look at my dumps

[PowerDiver]

I haven't submitted a dump in forever, but if you've never seen you ought to look. Go to

http://angband.oook.cz/ladder-browse.php?e=PowerDiver

click on "updated" to get them in time order and browse from the top.

Read the comments. That's where the actual info is.

In particular, Fuial is my crowning achievement.

[Antoine]

I haven't submitted a dump in forever, but if you've never seen you ought to look. Go to

http://angband.oook.cz/ladder-browse.php?e=PowerDiver

click on "updated" to get them in time order and browse from the top.

Read the comments. That's where the actual info is.

In particular, Fuial is my crowning achievement.

Remind us why you left?
A.

[PowerDiver]

Remind us why you left?
A.

I fought and fought and fought. To no avail. The game kept getting worse. The developers thought the insights of the best player in the history of the game were irrelevant. The thought the desires of newbs(*) were far more important. You know, ask the children if they want more sugar in their diet. They are the ones eating it after all.

V [by which I mean 3.0] deserved a champion, and I was the only one fighting for it, so I struggled on, alone, in an unending hopeless battle. I don't know how long I would have continued. I was already close to giving up. My involvement here was steadily diminishing.

Now this post gets *dark*. You are warned.

In April of 2012, my wife was diagnosed with breast cancer. She fought and suffered, and died in January 2016 at the age of 53. You are not supposed to compare these things, but I do anyway, and I understate when I say her experience was worse than average among breast cancer victims, and I was along for the ride. We had no kids, together since college, she was my whole world.

Once she was diagnosed, I couldn't continue any activities that might make me more aggravated, so leaving here completely was automatic.


DO NOT EXPRESS SYMPATHY


Hopefully, you are incapable of understanding my situation, not even a tiny bit at all. Your words are meaningless, which shouldn't be a problem. However, everyone in my position has multiple stories of people who think they are being supportive saying things so hurtful the choices are flee or manslaughter. No upside, huge possible downside. Do not express sympathy to me. Do not try to be supportive. Do not mention any bright sides. There aren't any.

A month ago, I would not have been open to returning. I was spiraling towards doom. Some days, it seemed the only reason to get through the day was that it would be rude to die before my mother. One day I realized if I was going to survive I would have to reinvent myself. That was October 1, after midnight, maybe I should call it October 2. Luckily or unluckily for you guys, it was after that when someone from long ago politely asked me to return, so I did.

There's a fair chance I will become toxic to the community, but before I started posting I told Nick he should feel free to ask me at any time to leave and I will.

I've decided as a life choice to be frank and direct. Don't ask me any questions if you don't truly want the answers. I will not dissemble, and I have a lot of animosity towards people I think ruined a game I loved.

My wife would have said "Get over it, already." She'd tell me it was stupid to harbor grudges. However, I'm likely to be an angry, bitter, nasty person for quite a while longer. I have to re-edit every post. I set my first goal posting here to be "adequately polite", but that doesn't mean I'll succeed at it.


That's the short answer. Eventually I'll post the long answer. I hope it will be therapeutic to write. The unfinished outline is already longer than this post, so it may take a while.


(*) I know someone corrected me about "noobs", but in a world that changes word meanings all the time, I stick to the classics. That some person spelled something phonetically because they didn't make the connection to "newbie" is no reason for me to follow suit even if afterwards everyone piled on. It should be clear by now I'm not someone who follows the crowd.


In case you forgot already:

DO NOT EXPRESS SYMPATHY


OK, I won't be a complete nazi about this. You can make remarks and ask personal questions without fear I'll put you on ignore forever. But please keep the point in mind, and restrain yourselves. If you say something you realize is inane trying to be polite or civil or whatever, keep in mind it sounds 10 or even 100 times as inane from the other side.

[Derakon]

I just want to thank you for being honest and straightforward. We may not always see eye to eye, but it's only through disagreement that we can hash out the really difficult decisions, the ones where people may not even realize something is wrong.

For what it's worth, regarding the game I believe the general developer stance has been largely to try to remove things that were "no-brainers". All the talk about traps, to take one example, is because 99+% of the time, players just spam one of their sources of unlimited trap detection every so often and then play on as normal. "Removing trap detection" was an attempt to address this ritualistic formality, to replace it by meaningful choices. Of course it also changed the class balance because now warriors did not need to (indeed, could not) carry trap detection items, thus easing their inventory pressure. The aim was to get all of these "systemic rituals" thrown out, and then to address balance. We're midway through dealing with the rituals; balance hasn't really gotten a serious look yet.

3.0 was the culmination of a good 15 years or so worth of very subtle tweaks and refinements. I think it's entirely fair to consider it to be the peak of "old Angband", and I mean nothing derogatory about calling it "old". The problem we ran into was that we had nobody willing to maintain such a mature game without making their own tweaks to it. So instead we got maintainers who experimented. And as you note, balance is very easily disrupted, so those experiments rapidly "loosened" a balance that had been gradually tightened over the course of those 15 years.

I do believe that the framework underlying "new Angband" is better than the framework underlying "old Angband". Many systemic mechanical problems with the game have been corrected, making balance tweaks is easier, and the potential scope of those tweaks is far larger than it used to be. Not to mention major strides in the cleanliness of the code! What we need, once the current round of features is completed, is a serious look at the game balance -- where we are, where we used to be, where we want to be, and how to get there. You're absolutely right that we have erred on the side of making the game easier for almost every change in the past several versions (though at least we're nowhere near the brokenness that was 3.2). My hope is that what we come up with for Angband 4.0 is something that has an accessible "default difficulty" that is still tightly-designed, and has scope for being scaled via game options to anywhere from "easy enough for a young child" to "hard enough for Eddie".

(the above should not be taken as any kind of desire to lead or even necessarily participate in the balancing effort beyond as a member of the peanut gallery )

[Thraalbee]

Thanks for sharing! Not that I agree with everything, but my recent wins are all randart ironman games and now even that challenge feels too predicable to make me start a new game. Sil is more fun since I can easily challenge myself far beyond my limits but a bit limited. A more difficult Angband with today's gui improvements would be great.

On the personal side: I'm ok now but 12 years ago, being the one getting the radiotherapy treatment instead of standing on the side was a blessing.

[Sky]

and the answer is ... because angband is becoming a single player game.

it's no longer a game run on a server where hundreds of people play, and getting the high score is the objective of the game. the objective of the game is now to win, and the competition is .. games released today.

just because modern games are easier, it doesn't mean they are less fun.

and a game that requires you to have an encyclopedic knowledge to progress is no longer seen as a plus; the time you dedicate to angband will not translate into, say, familiarity with computers which could one day make you money. there is no longer any excuse to devote so much effort to angband.

fun?

maybe it's fun for you, because you are - and i quote - the best player the game has ever had. you are so high in the clouds, you forgot what it's like to experience angband as a generic player.

i'm sorry if the new guy needs to tell you this, but it's true.

[Pete Mack]

Eddie-
Very glad to hear from you, regardless of the terrible circumstances. (That is all the sympathy I will offer, in correspondence with your wishes.) If you need a place to vent, it won't disturb me, at least. You could never be as disruptive or annoying as neo. (Note: not intended as a dare.)

I suspect there are some players here that will challenge your mettle. Even in the current version of Angband, 57k player turns is an impressive win. And Sil is a surprisingly good evolution. Really, do try it-the current comp, for example.

A few points (about the dev branch)
Traps are much nastier than they used to be. Sure, there is a brief time at the start where detection is easier. But there is no longer any avoiding them, since they always appear at strategic locations in hallways. (Paladin is currently a nightmare because of this.)

Fuzzy object detection I don't particularly like. But makes the game harder, and largely makes up for the increased ego item drop, which itself is much reduced from 3.5: The Unbalancing. That's pretty much where I quit working on the engineering deficit, which I felt was 3.0's greatest weakness. Thanks to Nick and others' yeoman work, the code and edit files are vastly cleaner than 3.0. This should make it easier to hack.

Rune based ID would fit very well with your (and Sil's) notion of experience. I agree that you should not get more EXP for multiply redundant kills, though it's not clear that killing an ant and an ancient dragon should yield the same exp.

I don't think the quiver is so unbalancing as all that. The real imbalance was from massive archery damage from brands, which has been largely but not completely nerfed.

The one recent change I now think mistaken is breath attenuation. At the very least, more classes of breath should be beams. (Basically everything except poison and sound)

[Nick]

Derakon has summed up the situation perfectly.

[Pete Mack]

Sky:
Make angband too easy and replayability goes in the toilet. I stopped playing because 3.5 really was boring. Sure 3.0 is was unaccessible, largely because all the help files and user's guides gave bad advice.

Yes, it is now played on PCs not timeshare systems. But thanks to Pav, there is still a ladder, and thanks to debo there are still competitions. I think you are wrong in both the specific and the general.

[AnonymousHero]

@PowerDiver: If I might ask a couple of (perhaps) seemingly silly questions...

#1: how many games have you played, do you think? (Obviously, I'm talking ballpark here.)

#2: how many times did you get to *extremely* low turn-counts before facing the big P?

Obviously we can all look at the dumps, but if we are to learn anything from the recent replication crisis in science circles, it's that "publication bias" is a huge deal, so I'd like to get a sense of the real numbers -- and not the numbers just based on your "published" characters. EDIT: (because I'm just now realizing...) I'm not insinuating anything here about you purposefully omitting to publish characters that didn't make it very far. It's just that a) nobody tends to publish characters until they make it past, say, level 20, or b) find a significant game-changing item at an earlier level.[1]

Apologies if you're already specified these things somewhere, I haven't really read all of your recent posts fully.

@Derakon + @Nick: +1

[1] ... which is why I really think we should have (by default) auto-uploaded statistics for all levels generated, options chosen/altered/etc. It's not that "metrics" magically makes all choices about how a game should be easy, but knowing what the game is will certainly make "should" decisions more informed.[2]

[2] (Doin' the David Foster Wallace thing here, but: ) For example: It would be really interesting to know how full a typical player's inventory is. If we had really detailed stats we could even figure out what the distribution was, what the most "kept" items were, if there was a "bimodal" thing going on wrt. whether it was new players who tended to keep more "junk", and whether more experienced players (#games as a proxy?) tended towards not even picking up things that would eventually be useless, how option:ironman (etc.) influenced these things, etc. It's going to be a hell of a lot of data, probably filled with a lot of false correlations, but a simulated world is effectively almost as complicated as a real world, so it's not that surprising.

[Antoine]

Eddie, do you have a mission here (beyond 'talk to people about stuff')?

[the Invisible Stalker]

The developers thought the insights of the best player in the history of the game were irrelevant.

Your self-description will probably seem like an exaggeration to people who don't remember the old days, but I'd say it's a fair characterisation. You completely changed the way a large number of people, including me, play. Even though I agree with most of your complaints about modern vanilla, I have to say there's something to be said in general for discounting the opinions the opinions of the best players. FA is, in my opinion, one of the best variants out there. Nick is reasonably good as an FA player, but hardly great. I suspect those facts are related.

[Pete Mack]

Your self-description will probably seem like an exaggeration to people who don't remember the old days, but I'd say it's a fair characterisation. You completely changed the way a large number of people, including me, play. Even though I agree with most of your complaints about modern vanilla, I have to say there's something to be said in general for discounting the opinions the opinions of the best players. FA is, in my opinion, one of the best variants out there. Nick is reasonably good as an FA player, but hardly great. I suspect those facts are related.

The thing is, he's right. Angband 3.2-3.5 had poor replayability because they really were too easy once you had mastered them. One reason I left back then was that I got bored with it. A better way to make it noob friendly is to add Maia class, not reduce replayability. Note how many old time players are playing Poschen now because V is too easy. Even removing CHR had the effect of making it too easy to buy _Tele self and making kevels 20-30 trivial for diving. Nearly every change made turned the game easier, except fuzzy object detection which I think was a sub-optimal fix in that it makes the game harder in part by increased grinding. Better to do it by making teleport expensive.

Edit: Further, the one recent game I enjoyed by far the most was because I didn't find more than 2 !CON before dl 80, and no ESP until level 72. It meant at one point I was deep in the dungeon with base speed +3 or +5, not because I didn't have good speed items--I did--but because I was wearing up to +20 CON in gear to avoid getting one-shotted by an awakened pass wall monster. In 3.0, situations like this were much more common.

[fizzix]

The thing is, he's right. Angband 3.2-3.5 had poor replayability because they really were too easy once you had mastered them. One reason I left back then was that I got bored with it. A better way to make it noob friendly is to add Maia class, not reduce replayability. Note how many old time players are playing Poschen now because V is too easy. Even removing CHR had the effect of making it too easy to buy _Tele self and making kevels 20-30 trivial for diving. Nearly every change made turned the game easier, except fuzzy object detection which I think was a sub-optimal fix in that it makes the game harder in part by increased grinding. Better to do it by making teleport expensive.

I kind of want to put some of my opinions out here, since I was part of the devteam during the 3.1-3.5 era.

The problem with 3.1-3.2 were that the drop rates were all messed up and it was way too easy to get good gear. This was a side effect of the war on junk which took a while to be corrected. I think 3.4 and 3.5 were a lot better because we cut down on the probability of getting obscene drops. I know this because I generated statistics and rebalanced drop rates for artifacts and most other gear to be in line with 3.0.6. I believe these balance changes kicked in on 3.4 or thereabouts. If you think 3.4 is more generous than 3.0.6, you are wrong. What is true is that 3.0.6 generated about 5 times the total amount of items, just a lot of it was useless, so you were required to sort through all that junk. That often meant you missed the good gear because who has the time for that. You can think of 3.4 as auto-squelching all the useless gear. For the rest of the changes, the only major one was the quiver. There is an argument that this is a significant enough effect to offset all the other ones, I admit that.

There *were* other changes that made the game harder, and we shouldn't gloss over them. Teleport other used to be a beam, and there were long posts decrying how it would be awful if it was a bolt. Destruction used to leave artifacts on the ground, creating an easy way to get the best loot from vaults. Pits used to be on every level and they only had a minimum cap, so you could get orc and troll pits at dlevel 99, which were often free potions of stat gain/exp (especially with the obscene item drop rates of 3.0.x). Ammo used to get multiplicative slays, so a x2 bow with x3 arrows would do x6 damage instead of x5. Ammo is still too powerful IMO, but I think the solution has more to do with combat balance than with removing the quiver. If there's one thing I vehemently agree on with Eddie it's that tweaking the details of combat (and other details like item drop rates, monster details etc.) is an extremely difficult and extremely important aspect to the design of angband.

The main problem with angband design and development from my pov is the same as with any game. If you play it enough you will eventually have seen enough that you know what to do in every situation. The game then becomes routine. Veterans who still want to play have to create their own challenges. The game furnishes some challenges for you (ironman etc.) but there will come a point where even that is too easy. It's hard to remember back in the day when you struggled over whether to use a confusion wand or to teleport away from a dangerous monster, but we were all there once. If you design a game for super-elite players it becomes a dead end. If Angband is difficult enough that you are required to make the correct decision between 3 !CMW and 4 acid arrows, then it's only going to be enjoyable for the very few people who essentially know every detail about the game. It will be great for those people, but horrible for everyone else. BTW this is exactly the role for variants games like PosChengband. They're supposed to provide new excitement for the players that have already gotten what they could out of vanilla. It is also why there's so much effort to allow players to tweak things in the edit files. *You* can make the game harder.

When I look at changes like conical breaths, I see something that opens the door for other things. In the current incarnation it probably makes things way too easy, but what it allows is the ability to stay in LOS of 2 or more powerful breathers without having a 1% chance of death. Once you have that you can create situations where tactical decisions can be more interesting. With ball (or beam) breaths, if you are within LOS of two powerful breathers, you immediately had to remove one of them, or leave. The chance of death on the next turn was just too unacceptably high (1-2%). This reduced tactical decision making, since the decision was forced. (And if you want to say, well let's limit the escape opportunities so you have to accept that 1-2% in some situations, that doesn't work either. Angband allows you to know what's ahead of you so the game just becomes a matter of spamming detection and avoiding anything remotely dangerous.) If we make no additional changes besides conical breaths it will be a bad change. If we limit the players ability to remove monsters or escape, it's a better change. In the transition period it looks terrible.

One last point. Angband gameplay has always been more tactics focused than strategy focused. What I mean is that it's a lot more about what decisions you make in the immediate situation and less about what long term decisions you make. And I think for a lot of players this is a draw. You don't have to worry about choosing a bad character build, or making a poor decision in the early game that removes your win conditions. The one major strategic element it did have is inventory management. And I can see why someone who likes strategic gameplay would decry anything that made inventory management easier. What I wonder about is why someone who love strategic gameplay is playing Angband and not one of the many other roguelikes that provide skill trees, or other similar instances.

[PowerDiver]

@PowerDiver: If I might ask a couple of (perhaps) seemingly silly questions...

#1: how many games have you played, do you think? (Obviously, I'm talking ballpark here.)

#2: how many times did you get to *extremely* low turn-counts before facing the big P?

#1 I have no idea. I was stock trading, at home, Pacific time zone, with a wife who often came home late because she was a top engineer. Lots of free time in the afternoon / early eve. At a guess, many hundreds over 15 years.

I mastered half a dozen or maybe a dozen variants, by which I mean wins with different classes and enough insight to offer patches if they were still under development.

E.g. I had heard S was good for id-by-use, which to me means no identify scrolls at all. It was close enough that the fact it wasn't was a horrible tease. So I modded the code and sent my changes to, I forget, Leon? He used about half my changes, I think. He also changed the fundamental aspect of the game, how much experience costs change as you increase the number of different skills, because of how I played.


#2 By the standards of the community, the vast majority of my games are ridiculously fast reaching M. I go down, and then I fight him. If I lose, I try to figure out why so I can do better the next time. I don't waste my playing time clearing pits that are difficult. Better to spend the [real] time chasing down high value targets.

The above is in a sense no longer true. For my modded stuff, I was playing as maintainer rather than player, clearing levels and looking for bugs.

I beat Moria a few times before I got bored, and I never died to an AMHD. I'd heard how scary they were, and I was confused. If you see one, run away.!?! I later decided they must be deadly to scummers, and everyone else was a scummer. Just a guess of course. I wasn't sophisticated enough to be formulating strategies at the time. I just played the way I played.


When I arrived in rgra, the advice was *universally* bad. Horrible. I simply didn't understand why people played the way they did. I read the help guides. I listened, trying to find signal in the noise, and I couldn't.

I tried to give advice on rgra, and just aggravated people.

I wrote the Tales of the Bold because I couldn't give advice that was useful. I couldn't understand what people were saying. They couldn't understand me. I was unable to figure out specifics, so I figured if I wrote *everything* maybe they could find what they needed.

By the time the community ever caught up with where I was, I had moved on already. Those old turncounts wouldn't look so low now, but at the time some people insisted I must be cheating.


I think your questions are misguided. Before the game was nerfed, I didn't want a win rate over 50%. I'd rather play sports with my peers than beat up on 8 year olds. There's also meta-strategy. The best strategy to win the game you are playing today is not the best strategy to win the most games ad infinitum. For example, I played tennis just well enough to be the weakest player on a poor high school team. I had a coach who wanted me to change my serve. If I was playing one match for something truly important, it would have behooved me to use my old serve. However, even in school matches, I used my new serve, as the coach wanted. That was better for winning in the long run.


I guess I'll share something else. My wife was a skilled player in her own right. But she was OCD about level clearing. Occasionally she'd have a game where she would complain she was descending too slowly, playing the same level over and over and over, and I'd tell her to take the stairs down before recall. She just couldn't do it; if she went down she'd have to clear the next level. There were things she didn't want to learn, and she'd ask me questions from time to time about which weapons to keep. If you grant her that, she still wasn't as good as Timo but she wasn't so far behind. She absolutely refused me when I asked to post her dumps. She didn't want to be mentioned at all.

Every change I recommended for V was tested by both of us. When I said that I knew how a change affected level clearers that's because I had the best such play tester available.

My wife was like Timo in another respect. She'd playtest for me to be kind, but she wanted to play the "official" version. As the game got nerfed, she lost interest and stopped playing. That's really pisses me off still. Really really really pisses me off.

I have to move on. I'll release my code to the dev team with no conditions, just plead that they don't cut and paste. I wrote it at a bad time, and I am truly embarrassed by how wretched the code is. Multiple fireable offenses. Long long ago, in a completely different situation, I gave some guy a sim I'd written quickly with his promise not to share it. It got out, propagated, and five years later I was still getting grief. I made a solemn vow never to release bad code again. However, now is the time for new beginnings, so what the heck.