Ascii Dreams poll

[andrewdoull]

And I kept the diff for precisely this purpose, if you want it.

At one point the algorithm I used was included in the Angband source - in init1.c

Which begs the question when that file was removed.

[Mikko Lehtinen]

Provided Angband is balanced around making sure each battle is challenging, I believe infinite dungeons is fine.

Of course, by that I mean Hengband infinite dungeons, not any other type...

This combination really could work!

OK, how do we make each battle challenging then?

Maybe there should be some monsters, or some monster abilities, that wouldn't get any easier to defeat as you gain levels. That way even a level 50 character would face some real danger even on DL 1.

One idea, just to pick an example: Sanity score, which would depend only on race and class but not on character level. It would work just like hitpoints against some monster attacks.

Or divide hitpoints into endurance and life, with life points not increasing with levels. Usually life points would be reduced only after endurance is at zero. But some kind of attacks would reduce life points directly, like curse spells and poison.

[Magnate]

At one point the algorithm I used was included in the Angband source - in init1.c

Which begs the question when that file was removed.

It was removed circa 3.1.2, when all the text file parsing was replaced with a new system that lives solely in init2.c. Thanks for the reminder - I'll dig up the algorithm for fizzix as well, in case he's interested in seeing how the changes were derived.

[fizzix]

It was removed circa 3.1.2, when all the text file parsing was replaced with a new system that lives solely in init2.c. Thanks for the reminder - I'll dig up the algorithm for fizzix as well, in case he's interested in seeing how the changes were derived.

Anything is useful. From what I gathered, the rebalancing mainly sorted out monsters in the first 40 levels, but there are some curious differences, like time hounds being moved up.

[ekolis]

Or divide hitpoints into endurance and life, with life points not increasing with levels. Usually life points would be reduced only after endurance is at zero. But some kind of attacks would reduce life points directly, like curse spells and poison.

Hmm, that sounds similar to Steamband's Hitpoints/Woundpoints system... except in Steamband, I think WP did increase with levels, and also, it was possible to lose WP even if you had HP remaining; the chance of losing WP instead of HP from any given point of damage was simply WP / (HP + WP).

[Malak Darkhunter]

This combination really could work!

OK, how do we make each battle challenging then?

Maybe there should be some monsters, or some monster abilities, that wouldn't get any easier to defeat as you gain levels. That way even a level 50 character would face some real danger even on DL 1.

One idea, just to pick an example: Sanity score, which would depend only on race and class but not on character level. It would work just like hitpoints against some monster attacks.

Or divide hitpoints into endurance and life, with life points not increasing with levels. Usually life points would be reduced only after endurance is at zero. But some kind of attacks would reduce life points directly, like curse spells and poison.

No offense and it might just reflect the type of player I am, but the first thing that popped into my mind reading this was "Uh Complicated, not going to play it" I'm not a big fan of complexity, but that is just my opinion.

[andrewdoull]

This combination really could work!

OK, how do we make each battle challenging then?

Maybe there should be some monsters, or some monster abilities, that wouldn't get any easier to defeat as you gain levels. That way even a level 50 character would face some real danger even on DL 1.

Sorry??

I'm not sure how B follows A here? There is no need to have 1st level monsters be challenging at level 50... what you need is monsters at the right depth be challenging at level 50.

[andrewdoull]

Anything is useful. From what I gathered, the rebalancing mainly sorted out monsters in the first 40 levels, but there are some curious differences, like time hounds being moved up.

That's because the monster power evaluation function underrates the danger of time attacks.

[Mikko Lehtinen]

Sorry??

I'm not sure how B follows A here? There is no need to have 1st level monsters be challenging at level 50... what you need is monsters at the right depth be challenging at level 50.

Why do you expect the player to be at the "right" DL for his power? The optimal way to play Angband is to play at a DL that isn't too risky.

(If by optimal we mean that we want *this* character survives and wins. If we want to minimize playing time, by contrast, the optimal way is to dive fast. But I would rather like to play for survival.)

I thought you meant this problem, but I guess you didn't.

My proposal would push the player to dive faster in search for better loot, because there's more danger at *any* level.

[Mikko Lehtinen]

I believe Angband would be a better game if the most fun way to play the game would also be the optimal one.

I'm not especially in love with my lifepoints-suggestion, it was just the first one that popped in my mind.

Again, a scoring system that rewards taking risks would actually accomplish this in a way: the most fun way to play would get you the most points.

I'm not sure whether it's enough by itself. Survival and winning is more important than score to most players. Many players are willing to do even super-boring things if it gives their characters a better survival rate -- just witness how people play MMORPGs. (One of my friends used to spend hours each day to collect unguarded gold bags in Guild Wars, until he was kicked out of the game because admins though he was a bot. ) That point was also discussed in Roguelike Radio.

Maybe Angband should flash a welcome screen to new players:

"Welcome to Angband!

This game is most fun if you dive fast and take risks. You will die more often, but don't take that too seriously. Taking risks will also give you more points."

We veterans know what is the most fun play to play Angband, but many newbies really don't.

[Mikko Lehtinen]

To clarify another point:

At the moment, with connected stairs on, it is all too easy to switch levels once they start to get dangerous. Adopting Hengband method of saving floors would get rid of this problem. A huge improvent, in my opinion.

(I suppose Andrew was suggesting Hengband method for this exact same reason? If not, let us know!)

[Magnate]

That's because the monster power evaluation function underrates the danger of time attacks.

I fixed that when I did the monster spell refactor. The side effects of time attacks now have significantly higher damage-equivalent values, which should increase the max_dam result for them.

In case anyone is interested, I put up Andrew's original code (see the eval_r_power function), and the diff from this code against 3.0.8's monster.txt. These can be compared with the current code in mon-power.c (same function names) and the monsters generated with the -r runtime option.

[andrewdoull]

Why do you expect the player to be at the "right" DL for his power? The optimal way to play Angband is to play at a DL that isn't too risky.

What kills most characters is the player playing not risky enough which results in deaths from boredom and inattention. Therefore the 'real' optimal way to play is be too risky...

If you want to go the whole way, the real fix for the problem is not having character levels. But I don't think that game is Angband.

[andrewdoull]

I fixed that when I did the monster spell refactor. The side effects of time attacks now have significantly higher damage-equivalent values, which should increase the max_dam result for them.

In case anyone is interested, I put up Andrew's original code (see the eval_r_power function), and the diff from this code against 3.0.8's monster.txt. These can be compared with the current code in mon-power.c (same function names) and the monsters generated with the -r runtime option.

Ah. So the code is still in there. Nice moving it to a runtime argument... has anyone tried this recently?

[LostTemplar]

At the moment, with connected stairs on, it is all too easy to switch levels once they start to get dangerous. Adopting Hengband method of saving floors would get rid of this problem. A huge improvent, in my opinion.

I am trying another simple solution currently, using stairs just costs a lot of food in my variant (with some other changes to make food mechanics interesting).