Brogue

This post is a summary of my assessment of the difficulty of 1.7 vs 1.6.4 and a proposed solution to increase the ratio of ascensions for moderate to experienced players.

It seems that the general consensus is that 1.7 is much harder than 1.6.4 but I'm not entirely sure this is correct. I think they are very different at various stages of the game and that melee is not only a more viable option in 1.7, but the best option.

In 1.7 the difficulty peaks with mid-depth melee monsters (ogre, wraith, trolls), not with horrors & dragons as was the case in 1.6.4. This is why I believe that players feel that 1.7 is much harder. In 1.6.4 you could often save up enchants until D12 before you had to commit to a build, but this is an almost certain path to death in 1.7. Now it seems that you need to spend every single enchant as soon as you get it on the best armor & weapon you have found by the time you start running into ogres.

My current strategy to make it past the midgame is to use the first 8-10 enchants on armor & a weapon. I have found that ogres, wraiths & trolls have been jokes with this enchant strategy. I attribute this to Pender's buffing of weapons and that I seem to find plate or enchanted splint in almost every single seed by D8.

I find that by the time I make it to the endgame (D20-26) that I have accumulated so many wands, staves & charms that the breadth of tactical solutions available makes it almost trivial. For instance, a single charm of haste allows you to smack around horrors in melee if you have +6 plate and a +6 decent weapon. A wand/charm of negation or fire immunity turns dragons into a slightly tense, but manageable experience. I could list 20 different item combos that make the endgame much easier than in 1.6.4.

I may be off in my assessment of 1.7 vs 1.6.4 but I have ascended 7 times out of the last 20 random seeds using this strategy. This is more times than I ever ascended from random seeds in 1.6.4. Many of these seeds ended in death mainly because I could not find a suitable set armor & weapon before I encountered wraiths. Another observation is that 1.7 is more random than 1.6.4 and that an ascension is made or lost by finding decent weapon & armor by D8.

Another explanation is that I just started playing Brogue in 1.6.4 have become a much stronger player since 1.7 has come out. While this is partially true I don’t think it accounts for my recent string of success.

Please discuss & dissect. I have a tendency to write as if I'm preaching when I'm really trying to encourage discussion & ideas.

Looking at my high score lists from 1.7, it seems that most of the deaths are in the first half of the dungeon. If I make it past that point, I usually ascend.

I'm getting a lot more deep deaths while I try for big lumenstone hauls, all of which would probably be ascensions if I wasn't so damn greedy, so there may be something in this, but I'd be hesitant to attribute it to anything other than luck without some more time to evaluate.

As a less than moderately good player, my 2 ascensions were all in 1.6.2.

I think I had the mistaken impression that the reduction in vaults was making it harder to select items for a good mid-game run. But every time I die it's an ogre or a wraith or a zombie, where in the past my deaths tended to involve swarms of lesser monsters. Perhaps the reduction in healing availability is getting me too.

I still haven't legitimately ascended.


Ever.


In any Roguelike

Tomahawk wrote:I still haven't legitimately ascended.


Ever.


In any Roguelike

I recommend Desktop Dungeons to boost your confidence. It took me something like 3 years before my first legitimate ascension in a roguelike, unless of course you count my own mod to TOME 2 that had a super race with all abilities and 3x HP...

I think Brogue is a good starting place as it can be done in a few hours instead of weeks as is the case for most of the currently popular RLs.

DD is okay at first but I rapidly realized that not every game is winnable no matter how clever the player is. I recommend DoomRL, it has a nice range of difficulty options.

ggoDeye wrote:

Tomahawk wrote:I still haven't legitimately ascended.


Ever.


In any Roguelike

I recommend Desktop Dungeons to boost your confidence. It took me something like 3 years before my first legitimate ascension in a roguelike, unless of course you count my own mod to TOME 2 that had a super race with all abilities and 3x HP...

I think Brogue is a good starting place as it can be done in a few hours instead of weeks as is the case for most of the currently popular RLs.

Don't get me wrong; I really like roguelikes (otherwise why would I be posting here?), and Brogue is my favorite one, hands down (after playing tons of Rogue, Moria, *Angband, and Dungeon Crawl/Dungeon Crawl Stone Soup). I just haven't been lucky/smart enough to ascend yet. I have a lot of YASD's.

I also play Desktop Dungeons a fair bit, but it just doesn't scratch the same itch for me.

@tinyrodent: just like not every game of DD is winable, not every roguelike (of any game) round is winnable either; in fact, that's the whole point!

Tomahawk wrote:@tinyrodent: just like not every game of DD is winable, not every roguelike (of any game) round is winnable either; in fact, that's the whole point!

I think you can guarantee a win in NetHack if you know what you're doing (I don't).

My first RL "ascension" was in Ragnarok/Valhalla (also my first RL) - I actually recommend this to anyone, even RL veterans - not the deepest or best balanced RL, but has some cool ideas and a great theme (Norse Mythology).

My second was in Incursion.

Brogue is only my third. In all my time spent playing the "big" RLs I've never ascended in ADOM or NH, for instance. I find a lot of these games suffer from the amount of information that is arbitrarily witheld from the player - one of the reasons I love Brogue!

Tomahawk wrote:just like not every game of DD is winable, not every roguelike (of any game) round is winnable either; in fact, that's the whole point!

I don't agree. A well designed game gives the player difficult challenges, not impossible ones.

DesktopDungeons frequently ends up feeling like a puzzle with pieces missing - even if perfectly played, the player often simply isn't given the tools for his chosen build to defeat the boss.

tinyrodent wrote:

Tomahawk wrote:just like not every game of DD is winable, not every roguelike (of any game) round is winnable either; in fact, that's the whole point!

I don't agree. A well designed game gives the player difficult challenges, not impossible ones.

DesktopDungeons frequently ends up feeling like a puzzle with pieces missing - even if perfectly played, the player often simply isn't given the tools for his chosen build to defeat the boss.

Agreed on DD. However, the random nature of roguelikes presents the possibility of unwinable rounds, and fairly frequently, in my opinion. You just feel as though you have more control.

It depends on the roguelike - 99% of games of Nethack and Dungeon Crawl: Stone Soup (assuming you don't pick a horrible race/class combo) are winnable.

Joshua Day wrote:For instance, I'm very fond of making Negation into a timed effect -- although exactly how it would work, I can't quite think.

If you mean in the code, just add a function over the bitfields negation applies to (behaviours and abilities) - that way the original bitfields are preserved and whenever negation is active things that work with it only see the negated version.

Although, would negation's effect on items in FOV be timed too? That might get silly

I have ascended twice in 1.7 and never before. The first time was a melee build and the second time doesn't really count because I scummed for a ring of wisdom+lightning staff. I also made heavy use of a haste & negation charm and a wand of negation. Horrors were a real problem for me, either requiring about 9 charges from the staff or a polymorph wand shot. Almost nothing else was a problem and I managed to only face down one dragon all game.

The only other RL game I've beaten is Ragnarok. I definitely recommend it to anyone who has not tried it yet! I've beaten DoomRL a few times too, but I guess I don't count it since it is (was?) relatively easy.

Maybe I just really suck at roguelikes.

ggoDeye wrote:My current strategy to make it past the midgame is to use the first 8-10 enchants on armor & a weapon.

This definitely improved my performance in my last play. I still died to having to fight a dozen acid moulds while naked though.

tinyrodent wrote:DesktopDungeons frequently ends up feeling like a puzzle with pieces missing - even if perfectly played, the player often simply isn't given the tools for his chosen build to defeat the boss.

I think you guys are short selling DD a bit here. When I was seriously playing it I could beat normal 80% of the time, and I expect the possible win rate to be up around 95%. Both the Alpha and Beta versions.

The only runs that are always a shutout are when you start surrounded by higher level monsters, or when you roll the worst boss for your class (and also roll mediocre gods and shops).

Man, how is DD beta these days? I really loved alpha, so I don't regret sending the designers $20, but some of the beta missions are so brutally punishing that I just completely lost interest a couple months after it came out. Also, grinding for gold strikes me as pointless (I wasn't particularly happy with the change to goblins in one of the late versions of alpha, either).

Back on topic, ggodeye's suggestion is an interesting one. I started playing at 1.6.4, and so the difficulty increase seems significant to me. Ogres, wraiths, trolls, and even centipedes and vampire bats can be serious encounters now, depending on how generous the RNG has been. In 1.6.4, by the time they showed up, they were, in most runs, easily dispatched. I'd never put enchants into anything but endgame gear before (OK. The first week or so I played, I'd put a few enchants into my darts, for some reason). I will try "wasting" a few scrolls to make my midgame gear more competitive and see if that improves my win rate.

Have had several ascensions in 1.6.4 and just 2 so far in 1.7. 11 lumenstones was my best tally, I think, in 1.6.4, but I managed a semi-legit 25 lumen run in 1.7 (see midweek beta 1). I suspect that once some of the overpowered charms (cough, protection, cough) are nerfed a bit, my ascension rate will stabilize somewhere between 3-6%.

Joshua Day wrote:

Patashu wrote:Although, would negation's effect on items in FOV be timed too? That might get silly

No, and this is one of the unpleasant inconsistencies -- negation would still have to negate items (and possibly terrain features!) permanently.

For negating effects on monsters, take this case: Would there simply be a NEGATED status, and while it's ticking down, no new status will stick? So if I NEGATE someone preemptively and then they're made DISCORDANT, they won't behave discordantly until negation runs down? I suppose that would be the most natural thing, and I can see some interesting effects from that.

(I could also conceive of some sort of pre-negation effect, where there's a fixed percent chance each turn that negation will actually set in, so if you're negated while levitating over magma you could get a turn to read a scroll of teleportation. Although in that particular case, maybe you should get such a turn anyway -- it's the turn while you would otherwise be falling.)

The inconsistencies aside, I really like this idea.

It seems to me that a creature can shrug off being negated by some sort of biological action like getting better from a cold where a negated sword would forever stay that way. I guess what I'm saying is that I hate it when my allies are permanently negated, and I would like this change.

I like how the negation effect works currently. The fact that it works poorly with allies is a positive, since it emphasizes solo strategies as a viable option. Having said that, I feel the negation charm is overpowered and should have a longer recharge schedule.

Overall I win about a quarter of the games I begin in 1.7. (I was well over 50% in 1.6.4, which I found far too easy.) The new combat dynamics invalidate portions of my old approach, and I'm still "learning" 1.7. In particular, levels 4 through 16 can be very tough. It gets easier toward the end, but lumenstone-hunting is still a challenge. The most I've ascended with so far in this version is 12.