Magnate's new egos

[Magnate]

Can you tell me what the devteam consider the current object distribution problems to be, or point me towards a thread or ticket that describes them?

Interestingly enough, I couldn't - they'd never been summarised in once place. So I wrote a starter for ten, though it's late and it's by no means comprehensive. I hope others will contribute, and if not I'll do a trawl of recent threads for the next version.

[Antoine]

We froze 3.3.0 and did a lot of fixing and polishing before it was released. We've released 3.3.1 with bugfixes, and are right on the cusp of releasing 3.3.2 with some more (and as the release manager for all those, let me tell you it isn't a whole lot of fun cherry-picking bugfixes, building and testing).

I think you've done a great job in terms of improving the dev process for 3.3 - feature freezes, retrospective bugfix releases and etc. I thought things were looking good for 3.4 as well, but if 3.4 sees a proliferation of new ego-items, then I think that will show that the new dev process still allows for too-radical changes.

I went back and read the thread Timo started and was surprised to recall that you were very supportive of the proposals at that time.

I was supportive of your dev process changes. But I also said you should
"- think very seriously before making a gameplay change which makes the game easier (ego-DSMs, offweapon art +blows/shots)"
"- think very seriously before changing an aspect of gameplay that has been around for a long time (pointy penalty)"

A bottom-up redesign of egos was the sort of thing I had in mind here.

But that's just my view

A.

[Antoine]

we know quite a lot about what was wrong with 3.3's distribution (as it's been roughly the same since 3.1.0

Can you tell me what the devteam consider the current object distribution problems to be, or point me towards a thread or ticket that describes them?

A.

[Magnate]

I think so too, but I think we end up agreeing to disagree on the pace of V development (it wouldn't be the first time).

It's funny you should say that. I went back and read the thread Timo started and was surprised to recall that you were very supportive of the proposals at that time. We froze 3.3.0 and did a lot of fixing and polishing before it was released. We've released 3.3.1 with bugfixes, and are right on the cusp of releasing 3.3.2 with some more (and as the release manager for all those, let me tell you it isn't a whole lot of fun cherry-picking bugfixes, building and testing). The rest of the proposals involved "introducing major changes earlier in the development cycle" so they can be properly tested (and if necessary reverted) before the next release. You now seem to be saying that major changes shouldn't be introduced at all - or maybe you were just operating with a different definition of "major" at the time. But look at how many suggested changes to ego items I *didn't* make!

[Antoine]

I'm sorry, I thought we were exchanging views quite politely and successfully.

I think so too, but I think we end up agreeing to disagree on the pace of V development (it wouldn't be the first time).

[EDIT] I won't be surprised if Timo makes some similar comments.

A.

[Nomad]

You aren't supposed to find (+0, +0) weapons deep in the dungeon either - that wasn't my intention, and means there aren't enough high-level affixes which boost hit and dam. This is the problem with Nomad's suggestion of reducing the number of hit/dam affixes, btw.

Given that the affixes are numbered, is there any issue with using duplicate names? i.e. Could you have an "of Damage" that gives 1d5 dam that's rated "good", and another "of Damage" giving 6-10 dam that's rated "great"?

Or could you somehow give affixes an alias - maybe have some sort of pseudo-theme with a name that's automatically used if the item has any one of a possible set of affixes? So all the hit/dam affecting affixes would show up named as an ego "of Slaying". (Perhaps combined with the suggestion fizzix made of showing the actual affix details in the item description.) Either of those would help reduce the flood of unfamiliar names required to specify many essentially similar hit/dam boosting affixes.

Another, less ideal solution to the lack of hit/dam might be to enforce a random boost to both in high-level themes (with a possibility of that boost being 0), or include it as an aspect of other affixes, i.e. a high level weapon affix comes with built-in hit/dam bonuses of 0-5.

Beyond that, I think flavour-wise material-based affix names (and, I'm afraid, blandly descriptive ones), are probably less intrusive than other types: an "Iron" Sword stands out less than a "Keen" one, an ego "of Sustain Strength" blends better than a "Robust" one. Go with clean and simple at first, change it up later when somebody inevitably complains about how all the boring old affix names are totally lacking in flavour.

[Magnate]

To Magnate the baffled

I think the problem is that you fundamentally don't believe in cautious, incremental development of V. That being the case there's not much point in me suggesting ways in which you could make change more gradual.

I'm sorry, I thought we were exchanging views quite politely and successfully.

So your answer to my previous questions is essentially "I don't think you should add anything new because I think change should be really cautious and incremental".

That's fine - it's a well-trodden debate that takkaria started on usenet in about 2006 and was thrashed out again only a few months ago in a thread started by Timo about the pace of development. I don't need to repeat my views again, and I'm happy for us to agree to disagree.

If there are any *other* reasons not to allow the new object generation algorithms to generate stuff that couldn't be generated by their predecessors, let me know. (Personally I think I have some misgivings about allowing of Warding to be generated on non-theme items, and a host of other detailed criticisms - just because I published it doesn't mean I think it's perfect.)

[Antoine]

To Magnate the baffled

With regard to your puzzlement about my suggestion "that 3.4 should change the range of items as little as possible"

I think the problem is that you fundamentally don't believe in cautious, incremental development of V. That being the case there's not much point in me suggesting ways in which you could make change more gradual.

A.

[EDIT: This is not meant to be an unfriendly or unhelpful response]

[Magnate]

I would strongly support that. I don't see the benefit of balancing 3.4 based on its similarity to 3.3 when we know 3.3's object distributions are dicey.

As were 3.2's, and 3.1's. It's a big ask, but hopefully someone will be motivated to do it.

That said, we know quite a lot about what was wrong with 3.3's distribution (as it's been roughly the same since 3.1.0), so even if nobody backports the stats code, we can still make sure 3.4 is *better* than 3.3, even if it's not perfect.

[Antoine]

On your other point, the stats module will facilitate comparison with 3.3, but we don't have stats for earlier versions. But some devs have discussed backporting the stats code to 3.0.x precisely so that we can make these comparisons.

I would strongly support that. I don't see the benefit of balancing 3.4 based on its similarity to 3.3 when we know 3.3's object distributions are dicey.

A.

[Magnate]

I should add that this is a really great coding and design effort, so hopefully there is some way to proceed.

Thank you. I'll be the first to say I'm not the best person to decide on flavourful names for things - there's lots of improvement to be done there. Also, fizzix's stats are an excellent illustration of how this can be balanced - we can adjust the T: lines in ego_item.txt so that SI appears earlier, slays are less common, etc. It's fiddly work but it's iterative - large improvements come quite quickly then fine balancing takes ages.

[Antoine]

Antoine. What if the name doesn't change but the description does? So it's not called a "sharp broadsword" but it has a description line oon the Inspect screen that says, "it is sharper than usual"? Is this still a flavor break?

For me that's fine, an improvement on the status quo even.

A.

[Magnate]

Yes that's correct. For example previously after a certain depth you could not find weapons with (+0, +0). Now roughly about half of the weapons deep in the dungeon have no magical bonuses.

On the other hand, there are a lot of weapons now with extra dice or extra sides. I am not able to check most of these subtleties currently. Only very broad things. The only real flags to me are that SI is so rare early and that extra blows seems to be very common.

Well none of this was intentional - I did say very early on in the thread that my effort was focused on making it work without crashing or other painful bugs, not on ensuring perfect balance. So let's by all means make SI more available, slays/brands/blows deeper and rarer. This is a matter of adjusting the T: lines in ego_item.txt, by changing the affix level, the commonness, and/or the min/max depths.

You aren't supposed to find (+0, +0) weapons deep in the dungeon either - that wasn't my intention, and means there aren't enough high-level affixes which boost hit and dam. This is the problem with Nomad's suggestion of reducing the number of hit/dam affixes, btw.

Speaking of the naming problem (Nomad's suggestion was to reduce the number of affix names), I'd be really really happy if someone came up with some better names for any of the affixes. We do need a lot of names, but I've always thought that was part of the flavour.

[fizzix]

It's a little worrisome that you say branded and slay items are 75% more common, that seems like too much but I think those could also be weapons of slay orc (+0, +0) correct? So even if the slays and or egos are more common there is a greater power variation between items with the same flags, right?

Yes that's correct. For example previously after a certain depth you could not find weapons with (+0, +0). Now roughly about half of the weapons deep in the dungeon have no magical bonuses.

On the other hand, there are a lot of weapons now with extra dice or extra sides. I am not able to check most of these subtleties currently. Only very broad things. The only real flags to me are that SI is so rare early and that extra blows seems to be very common.

[Jungle_Boy]

In general with affixes. Branded and slay weapons are about 75% more common
Westnernesse, Gondolin, Defender and HA weapons are about 33% more common.
edit2: To put this in context. The first level with a 1% or greater chance of finding a westernesse weapon in current V is about level 16. In Affixes, it's level 62!
Weapons with extra blows are about 4x more common with affixes, which probably should be verified and changed.

Also, my code is finding Morgul weapons once again. Is this by design? It's possible that it's buggy.

First, I'd like to say I like this idea in theory. I have not tested it but it seems like it should provide a smoother step between magic items and ego items and even between artifacts and egos which I think is a good thing.

It's a little worrisome that you say branded and slay items are 75% more common, that seems like too much but I think those could also be weapons of slay orc (+0, +0) correct? So even if the slays and or egos are more common there is a greater power variation between items with the same flags, right?