abilities won't go up any more, but... that'd take quite a while.

[Octopuss]

Oh the above posts were a bit hard to decypher for me Cool then.

But damn, it will take a "while" indeed. Working on 4th master ninja and it's taking a damn long time! I seriously cannot imagine anyone doing this the regular way.

[Seth]

Hehe, nice
BTW i suggest to use Rapier for fighter levels - "JAB" all the way

Are you planning to use this ultimate character in CSB?

[Octopuss]

Hmm not really, not much interest in CSB - but who knows.
Jab? For me it's chopchop

broke 4th master priest yesterday - experience from casting is apparently much faster than fighting..

[Maven]

I have found there is an awful lot of misinformation out there. Probably in part due to so many different versions of the games that do in fact function differently. So much is unreliable I decided to write my own experience monitoring program and check things for myself. I got some interesting results. Here are some things I found.

Version is FTL DM running with DOSBOX under SuSE Linux. Experience tests were done on level 3.

• Wisdom does affect mana regeneration. There are two parts to mana regeneration--how often I receive new mana and how much I receive at a time. The only thing that has affects how much I receive is the max mana stat. If I have less than 40 max mana, I get 1 mana per Mana Tick. 40-79 gives 2 mana per Tick. 80-119 gives 3... How often I receive new mana is affected by my Wisdom and my Priest and Wizard levels. Raising Wisdom with Dain Potions gives me significantly more mana. Up to a maximum. Sophia says the max for Wisdom + Priest level + Wizard level is 112.
• Being "close to" monsters isn't a useful idea. Being "toe to toe" with monsters is. Fighter Club, Thrust, and Swing and Ninja Fight experience is max when fighting toe to toe with monsters. That means if I'm facing them I have to be on the front line. Characters in the rear that give me the "can't reach" message are not toe to toe. I have had some luck being on the back row while the party is facing away, but it isn't totally consistent.
• Melee experience gets some of the bonuses when I'm facing a door or wall. Not quite as much as toe-to toe with actual monsters, though.
• Warcry gives Warrior Parry experience any time it is used. I only get the Priest Influence experience when toe to toe.
• Horn of fear gives very wimpy experience when not toe to toe. Like 4, even while being attacked. Toe to toe gives 44.
• Practicing skills when monster attack timers have run out results in very small amounts of experience. Typically 4-6 for melee, 14-20 for throwing. Spell casting seems to give half experience.
• There are two monster attack timers. One is about 6.5 seconds. Using any skill withing 6.5 seconds after being attacked gives max experience. I don't actually have to get hit.
• I still get some experience bonus until the second timer runs out in about 28 seconds after the attack. It seems to be about half. That might depend on the level of the dungeon, though. I think level 3 has a 2X modifier, so that might account for the half. So for example if I get attacked by a screamer and then kill it, I could still get bonus experience for using the rope to climb down. For the next 6.5 seconds I get 60 experience per use. After that, until the 28 second point, I get 30 experience per use. After that, I only get 14 experience per use.
• When throwing items, it makes absolutely no difference where I throw them, whether I hit a monster or not, whether there is actually a monster anywhere around or not, or whether I'm toe to toe or not. The only thing that makes a difference is how long it has been since I have been attacked by a monster.
• Weight of the thrown item isn't important. It does make a small difference whether I have the item in hand or in the primary weapon char slot. Throwing a boulder gives me 32 Ninja Throw experience. Throwing a rock from the hand gives me 72. Choosing "throw" from the action menu when there is a rock in my weapon slot gives me 84. Daggers and Shurikens also give me 84. Arrows and wooden clubs only give 76.
• When using the bow or sling, I get 80 Ninja Shoot experience.
• Spells seem to give me slight variations in experience depending on which character casts the spell. I'm not sure what causes it.
• The Torch spell gives me, from Lo to Mon, 160 264 404 520 676 804 Fire experience
• Poison Cloud spell gives me, from Lo to Mon, 332 528 744 936 1160 1400 Water experience
• When a spell fails on the Lo or Um level, I get half experience. On the Mon level, about one quarter experience. Don't know about in between.
• The Ya potion is by far the most efficient mana/experience wise. For wizard spells, the Poison Cloud is most efficient.

[Zyx]

Exciting investigation!

[sucinum]

Are you sure about the wisdom part? I did some test a bit ago and found out that the mana ticks are more often the higher your Wis is. Food and water consumption increase accordingly.

[Sophia]

Wisdom has absolutely NO effect on mana regeneration.

Not true, at least for the Atari ST version that CSBwin is based on. However, the actual formula is so bizarre I doubt if you could ever figure it out by just observation. Digging through the CSBwin source code, the answer is found in TenSecondUpdate.

The short version is as follows: using the game clock, it generates a value that is a multiple of 16 between 0 and 144 inclusive. The exact sequence is somewhat... esoteric (that is, the algorithm is crazy), but DSB approximates it well enough by simply going from 16 to 144 by 32's, and then from 0 to 128 by 32's, and then repeating the process, so you could simply imagine it working this way. Anyway, your wisdom, wizard level, and priest level are added together. If this sum is greater than or equal to this generated clock value, you recover 1/40 of your total mana, as you observed, and lose stamina equal to 7 times the amount recovered, or (16 - (wizard level + priest level)) times the amount recovered, whichever is greater. If sleeping, everything is doubled (that is, you recover 1/20 your total mana, etc.), and the sequence of clock numbers progresses faster.

There are two monster attack timers. One is about 6.5 seconds. Using any skill withing 6.5 seconds after being attacked gives max experience. I don't actually have to get hit.
I still get some experience bonus until the second timer runs out in about 28 seconds after the attack. It seems to be about half. That might depend on the level of the dungeon, though. I think level 3 has a 2X modifier, so that might account for the half. So for example if I get attacked by a screamer and then kill it, I could still get bonus experience for using the rope to climb down. For the next 6.5 seconds I get 60 experience per use. After that, until the 28 second point, I get 30 experience per use. After that, I only get 14 experience per use.

The two key intervals are 25 ticks and 150 ticks, so the exact amount of time depends on how fast your game is going and whether or not you're moving around (due to the extra tick), but your values are close. Within 25 ticks, any experience with a subskill (that is, anything greater than skill 3) is doubled. After 150 ticks, any experience gained based on fighter or ninja skills is halved. And yes, each level does have an experience multiplier, which also affects the amount gained.

Spells seem to give me slight variations in experience depending on which character casts the spell. I'm not sure what causes it.

It's just a random value from 0 to 7 that is added to the experience gained.

When a spell fails on the Lo or Um level, I get half experience. On the Mon level, about one quarter experience. Don't know about in between.

The skill level required for a spell to succeed is the sum of its difficulty (stored by spell) and its power level (a number from 1 to 6). You lose half of the experience for each level below the required skill that yours is. So, if you fail by one level, you get half, and so on.

Note that you can still succeed at casting a spell even if your skill is below the required value, because you get a "second chance," when your wisdom is compared to a random number between 0 and 127. If you pass this check, you're able to cast the spell, no matter how difficult it is.

[sucinum]

Anyway, your wisdom, wizard level, and priest level are added together.

Just a little side question so i can estimate the impact of Wis better: is wizard/priest level factored as neophyte = 1, novice = 2 etc. or anything else?

[Sophia]

Just a little side question so i can estimate the impact of Wis better: is wizard/priest level factored as neophyte = 1, novice = 2 etc. or anything else?

That's how it is. So wisdom is much more important, as five extra points of wisdom are like gaining five levels!

[Maven]

Well I discovered last night that I had made a major error in my findings. Popped in this morning to find that this forum is more active than I had thought, and I had not only people pointing out my error, but explaining it in detail. I'm going to edit my original post, so I'll explain it here.

My original findings were:

• Wisdom has absolutely NO effect on mana regeneration. The only thing that has affects mana regen is the max mana stat. And Time. And Sleeping. Mana Ticks are just less than 30 seconds. If I have less than 40 max mana, I get 1 mana per Mana Tick. 40-79 gives 2 mana per Tick. 80-119 gives 3...

I had just noticed that the mana goes up by the same amount every tick. I figured if Wisdom made some difference it would go up more at a time. I totally didn't think of the possibility that the Mana Tick times might vary. Well, last night I noticed that two of my characters that had similar max mana stats were gaining mana at different rates, so I actually timed it. Well, Wisdom DOES make a difference. Higher Wisdom makes your mana ticks happen more often. Sophia went into detail about the algorithm, but suffice it to say that I think it is worth making some Wisdom potions. I haven't fully analyzed how much additional Wisdom the potions give you, or how quickly it wears off, but some seem to be worth it. It looks like at lower Wisdom levels the potions give me a decent boost that diminishes as the Wisdom increases. So there will probably be a diminishing returns point after which it costs more mana than it regenerates.

[beowuuf]

You had me worried since this was my experience in the past.

Welcome to the forums btw!

[sucinum]

Just a little side question so i can estimate the impact of Wis better: is wizard/priest level factored as neophyte = 1, novice = 2 etc. or anything else?

That's how it is. So wisdom is much more important, as five extra points of wisdom are like gaining five levels!

Now I even know why i always favored chars with a high Wis.

Thanks a lot - and awesome bikini.

[Maven]

Wisdom has absolutely NO effect on mana regeneration.

Not true, at least for the Atari ST version that CSBwin is based on. However, the actual formula is so bizarre I doubt if you could ever figure it out by just observation. Digging through the CSBwin source code, the answer is found in TenSecondUpdate.

The short version is as follows: using the game clock, it generates a value that is a multiple of 16 between 0 and 144 inclusive. The exact sequence is somewhat... esoteric (that is, the algorithm is crazy), but DSB approximates it well enough by simply going from 16 to 144 by 32's, and then from 0 to 128 by 32's, and then repeating the process, so you could simply imagine it working this way. Anyway, your wisdom, wizard level, and priest level are added together. If this sum is greater than or equal to this generated clock value, you recover 1/40 of your total mana, as you observed, and lose stamina equal to 7 times the amount recovered, or (16 - (wizard level + priest level)) times the amount recovered, whichever is greater. If sleeping, everything is doubled (that is, you recover 1/20 your total mana, etc.), and the sequence of clock numbers progresses faster.

If this is true, it should be pretty easy to grasp how this works. If my Wisdom + Wizard + Priest >= 144, I should receive mana every tick. For every 16 points less than 144, I should skip receiving mana on 1 tick in ten. Exactly which ticks don't receive mana is probably supposed to look random, but over time it should average about 10% less mana for each 16 wisdom points.

I made a preliminary test for this. Two characters have priest and wizard levels of = Master, which is 11 each. One character has wisdom of 58. The other one I buffed up to 170. Then I ran them both out of mana and watched how much they gained. According to calculations, the lower wisdom character should receive mana at 60% of the rate of the buffed character. I watched mana regen for 35 mana ticks. The buffed character did indeed receive mana on all 35 ticks. The lower one received mana on 27 of the 35. That's 77%, which is a bit higher than the expected 60%, but could be within margin of error for one test I think.

Then I watched the buffed character to see how low my Wisdom would go before it skipped a tick. It was all the way down to 111 before it missed one. That's 48 ticks into the 90% range. I must have been really lucky. Odds on that are less than one in a hundred. But I'll do more testing before I make too much fuss.

At this point I'll say I think the algorithm is probably correct, but the actual numbers may be a bit off for my version of DM.

At any rate, using Dain potions to raise Wisdom certainly CAN give me lots more mana. The diminishing return point is different than I expected. For beginning characters, I pretty much have to max out my Wisdom to even get to the break even point. The break even point is like 12 or 13 potions stacked, and even my worst beginner had too much wisdom not to max out with that many potions. Assuming 9 Wisdom points per potion. If I KEEP my wisdom at those high levels over time, it becomes very profitable very quickly. There is another issue, however. As soon as I reach 40 max mana, the regen multiple kicks in and moves the break even point down to 7 potions. Since the Dain potions is a level 4 Priest spell, chances that I'll even be able to cast that spell reliably with less than 40 max mana are pretty low. The higher my max mana, the less potions I have to use to break even. My = Master characters break even at 2 or 3 stacked potions. If I buff them up to 144, I'm getting major returns on my mana investments.

Bottom line is, I'm going to buff all my characters up to over (144 - (priest + wizard levels)) as quickly as I can and keep them up there. I'll have more mana than I can shake a stick at.

[Sophia]

At this point I'll say I think the algorithm is probably correct, but the actual numbers may be a bit off for my version of DM.

The actual numbers were off, but I think that was my mistake, not a version issue. The short answer is that the "key number" is 112, not 144.

Out of curiosity, I generated the exact sequence used by DM, and it turned out to be:
0, 64, 32, 96, 16, 80, 48, 112

DSB turned out to have gotten it right too, I just misread the code, seeing a >= and thinking "inclusive" when that was really the values it was cutting off. So, the sequence used by DSB is:
0, 32, 64, 96, 16, 48, 80, 112

Not exactly the same, but close enough for a DM player!

Most significantly, DSB preserves the "gap" in the sequence: starting at 0 and then starting over at 16. I remember, now that I think about it, a time when it didn't do this, and people complained that DSB regeneration was too smooth, and didn't have the "stop and start" effect they were used to in DM.

[Maven]

The actual numbers were off, but I think that was my mistake, not a version issue. The short answer is that the "key number" is 112, not 144.

That would change my test to 6/8 or 75% instead of 60%. 27 / 35 is 77%. That falls right in line. However, I'm pretty sure my test character skipped a tick then the wisdom was 111. Subtracting the 22 points for Wizard + Priest levels, it shouldn't skip a tick until the wisdom gets down to 90.

This is nice, though, because it is a lot easier to keep up above 112 than 144. My beginner test characters get less of a wisdom boost when I drink a potion when my wisdom is already over 120. The bonus falls to 5 instead of 9. If anything over 112 is wasted, I like this better.

[Maven]

Note that you can still succeed at casting a spell even if your skill is below the required value, because you get a "second chance," when your wisdom is compared to a random number between 0 and 127. If you pass this check, you're able to cast the spell, no matter how difficult it is.

Hmmm. I just noticed this.

So... does that mean that if I buff my wisdom above 127 I will ALWAYS succeed any spell, no matter how difficult? This could make Dain potions insanely powerful. I'm going to have to test this. Imagine casting a Mon Lightning Bolt from a neophyte wizard and gaining two wizard levels with one spell.

[beowuuf]

Catch 22 though, isn't it - how are you going to get a dain poton that powerful and keep it coming to keep pace with the wisdom degredation under normal gameplay conditions without the mana and wisdom of high levels to get you the potions in the first place!

Try it though, it might be an interesting mechanic to have a powerful monster at a very low level, and the low level player has a set of dain and ee potions sitting around to give a single shot a killing the critter, with suitable reward if he can pull it off.

[Sophia]

So... does that mean that if I buff my wisdom above 127 I will ALWAYS succeed any spell, no matter how difficult? This could make Dain potions insanely powerful. I'm going to have to test this. Imagine casting a Mon Lightning Bolt from a neophyte wizard and gaining two wizard levels with one spell.

I took a look at the actual source code to make sure I was remembering this right, and I found out it's even more interesting.

What it actually does is generate a number between 0 and 127, and compare it to your wisdom + 15, but caps it at 115. This means that 100 is the highest you need to go, but there will always be a chance of failure. In addition, you have to pass this check once for every level below the spell's "minimum" you are. That is to say, if you're a neophyte and you try to cast a mon level spell, you have to pass the check at least 6 times. Even if your wisdom is buffed to 100, that is still only a 55.12% chance of success (that is, (115/127)^6)

[Paul Stevens]

Clever!
There is lots of cleverness in that program.

[Octopuss]

Is the claim about chars standing in the rear getting less xp (when ther party is being attacked) true or not?

[Maven]

Catch 22 though, isn't it - how are you going to get a dain poton that powerful and keep it coming to keep pace with the wisdom degredation under normal gameplay conditions without the mana and wisdom of high levels to get you the potions in the first place!

Try it though, it might be an interesting mechanic to have a powerful monster at a very low level, and the low level player has a set of dain and ee potions sitting around to give a single shot a killing the critter, with suitable reward if he can pull it off.

Yes, and I thought of that also. However, you only have to have your PRIEST level up high enough to cast the Dain potions. That's why my example had a neophyte Wizard. I don't think I would really be neglecting my Wizard practice like that in a real game. So I agree, it would just be an interesting experiment. Not likely to have that kind of usefulness in a real game.

However, I have run some of the numbers as to mana/experience efficiency, and in general the higher the power level of the spell, the higher the mana/exp ratio. It would have been worth it to train on Mon Ful spells or Mon Oh Ven instead of Lo Ful if success could have been guaranteed. Probably not if it's only 55% sccess rate.

[Maven]

Is the claim about chars standing in the rear getting less xp (when ther party is being attacked) true or not?

I don't know which claims you're referring to, but I've found at least one example where the rear rank gets less exp. That is with Warcry. If you're not toe-to-toe, you don't get the Priest Influence component of the exp.

Of course you don't get other melee exp if you're not toe-to-toe, either, but in their cases the program lets you know with the "Can't Reach" error message.

You might be able to get away with shield blocking from the rear rank, and that might have a toe-to-toe requirement for fighter parry experience. I'll put that on my list of things to test when I get to them.

As far as all the claims about getting experience for getting attacked or hit... Not in my version of DM. The only effect getting attacked or hit has on exp is the exp bonus. You still have to make an attack or cast a spell to actually get exp. You just get more of it if you have been attacked recently. This is a lot different than other role playing games, where you get exp for actually killing things or acquiring treasure.

[Octopuss]

Hm, well, I'd like to see someone check the code to be absolutely sure.
I wanted to increase the melee skill associated with chop, but the supply of screamers is limited, so I thought about putting the party in the room (do I need to mention which? ) with front chars chopping at the doors while being attacked from behind. But if the xp is really lowered, then there is probably no point in doing so.

[zoom]

I've found at least one example where the rear rank gets less exp. That is with Warcry. If you're not toe-to-toe, you don't get the Priest Influence component of the exp.

are you postive about this one? I really am sure that you get the priest experience regardless of rear rank or front rank when facing a monster. If you want to tell us that there is still priest experience but only less so when in the rear ranks, please say so. I do not get a clear statement out of your sentences in this regard. Anyway thanks for looking into this stuff always interesting.