Crawl with difficulty setting

[Sandman25]

monster_hp_difficulty=100
Possible values are 1-1000
Monster HP is changed as
hp = hp * monster_hp_difficulty / 100;
Default is 100.

You can download the patch at https://www.dropbox.com/s/kwk24bfxtim3o ... patch?dl=0
or download trunk crawl (as of 06/04/2015) with this and other my changes at
https://www.dropbox.com/s/f9mkoesujx03a ... l.zip?dl=0
My other changes include:
1) New option turns_for_comes_into_view_again = 10
Default is -1 i.e. the message is disabled.
2) Damage dealt to monster is displayed for most sources
Cannot be hidden
3) To_hit for damaging spells (success chance similar to hexes)
Cannot be hidden
4) New option confirm_action += Borg
confirm_action += Death's Door
You will get confirmation dialog before casting the spells
Default is none

[bcadren]

Making increased difficulty by making Centaurs actually kite intelligently would be a much more interesting difficulty jump than +X% HP.

[Sandman25]

Sorry, I am not expert in crawl code or C++.
Not sure why I made 1000 max, even 200 is extremely hard.

[Sar]

You could always try playing 4.1 beta, it has kiting centaurs (or so I heard) and much more!

It's available for online play on CDO. It's not Stone Soup though, just DC, so no autoexplore sadly!

[wheals]

I believe that autoexplore was one of the things that actually got backported from the "travel patch" (greensnark's patch that eventually mutated into Stone Soup) into 4.1. IIRC it's bound to ctrl-O.

[dpeg]

wheals: Yes, that's true. Darshan did this before (or maybe at the same time) as he and Erik were preparing DCSS 0.1.

Regarding the OP, I agree that scaling monster HP is one of the best ways to achieve difficulty settings. (Another option would be a global experience modifier.) I don't think Crawl should ship with difficulty settings, but nonetheless I'd be interested in feedback. (E.g. estimates which monster HP increases still allow for reliable wins on easy combinations.)

[Sandman25]

Global experience modifier will result in less HP and in increased chance of unavoidable deaths, I don't want to make it easier to die, I want to make it harder to kill things.

Here is some feedback. I've never won TeAE but it is too easy by default, here is why I created the patch. Initially I went with 200% HP for monsters. I died to first monster (dungeon entrance was a 1 tile room with a door, I opened it and found Ooze who blocked my path out of the room, I ran out of MP and UC didn't help). That was unavoidable death.
So 200% HP is unreliable.
After that I changed difficulty to 150% HP. First game was too lucky (2 potions of experience on D4) so I didn't care much about the character and it died in Lair 4 to a Spiny Frog, it wasn't fun to play 150% HP when you have Conjurations/Fire 12 on D4.
I was very careful with next TeAE, it's just got silver rune and is very fun/challenging. viewtopic.php?f=12&t=16433
Saint Roka with 300 HP was incredibly hard (Haste, Conjure Flame, Sticky Flame, Bolt of Fire, Lightning Bolt, Conjure Ball Lightning), I was able to kill it from 3rd attempt.
Jory with 270 HP was blocking Crypt so I killed it after clearing Depths, using Haste, Tornado, Silence and melee (+7 scimitar of slicing), didn't kill it from first attempt, retreated, waited for contamination to disappear and repeated the same, Jory wasn't still at full HP.
Vs Curse Skulls I had to use might. Elf was fun with 120 HP Deep Elf Annihilators and 150 HP Deep Elf blademasters. I am not sure if it can be called reliable but 150% HP is definitely playable.

[linear]

I don't think Crawl should ship with difficulty settings

I think Crawl should ship with difficulty settings. Sandman's HP mod is one of many ways to help players stick with the game longer. Good to see this.

[dpeg]

linear: A large part of Crawl is about getting a first win, then counting wins, and special wins. Having to qualify by this how those wins were gotten seems really impractical to me. My hope would be that feedback from a branch or fork like Sandman's might allow us to tune the game better.

[Sandman25]

My hope would be that feedback from a branch or fork like Sandman's might allow us to tune the game better.

How? By nerfing Tornado and Orb of Destruction?
I have just killed first OoF in Zot 1, I checked message log and found that I dealt about 50 HP with weapon, everything else was dealt by Tornado (OoF had 225 HP).
And here is "fight" vs Bone Dragon:

  Code:

The orb of destruction hits the bone dragon.(damage 80)
The bone dragon is moderately damaged.
Casting: Orb of Destruction
Confirm with . or Enter, or press ? or * to list all spells.
Aiming: Orb of Destruction
Press: ? - help, Shift-Dir - straight line, f/p - bone dragon
Aim: a bone dragon (moderately damaged)
The orb of destruction hits the bone dragon.(damage 62)
The bone dragon is heavily damaged.
Casting: Orb of Destruction
Confirm with . or Enter, or press ? or * to list all spells.
Aiming: Orb of Destruction
Press: ? - help, Shift-Dir - straight line, f/p - bone dragon
Aim: a bone dragon (heavily damaged)
The orb of destruction hits the bone dragon.(damage 41)
The bone dragon is severely damaged.
Casting: Orb of Destruction
Confirm with . or Enter, or press ? or * to list all spells.
Aiming: Orb of Destruction
Press: ? - help, Shift-Dir - straight line, f/p - bone dragon
Aim: a bone dragon (magical condensation, severely damaged)
* * * LOW MAGIC WARNING * * *
The orb of destruction hits the bone dragon.(damage 59)
The bone dragon is almost destroyed.
The wavering orb of destruction hits the bone dragon.(damage 56)
You blow up the bone dragon!

Bone Dragons have 228-307 HP. 5 Orbs dealt 298 damage, 59 on average.

[dpeg]

Sandman: Of course. Tweaking monster parameters and spell parameters is an on-going process. More numerical feedback can only help.

[Sandman25]

linear: A large part of Crawl is about getting a first win, then counting wins, and special wins.

Is it possible to count 150% monster HP as special win? I mean some games require players to win normally before allowing to play in some challenge modes. After player gets greaterplayer status, it could be interesting to get "greaterplayer150" status for winning with all species when monsters get 150% HP. Then "greaterplayer200" and so on.

[mps]

While Sandman25 was away, we replaced his usual dungeon crawl with new one half monster hp crawl. Let's see if he can tell the difference...

[Sandman25]

While Sandman25 was away, we replaced his usual dungeon crawl with new one half monster hp crawl. Let's see if he can tell the difference...

As you can see I display damage so mistake cannot happen here, I read messages regularly to make sure the monsters have 150% HP (for example, killing fire dragon in 2 turns by double-zapping was unexpected), also it is fun.

[Sandman25]

I have just killed Tiamat, it had lajatang of speed. If I checked wiki before fighting it, I wouldn't fight. 570 HP, omg. I had to use 2 potions of healing to survive.

[savageorange]

Have you considered putting this (the full git repo, not just this patch) up on github? It would make trying this out easier for people who don't use Windows.

[KoboldLord]

Based on my experience in playing Slash'EM, giving non-threatening monsters extra hit points just makes killing them a tedious chore. It's probable that with any given hp multiplier there will be a sweet spot of monsters that are made more interesting by the change, the vast majority of monsters are not the sort that will be improved by tripling their hit points. You'd just spend more time kiting them around under carefully controlled conditions, and we already have plenty of that in the game.

[Sandman25]

Have you considered putting this (the full git repo, not just this patch) up on github? It would make trying this out easier for people who don't use Windows.

How is it different from applying the patch? All those patches which are in my executable are available on mantis. Sorry, I don't even have github account and I don't think many players would use it.

[Sandman25]

Based on my experience in playing Slash'EM, giving non-threatening monsters extra hit points just makes killing them a tedious chore. It's probable that with any given hp multiplier there will be a sweet spot of monsters that are made more interesting by the change, the vast majority of monsters are not the sort that will be improved by tripling their hit points. You'd just spend more time kiting them around under carefully controlled conditions, and we already have plenty of that in the game.

Is Ooze a non-threatening monster? I died to it with 200% HP multiplier, it was first monster which I met. All those gnolls, Orc Priests, Jackals and Adders become much more interesting/challenging. Even plain Orcs in Mines caused some problems, you cannot kill them fast enough to reach Orc Sorcerers/High Priests so you have to change tactics.

I won the 150% game but my character was almost perfect (Sticky Flame/Conjure Flame/Fireball/Silence/Haste/Tornado/Bolt of Fire/Bolt of Lightning/rN+++/rF++) and it had problems literally at all stages.

[savageorange]

Have you considered putting this (the full git repo, not just this patch) up on github? It would make trying this out easier for people who don't use Windows.

How is it different from applying the patch? All those patches which are in my executable are available on mantis. Sorry, I don't even have github account and I don't think many players would use it.

The difference is, getting a patched version currently requires:
* clone git master
* apply the patch via 'git apply < your_patch.patch'
* potentially have to resolve patching conflicts, if git master has changed sufficiently.
* compile and install

Whereas if it were on github, it can be used directly as a build source, for example simply
reinstalling this package, editing the PKGBUILD to specify "git+https://github.com/sandman25/crawl-ref.git" instead of "git+https://gitorious.org/crawl/crawl.git" at the start of the process. Very straightforward, no user intervention aside from choosing where to clone from.

(also, patches on mantis have to be hunted down. Searching 'sandman25', 'sandman', 'sandman25 patch' or 'sandman patch' on mantis displays no patches by you*. Putting your changes on github means it's all collected sensibly.)


* the correct search strings appear to be 'confim_action' (sic), 'turns_for_comes_into_view_again', 'to_hit', 'damage' (and one must also use advanced filters to specifically select reporter = Sandman25 for this case); and this particular monster hp setting does not appear to have a patch on Mantis.

[Sandman25]

Sorry, I am not interested in rebasing my changes to keep them up to date with trunk.
Also I don't understand what you mean as package etc. (I am a windows guy).

Hopefully this post will be helpful, it has some links to the patches.

Right, I didn't post monster HP patch on mantis because it will never be accepted and I wouldn't like to get "why is it on mantis?" again.

[savageorange]

Sorry, I am not interested in rebasing my changes to keep them up to date with trunk.

That's exactly what I am saying: If you don't maintain a proper repo, then people will HAVE TO rebase your changes when your patches break. If you do maintain a proper repo, then you can provide exactly the version that you want and no adjustment is necessary, because you are providing all the necessary data for that exact version, not just a patch. It's easy for you and easy for them - nobody has to bother rebasing unless they specifically want to.

Also I don't understand what you mean as package etc. (I am a windows guy).

Just things that you can install, in one or two clicks. If you've ever used Steam, or the Google Play store on an Android based phone, imagine basically everything on your system could be installed/uninstalled/etc with a similar level of ease.

Windows has things that resemble packages (setup.exe's and .msi's, with InstallShield). It's just messier and less consistent than an actual formal package management system.

Hopefully this post will be helpful, it has some links to the patches.

Thank you, that is IMO an important link to give whenever you mention them (or to put in your signature).

[Sandman25]

Sorry, I am not interested in rebasing my changes to keep them up to date with trunk.

That's exactly what I am saying: If you don't maintain a proper repo, then people will HAVE TO rebase your changes when your patches break. If you do maintain a proper repo, then you can provide exactly the version that you want and no adjustment is necessary, because you are providing all the necessary data for that exact version, not just a patch. It's easy for you and easy for them.

Also I don't understand what you mean as package etc. (I am a windows guy).

Just things that you can install, in one or two clicks. If you've ever used Steam, or the Google Play store on an Android based phone, imagine basically everything on your system could be installed/uninstalled/etc with a similar level of ease.

Windows has things that resemble packages (setup.exe's and .msi's, with InstallShield). It's just messier and less consistent than an actual formal package management system.

Hopefully this post will be helpful, it has some links to the patches.

Thank you, that is IMO an important link to give whenever you mention them (or to put in your signature).

No, I have never used Steam or Google Play but I think I got the idea. I will register and will try to create the github repository.

[savageorange]

No, I have never used Steam or Google Play but I think I got the idea. I will register and will try to create the github repository.

Thank you for your time and consideration
I am familiar with the process of forking an existing repository on github, and am happy to help resolving any issues you may encounter.

[Hirsch I]

I have never used Steam or Google Play

[KoboldLord]

Based on my experience in playing Slash'EM, giving non-threatening monsters extra hit points just makes killing them a tedious chore. It's probable that with any given hp multiplier there will be a sweet spot of monsters that are made more interesting by the change, the vast majority of monsters are not the sort that will be improved by tripling their hit points. You'd just spend more time kiting them around under carefully controlled conditions, and we already have plenty of that in the game.

Is Ooze a non-threatening monster? I died to it with 200% HP multiplier, it was first monster which I met. All those gnolls, Orc Priests, Jackals and Adders become much more interesting/challenging. Even plain Orcs in Mines caused some problems, you cannot kill them fast enough to reach Orc Sorcerers/High Priests so you have to change tactics.

I won the 150% game but my character was almost perfect (Sticky Flame/Conjure Flame/Fireball/Silence/Haste/Tornado/Bolt of Fire/Bolt of Lightning/rN+++/rF++) and it had problems literally at all stages.

The ooze you encountered is only one of several possible failure states for a difficulty mod to avoid. The extension of fights against non-threatening monsters is a separate failure state, and one that I have encountered in other games.

In the case of plain orcs, outside of a few Orc entry vaults and the like, it wouldn't make any difference to difficulty if every one of those vanilla orcs had 200hp apiece. If they're too resilient to just roll over, you just separate their pack and kite them around until they die. This sort of activity is prevalent enough in the base game, and I don't think a simple hit point multiplier is really quite capable of any consistent difficulty modification. The job really calls for a re-design of every monster for every difficulty level to make sure that the stats getting improved are actually meaningful stats and that they are getting improved in appropriate ways, but that would be quite a lot of work.

[Sandman25]

Thank you for your time and consideration
I am familiar with the process of forking an existing repository on github, and am happy to help resolving any issues you may encounter.

Please let me know if https://github.com/Sandman25DCSS/difficulty/ works for you.

[Sandman25]

The ooze you encountered is only one of several possible failure states for a difficulty mod to avoid. The extension of fights against non-threatening monsters is a separate failure state, and one that I have encountered in other games.

It was possible with 100% HP too, the mod just stressed the problem. There should be a guaranteed monsters free area, otherwise presence of unavoidable deaths is um, unavoidable. My point was that "harmless" monsters are not exactly harmless when they have high HP and block path.

In the case of plain orcs, outside of a few Orc entry vaults and the like, it wouldn't make any difference to difficulty if every one of those vanilla orcs had 200hp apiece. If they're too resilient to just roll over, you just separate their pack and kite them around until they die. This sort of activity is prevalent enough in the base game, and I don't think a simple hit point multiplier is really quite capable of any consistent difficulty modification. The job really calls for a re-design of every monster for every difficulty level to make sure that the stats getting improved are actually meaningful stats and that they are getting improved in appropriate ways, but that would be quite a lot of work.

I don't think kiting is as easy as you claim it to be. I died several times to a 200% HP almost dead Ogre because it had too high HP and I was blocked by other monsters. Basically you need to have clear levels and/or advantage in speed. Swiftness is hardly enough, for example. When you cannot kill monsters which block stairs, you can have unusual problems. For example, my TeAE almost died when I tried to stairdance with heavy hitter and some other monsters because all 3 stairs were blocked by monsters from the same huge vault on D11. The same happened again on U2 (Mara + Saint Roka, later Norris), again close stairs.

[savageorange]

Please let me know if https://github.com/Sandman25DCSS/difficulty/ works for you.

Currently building from this...

EDIT: Actually, looks like 'git describe' is unhappy with it; it was unable to extract a sensible version number. Looking more closely at it, it seems that you have done something other than a straight fork -- your repository does not include any tags.
That's a little odd; probably it means that whatever you used to clone the crawl repository only cloned some of the metadata, not all of it
(ie. did something other than plain 'git clone http://github.com/crawl/crawl.git').

Anyway, the easiest fix should be to tag commit ba81fef2d087a556a87a8045e787d33bd098343c as '0.17-a0' -- this is the commit tagged 0.17-a0 in trunk, and it is included in your repo. On the commandline, this is done with 'git tag -a', as in 'git tag -a 0.17-a0 ba81fef2d087a556a87a8045e787d33bd098343c'. This allows 'git describe' to successfully produce a description, which allows a proper version string to be constructed and compilation to complete successfully.

So performing that action and then pushing the updated repository should cause everything to work OK.

[tabstorm]

I like this idea, but I actually think making the numbers in crawl smaller could be a better way to improve difficulty.

[savageorange]

Thank you for your time and consideration
I am familiar with the process of forking an existing repository on github, and am happy to help resolving any issues you may encounter.

Please let me know if https://github.com/Sandman25DCSS/difficulty/ works for you.

As it is, there seems to be some missing metadata preventing compilation from fully completing. I have described this in my edit above, including a solution.

(I now have a working build, after carrying out the above solution locally. Plays normally and damage numbers are showing, so I think it's all good, haven't tested any other part of it yet.)

[Sandman25]

As it is, there seems to be some missing metadata preventing compilation from fully completing. I have described this in my edit above, including a solution.

(I now have a working build, after carrying out the above solution locally. Plays normally and damage numbers are showing, so I think it's all good, haven't tested any other part of it yet.)

Sorry about that. I am not a big fan of command line (mildly speaking) so I used egit plugin for eclipse when working with git. I created a release and a tag using github UI, hopefully it will fix the issue for other interested players.

https://github.com/Sandman25DCSS/diffic ... /0.17_diff

[Sandman25]

I like this idea, but I actually think making the numbers in crawl smaller could be a better way to improve difficulty.

Sorry, I missed the comment. What do you mean? You can set difficulty to any integer number between 1 and 1000.

[savageorange]

As it is, there seems to be some missing metadata preventing compilation from fully completing. I have described this in my edit above, including a solution.

(I now have a working build, after carrying out the above solution locally. Plays normally and damage numbers are showing, so I think it's all good, haven't tested any other part of it yet.)

Sorry about that. I am not a big fan of command line (mildly speaking) so I used egit plugin for eclipse when working with git. I created a release and a tag using github UI, hopefully it will fix the issue for other interested players.

https://github.com/Sandman25DCSS/diffic ... /0.17_diff

This seems to work. Surprising, because my test indicated that the tag needed to be annotated, whereas that tag is un-annotated. Well, whatever, it works


Tabstorm's comment, I think, refers to modifying player HP as well as monster HP; and possibly other stats. It's been occasionally discussed by devs, the idea that a lot of crawl's numbers, especially stats, would be more meaningful if they were smaller.

[daggaz]

The comment about increasing HP also increasing tedium certainly has merit. Another way to do this is to implement a slider on damage, that way you can up the danger without changing how long it takes to kill them. Personally, Id stick with both options.

[Sandman25]

The comment about increasing HP also increasing tedium certainly has merit. Another way to do this is to implement a slider on damage, that way you can up the danger without changing how long it takes to kill them. Personally, Id stick with both options.

Character HP is not increased and if monsters have 150% max HP it means you are losing 50% more HP in a typical fight. For example, if you usually lose 10 HP vs a Kobold, now you will lose 15 HP. When your max HP is 25, it means a lot.
I am playing MfGl with 150% HP (going for Oka or maybe Makhleb), first character died after shafting from D2 to D4 without any chance to survive, second is still alive on D3. The only tedious moment was when I killed a single worm, that was free XP with trident, another slightly boring moment was when I killed a goblin with a single stone (damage 3), I checked wiki and goblin's min HP is 2 indeed. Even 2 Worms + Kobold + Goblin fight was interesting, I had to run a lot and monsters were joining the fun.

[Rast]

Presumably, with a good enough vampiric weapon, even 1000% monster HP Crawl is winnable.

Just 900% more tedious.

[Sandman25]

Presumably, with a good enough vampiric weapon, even 1000% monster HP Crawl is winnable.

Just 900% more tedious.

Yes, vampiric will be a problem. Actually you can change the option to 1 when you get vampiric executioner's axe, all new monsters will have 1% of their normal HP

[wheals]

Meatsprint mode!

[Sandman25]

Even better IMHO. I tried it in wizmode, killed a Hill Giant with a stone.