Balance issues in current V.

[Pete Mack]

TL;DR: After the recent changes to pathfinding, I figured balance might be affected for some common cases. To test it, I rolled up a Half-Troll Warrior. The result: very tedious play. I rather expected this; in fact I've expected it for a while, which is why I haven't played this combination in many years.

Pathfinding
Better pathfinding penalizes poor stealth and poor device skill. With a half-troll warrior, annoying monsters like GMMs keep waking up, and keep coming back. Dangerous monsters like Carcharoth keep coming back, and the danger is amplified because of the high fail rates. (Yes, you can game this using breath attenuation, but again: tedious.)

Further, this really hurts with no-ESP play. If you don't find ESP early, you are at significantly greater risk from monsters coming out of nowhere to kill you. It's not enough to detect each region once. Since instakill monsters are much likelier to come back, you either need ESP, or you need frequent detection, which is right out. The light rod technique becomes much less reliable. In this game, I found an amulet of ESP at 1700', so this isn't a problem. Without it, the temptation to scum 2000' for rPoison becomes much greater. (The same holds for priest-paladin, albeit to a lesser extent.)

There are a lot of moving parts here: breath attenuation, fail rate, stealth, pathfinding. Of course attenuation tends to make non-melee fights trivial, so it's not so good to rely on it as a solution.

Blow count and damage.
The other balance issue that applies particularly to H/T Warrior is the near impossibility of using heavy weapons after the fractional-blows change. The result is that unless you are lucky with lots of DEX, the optimal HT/Warrior weapon is essentially Aglarang. Yes, it's a pretty nice weapon, but it means that end-game Rogues are likely doing as much or more damage per turn.


Analysis
The root cause of this trouble is that under the current design, DEX is more valuable than STR, especially for warriors. To get to 6 blows with a weapon of 11lb or more, you need DEX 18/200--even with STR 18/220. (Note that Aglarang has weight 5lb and +5 DEX, and you see why it's so valuable.) Further, even with DEX 18/200, you can't get 6 blows beyond 15lb. With a BoC, you are already down to 5.9, and that's at S18/220, D18/200. Beyond that, things drop off slowly, but only if you have DEX 18/200. For a Half-Troll Warrior, even getting 5 blows with a heavy weapon is problematic.

In comparison, a Rogue gets 5 blows with a MoD, the heaviest weapon in the game, at STR 18/200, DEX 18/200. That's (a) possible, and (b) relatively easy to get--especially for the most common races (Hobbit, Kobold, HE). With the Glaive of Pain or Scythe of Slicing, S18/150,D18/200 or S18/210, D18/180 gets 5 blows. Getting 5 blows with a BoC is trivial.

The difference, of course, is that dexterity adjustment used to top out at 18/150, which is relatively easy to get--even for a H/T Paladin, who needs DEX +9.

Possible fixes:
* Give warrior back his extra blow.
* give more DEX to warrior.
* Reduce race DEX penalties by half (I hate this, since it affects other classes.)
* Make adj_str_blows superlinear at the last few rows.
* Give Warrior higher strength multiplier (6 instead of 5)
* Increase benefit of DEX between 18/130 and 18/200.

Doubtless there are more.

[Estie]

As for str, dex and blows, it has been a long time beef of mine that for light weapons, more str gives more blows.
Long ago I suggested a system where dex determined amount of blows whereas str put a cap on the weight for weapons that allowed for the maximum amount of blows; insufficient str for a wielded weapon would reduce the #blows gradually.

Various ways to give the warrior more attacks than the spellcasters are possible, but the most simple and sufficient one is simply to use one attack table (caster, 1-4 blows), add 1 blow for hybrids and 2 blows for warriors.

Better pathfinding, or ai in general, does not necessarily lead to a better game. That being said, I have no problem with warriors getting the shaft from monster ai, IF they get more combat power to compensate. Currently a warrior is basically a paladin without spells.

[Sky]

hear, hear!

[Pete Mack]

The whole STR=weight DEX=Blows issue is solved in Eddieband, which is why I mentioned it. It's not an issue when DEX value maxes out at 18/150. But when it maxes at 18/200, you had better get the details exactly right--and that means CL 40 Half-Troll Warriors using the Glaive of Pain should do a whole lot of damage--and certainly more damage than a CL 50 Hobbit Rogue with the same weapon.

And to be fair, a starting warrior is noticeably better than a paladin with no spells due to the difference in starting stats, though this advantage becomes immaterial in the endgame, when a Paladin has infinite heal, and vastly more inventory flexibility.

As for str, dex and blows, it has been a long time beef of mine that for light weapons, more str gives more blows.
Long ago I suggested a system where dex determined amount of blows whereas str put a cap on the weight for weapons that allowed for the maximum amount of blows; insufficient str for a wielded weapon would reduce the #blows gradually.

Various ways to give the warrior more attacks than the spellcasters are possible, but the most simple and sufficient one is simply to use one attack table (caster, 1-4 blows), add 1 blow for hybrids and 2 blows for warriors.

Better pathfinding, or ai in general, does not necessarily lead to a better game. That being said, I have no problem with warriors getting the shaft from monster ai, IF they get more combat power to compensate. Currently a warrior is basically a paladin without spells.

[Gwarl]

I think green book classes are always going to be easy mode, and we should leave them out of it.

I think the curvature of the weight/str/dex/blows interactions needs some fairly deep analysis and subsequent correction. In fact I'd like to have a go at that myself.

[Derakon]

I like the idea of using a single blows formula and granting bonus blows to hybrids / warriors. You could do it as a level-based thing, like, warriors get 1 bonus blow at level 5 and another at level 25, paladins/rogues get 1 bonus blow at 25, rangers get 1 at 30 (just to throw some numbers out there).

[Pete Mack]

And speak of the Devil: I'm getting a respectable but not phenomenal 6.8* blows, 538 damage/round with Aglarang, which I've used exclusively since DL 47. (Prior to that I'd been using a *thanc.) I've now found a more traditional end-game weapon:
...
A mystical staircase appears!
You have slain Sauron, the Sorceror.
You see a Mace of Disruption of *Slay Evil* (+16,+13) <+2>
...
Combat info:
3.5 blows/round
With +0 STR and +1 DEX you would get 4.0 blows.
Average damage/round: 523 vs Undead, 422.3 vs Evil, and 320.2 vs others.

:sad trombone:

* And really, It's hard to believe it's even possible to get less than max blows with an endgame-strong character and a +5 DEX, 5lb weapon.

[Nick]

I'm not completely unsympathetic to these arguments, but I think they're matters of opinion rather than clear changes that need to be made. Given that monster behaviour and player classes and races are specifically up for examination in the plans for 4.2, and that the sort of alterations required here would potentially have wider effects, I think it's best to leave it as is for 4.1.0.

[Pete Mack]

Mmm. I understand your reasoning. But other changes are going in that affect balance even more, notably last minute randart changes and breath attenuation. There's getting to be a serious technical deficit on balance issues, I fear.

[Gwarl]

Mmm. I understand your reasoning. But other changes are going in that affect balance even more, notably last minute randart changes and breath attenuation. There's getting to be a serious engineering deficit on balance issues, I fear.

Don't forget status effects.

[gameplay appreciator]

Agree that it would be good to flatten stat dependence of bpr and introduce an experience level component. It would be quite rad, in my opinion, if the role of stat potions could be reduced or even eliminated in some future version. Though it is not necessary to scum for them, the incentive is definitely there.

About pathfinding, there's good reason to rethink the situation vis a vis ESP and detection, all the more so given pathfinding that frequently moves monsters from a sleeping state to an actually dangerous state. The difference between ESP and detection from magic is largely convenience. There is nothing stopping you from spamming detection spells and in fact it is a very good thing to do from the perspective of winning the game. Detection spells are actually better at detecting monsters than ESP so you should use them even if you have ESP! Now that monsters aren't just going to get stuck on a wall most of the time when they wake, this is even more true.

If you are willing to grant the player access to monster detection magic, you've already given away the farm from a design perspective. Yes, some players, maybe most, are too "lazy" (i.e. actually want to enjoy playing the game instead of spamming the same spell constantly out of combat) to use detection to its full potential, but this is no reason to punish them. Monster movement is deterministic, at least for dangerous monsters, so the experienced player knows with pretty high precision where a monster is N turns after casting detection for small N. There is no real difference between seeing them once every 10 turns and every turn as with ESP.

Honestly, killing detection would be a step in the right direction. What is the point of mechanics like light or limited line of sight when you can just use detection magic? Why allow core mechanics like that to be made irrelevant by a spell that appears in the book your guy starts with? Once you've decided to move away from the one-shot-kill-out-of-darkness paradigm, you gotta start rethinking the basic questions.

[mixer]

Agree that it would be good to flatten stat dependence of bpr and introduce an experience level component. It would be quite rad, in my opinion, if the role of stat potions could be reduced or even eliminated in some future version. Though it is not necessary to scum for them, the incentive is definitely there.

About pathfinding, there's good reason to rethink the situation vis a vis ESP and detection, all the more so given pathfinding that frequently moves monsters from a sleeping state to an actually dangerous state. The difference between ESP and detection from magic is largely convenience. There is nothing stopping you from spamming detection spells and in fact it is a very good thing to do from the perspective of winning the game. Detection spells are actually better at detecting monsters than ESP so you should use them even if you have ESP! Now that monsters aren't just going to get stuck on a wall most of the time when they wake, this is even more true.

If you are willing to grant the player access to monster detection magic, you've already given away the farm from a design perspective. Yes, some players, maybe most, are too "lazy" (i.e. actually want to enjoy playing the game instead of spamming the same spell constantly out of combat) to use detection to its full potential, but this is no reason to punish them. Monster movement is deterministic, at least for dangerous monsters, so the experienced player knows with pretty high precision where a monster is N turns after casting detection for small N. There is no real difference between seeing them once every 10 turns and every turn as with ESP.

Honestly, killing detection would be a step in the right direction. What is the point of mechanics like light or limited line of sight when you can just use detection magic? Why allow core mechanics like that to be made irrelevant by a spell that appears in the book your guy starts with? Once you've decided to move away from the one-shot-kill-out-of-darkness paradigm, you gotta start rethinking the basic questions.

Would it not be easier to generate a new Shrieker style mushroom that makes a noise if detection or ESP see it (does not have to go off immediately). Easy to kill if not ESP i.e. only makes a noise if detected or ESP.

[gameplay appreciator]

Removing monster detection seems pretty easy to me, so no I wouldn't think the new shrieker thing would be easier. I think you would still want to spam detection if there were a monster like what you suggest anyway.

[Huqhox]

Having a monster that aggravates when detection/ESP is used wouldn't be enough to prevent me using it, certainly, or even have any concerns about using it

Although it's not a bad idea for a monster

[Pete Mack]

Would it not be easier to generate a new Shrieker style mushroom that makes a noise if detection or ESP see it (does not have to go off immediately). Easy to kill if not ESP i.e. only makes a noise if detected or ESP.

Getting rid of detection is an interesting idea, but the game won't be angband anymore. Abd why stop there? Get rid of teleport spells, too. Of course the result is that everyone would have to play Paladin for every character, but so what? They're both way over powered techniques that really interfere with balance.