Competition with two savefiles?

[Timo Pietilä]

I had a recent discussion about is Priest fast or not and my "opponent" used Half-Troll priest as example while I claimed that Dwarf Priest is much faster because of much better WIS.

I'm wondering, would it be possible to make comp with two safefiles: One for Dwarf and one for Half-Troll and see which one does better job. Dwarf is natural priest char, but Half-Troll has advantage of regeneration and better STR which might show up with much lower turncount for early game.

To compensate that I suggest that this comp, if it falls into comp rules, uses player active turns (no penalty for resting for mana) instead of player turns.

Interested? I can make two comp chars with standard artifact set to see which is faster.

[Nick]

Certainly. We have actually had one like that a little while ago, with two varieties of UnAngband mage. It technically appears as two competitions (81 and 82, with different ladders). And player active turns will be fine.

Just send me the savefiles at nckmccnnll@yahoo.com.au (the spammers already have that one...), and it will be done. The usual thing is to have disconnected stairs and as complete a monster memory as possible - that said, if you're submitting the savefiles, you can do it however you like.

[Derakon]

To compensate that I suggest that this comp, if it falls into comp rules, uses player active turns (no penalty for resting for mana) instead of player turns.

That negates a huge portion of the half-troll's advantage. He doesn't have to rest for mana because he can effectively regenerate it as he moves.

That said, I'm curious what the results of this comp would be. My priest is currently stuck at 4900' looking for a good damage source, and has been for the past 100k turns or so...

[Timo Pietilä]

Certainly. We have actually had one like that a little while ago, with two varieties of UnAngband mage. It technically appears as two competitions (81 and 82, with different ladders). And player active turns will be fine.

Just send me the savefiles at nckmccnnll@yahoo.com.au (the spammers already have that one...), and it will be done. The usual thing is to have disconnected stairs and as complete a monster memory as possible - that said, if you're submitting the savefiles, you can do it however you like.

Posted, check your mailbox. I made one little twist to it so that it wouldn't be too easy.

[PowerDiver]

For priests, I don't think it won't be close at all. If you want to make it interesting, I think it has to be paladins.

[Magnate]

For priests, I don't think it won't be close at all. If you want to make it interesting, I think it has to be paladins.

Because? I'm with Derakon - I think we should use player turns not active turns, but I don't see why paladins would be more interesting than priests.

Kudos to Timo for the idea. Can we try both chars or must we choose just one?

[PowerDiver]

Because? I'm with Derakon - I think we should use player turns not active turns, but I don't see why paladins would be more interesting than priests.

Kudos to Timo for the idea. Can we try both chars or must we choose just one?

The question motivating this comp is how much a half-troll's fighting skills compensate for a severe wisdom bonus on a priestly spellcaster, compared to dwarves.

It is possible that to make it interesting, it needs to be a paladin with only 5 points put into STR so the half-troll starts with an extra blow over the dwarf.

Timo wants to be able to rest without penalty, because that is how he likes to play. One of the benefits of creating a comp is being allowed to set the rules. That weakens half-trolls comparatively speaking, making it perhaps more important to do something like what I suggested above.

The initial discussion was between a few people talking about different things, perhaps without realizing it. If Timo wants only a priest comp, then so be it. I just don't see how a half-troll can be competetive as a priest counting active turns.

[Timo Pietilä]

The question motivating this comp is how much a half-troll's fighting skills compensate for a severe wisdom bonus on a priestly spellcaster, compared to dwarves.

Not quite. It is about WIS being most important stat for priests, but it also is a test does it actually compensate Half-Troll regen as well as better fighting.

It is possible that to make it interesting, it needs to be a paladin with only 5 points put into STR so the half-troll starts with an extra blow over the dwarf.

Timo wants to be able to rest without penalty, because that is how he likes to play.

Not quite that either. Question was which is faster, and because I don't refer to turns for "how fast you are", but how you actually experience it, player active turns is closer to truth than any other counting method. I probably use much more RL time pondering should I use this armor with this headgear and that weapon or should I use this other combination than I use resting. Resting takes less than a second RL time.

If the game could set real-time counters and could magically detect when you are distracted by, lets say your favorite team making incredible goal in World Cup, then that would work. Unfortunately that is not an option.

One of the benefits of creating a comp is being allowed to set the rules. That weakens half-trolls comparatively speaking, making it perhaps more important to do something like what I suggested above.

The initial discussion was between a few people talking about different things, perhaps without realizing it. If Timo wants only a priest comp, then so be it. I just don't see how a half-troll can be competitive as a priest counting active turns.

It is about regen more than anything else. Does is actually make a difference? I think it does, but does it make enough that it compensates clear benefit of great WIS bonus? WIS is probably easiest stat to max, Blessed, *slay evil*, Holy Avenger, early WIS amulet, multiple WIS-boosting helmets, so does that initial WIS bonus really make that much difference? Half-Troll priest is probably much better than Half-Troll mage, because INT-boosting items are much less common or less useful. Half-Troll regen also helps it to clear levels faster, which means less spawned monsters, which might make a difference.

In other words, I really don't know which is faster. They require a bit different playing strategies in start of the game, but after all things considered they can be quite close to each other for start to end -speeds.

[Magnate]

Not quite. It is about WIS being most important stat for priests, but it also is a test does it actually compensate Half-Troll regen as well as better fighting.
...
It is about regen more than anything else. Does is actually make a difference? I think it does, but does it make enough that it compensates clear benefit of great WIS bonus?

This is precisely why Active Turns is the wrong measure. Game turns or player turns, but not active turns.

[Timo Pietilä]

This is precisely why Active Turns is the wrong measure. Game turns or player turns, but not active turns.

Active turns measure real time better than anything else. For the same reason you say Active turns is wrong measure I say that game turns and player turns are wrong measure. They give a very wrong picture of which one actually matters for fast game.

It is obvious that regen makes a big difference in game or player turns. It just doesn't matter because you don't notice that time used to rest.

[Pete Mack]

If you really want to compare the value of regeneration with the value of a lot of mana and spells, you have to include resting turns--that's exactly when regen pays off the most. The other time is that for utility spells near zero mana, you do better with less mana and faster regen, but that's not nearly as big a factor.


Anyway, this is the first comp to catch my fancy in a while. I just hope that the two characters are optimized for strength according to race.

Addendum.

OK, now I really hate the new purple uniques color scheme. I saw a fast moving h at dl 5, and figured no problem--just Smeagol.
Bullroarer hits you<2x>

Oh no! Only 1 phase door and no !oil. I was dead in no time. So much for my troll priest.

[Timo Pietilä]

If you really want to compare the value of regeneration with the value of a lot of mana and spells, you have to include resting turns--that's exactly when regen pays off the most.

That is a exact reason why I should not include resting turns. It twists the count to favor turns that do not use time. There is no benefit from regen in terms of game speed during resting, but player or game turns do count that anyway, so I have to use active turns instead.

OK, now I really hate the new purple uniques color scheme. I saw a fast moving h at dl 5, and figured no problem--just Smeagol.
Bullroarer hits you<2x>

It gets worse. That change turned many non-uniques to colors that are not distinguishable from much more dangerous non-uniques. IMO idea behind turning all uniques to purple does exactly opposite of what it was mean to do. Now you can't tell the difference between uniques and monsters are more difficult to distinguish from each other.

[Magnate]

That is a exact reason why I should not include resting turns. It twists the count to favor turns that do not use time. There is no benefit from regen in terms of game speed during resting, but player or game turns do count that anyway, so I have to use active turns instead.

I understand now - I hadn't realised you wanted to measure "fast" in *real* time. Then yes, active turns is best - though you could always ask people to measure their real time. Or you could postpone the comp until I can write a platform-independent routine to store the elapsed real time in the savefile ...

It gets worse. That change turned many non-uniques to colors that are not distinguishable from much more dangerous non-uniques. IMO idea behind turning all uniques to purple does exactly opposite of what it was mean to do. Now you can't tell the difference between uniques and monsters are more difficult to distinguish from each other.

Opinions on this are highly polarised and not entirely objective. A number of people find it much better to know that anything purple is a unique, and can easily establish which unique it is (e.g. by looking at the monster list, or using 'l'ook). Yes, there are a small number of monsters I changed from purple to red which are now a little bit confused with their kin (nexus Q, barrow wights, etc.). I have a ticket open to correct these after the current comp.

[Timo Pietilä]

I understand now - I hadn't realised you wanted to measure "fast" in *real* time. Then yes, active turns is best - though you could always ask people to measure their real time. Or you could postpone the comp until I can write a platform-independent routine to store the elapsed real time in the savefile

We would need a "pause" -command too for that. As I wrote people might get distracted and leave game running while they do other things.

...Opinions on this are highly polarised and not entirely objective. A number of people find it much better to know that anything purple is a unique, and can easily establish which unique it is (e.g. by looking at the monster list, or using 'l'ook).

I think that one was pretty well done in older color scheme. There were few difficult ones like ringwraiths/black wraith/Nightwing/Nightcrawler, but most of them were very easily distinguishable from rest of the population. If not from immediately following their movements usually made a difference pretty clear.

Yes, there are a small number of monsters I changed from purple to red which are now a little bit confused with their kin (nexus Q, barrow wights, etc.). I have a ticket open to correct these after the current comp.

My first Phantom scared me in this new scheme. It is now red, which is Dreadlord -color.

Also have there been some change in how those colors are made? That "v" (violet) color is much darker in my screen than it used to be. At first I thought that there is a completely new color for uniques.

[Magnate]

I think that one was pretty well done in older color scheme. There were few difficult ones like ringwraiths/black wraith/Nightwing/Nightcrawler, but most of them were very easily distinguishable from rest of the population. If not from immediately following their movements usually made a difference pretty clear.

Only for experienced players. There was no logic to the previous colouring of uniques, and less experienced players had to learn each of them.

Also have there been some change in how those colors are made? That "v" (violet) color is much darker in my screen than it used to be. At first I thought that there is a completely new color for uniques.

Yes, d_m introduced 256-colour support for 3.1.2 - in practice this means we have something like 26 usable colours rather than 16. So it is possible to have a colour that's only for uniques, without too much restriction on the rest of the monster list. I remain convinced that this is the right approach.