[V] Question on Extracting Monster Power

[Magnate]

I think this is right, though I sought to partially address this by adding spell and melee power. I believe a lingering problem is spell power is scaled by frequency but melee is not. E.g., a 1_IN_3 monster's spell power is multiplied by 1/3 but her melee is not multiplied by 2/3 and it probably should.

Definitely, now that you're adding them. Good catch. This would boost 1_IN_2 casters with weak melee ...

[Derakon]

Erratic movement should also penalize melee power without affecting spellcasting power.

Though it always struck me as weird how a creature that can't walk a straight line can fire arrows 200' with perfect accuracy...

[nullfame]

Erratic movement should also penalize melee power without affecting spellcasting power.

Though it always struck me as weird how a creature that can't walk a straight line can fire arrows 200' with perfect accuracy...

Haha. I don't know if there are any of those but good point.

FWIW, a melee penalty for erratic movement already exists. IIRC it reduces damage by 50% for RAND_50 and 25% for RAND_20, then adds back the 1/8 chance of a random movement hitting the player. A lot of the old code was pretty well thought-out. I'm not saying don't keep making suggestions/observations, just trying to give credit where it is due.

[nullfame]

Well, upon further inspection, my changes have thoroughly broken rebalance. For people that don't know what this obscure run-time option is, it "rebalances" the monsters depth, rarity, and experience based on the power ratings of the entire monster list. E.g., so a scrawny cat in town is rarity 3 normally, but isn't really dangerous so it is made rarity 1 on rebalance. Greater mummies get moved deeper, etc. I was totally unaware such a option existed and doubt many people use it. But, nonetheless, my changes may be a step backwards since they don't improve gameplay and may break something that was working.

I have a set of changes that partially addresses this but doesn't fix rebalance 100% (it changes greater mummy from rarity 170 to 11, which is still way too high but not hyperbolic). I intend to keep working on the monster power algorithm but could use a little advice:

1. Submit an updated patch with my partial fix to rebalance.

2. Don't commit my changes, keep following this awesome thread on an arcane calculation making improvements, and fix rebalance when the power rating seems otherwise right.

Now that I think I have a pretty firm grasp on the current algorithm, I wonder if the power rating was meant for rebalance alone, never to be otherwise useful. Of course if that's true why is it being saved in the monster struct.

[Derakon]

How does your patch break the rebalance option? I wouldn't worry about it too much as I've never even heard of it, unless your patch is doing things like making AMHDs show up at dlvl 15.

[Magnate]

Well, upon further inspection, my changes have thoroughly broken rebalance. For people that don't know what this obscure run-time option is, it "rebalances" the monsters depth, rarity, and experience based on the power ratings of the entire monster list. E.g., so a scrawny cat in town is rarity 3 normally, but isn't really dangerous so it is made rarity 1 on rebalance. Greater mummies get moved deeper, etc. I was totally unaware such a option existed and doubt many people use it. But, nonetheless, my changes may be a step backwards since they don't improve gameplay and may break something that was working.

I have a set of changes that partially addresses this but doesn't fix rebalance 100% (it changes greater mummy from rarity 170 to 11, which is still way too high but not hyperbolic). I intend to keep working on the monster power algorithm but could use a little advice:

1. Submit an updated patch with my partial fix to rebalance.

2. Don't commit my changes, keep following this awesome thread on an arcane calculation making improvements, and fix rebalance when the power rating seems otherwise right.

Now that I think I have a pretty firm grasp on the current algorithm, I wonder if the power rating was meant for rebalance alone, never to be otherwise useful. Of course if that's true why is it being saved in the monster struct.

Definitely #2 - don't worry about rebalance at all. It was put in a while ago but has never been very widely used. Ultimately it should be unnecessary because the monster list should be properly balanced (i.e. rebalancing should be a no-op), but in the meantime we can live without it while the power algorithm is improved. When we bring it back it should be to help fine-tune the power algorithm until it is eventually a no-op.

A quick bit of history, in case anyone is interested: the monster power rating started life in randart.c purely as an input to slay_power (which is now in obj-power.c) - i.e. in order to rate slays and brands empirically. Using it to rebalance came after that, and was entirely UnAndrew's work - he improved it dramatically.

[andrewdoull]

It's also true that monster power tries to encode two different things:

1. The chance that the monster will kill you even if you try to avoid it
2. The difficulty in killing the monster if you fight it

While they are harder to avoid, I do think Dark Elven Sorcerers are way easier to kill than Storm Giants.

There is some attempt to do this: there's actually two values calculated. Power is option 2; there is also a highest_threat value which mimics some of option 1.

The rebalancing algorithm works more effectively if both are used. At the moment, the highest_threat is scaled by the frequency with which the monster casts spells - I could see an argument against that.

Andrew