Gold and no_selling

[takkaria]

Thank you. What that tells me is that for the first FORTY dungeon levels the drops have hardly changed at all, which is completely contrary to what people are saying.

This was the average before the infinite variance stuff was put in, which accounts for the difference, I think.

[takkaria]

How was that data collected? It'd also be good to see a zoom-in on the first 30 levels or so; the Y scale's gotten flattened because of the high upper end. I know for a fact that I never saw single-digit treasure drops in the old days (and I'm pretty certain that getting less than 20 gold from a drop was a rarity), and now I see them all the time early on.

I made a spreadsheet (no idea where it is now, sadly) that used the game's own function and plotted it. It was quite complex.

Here's another graph of the old gold distribution I found: http://rephial.org/research/cur_gold_drops.pdf. This one was made by running 1000 of each 5 dungeon levels and plotting the amount of gold found if you killed all monsters on each one. Not sure what it looks like now, but I guess fizzix does.

[PowerDiver]

http://rephial.org/research/avg_gold_drop.pdf should tell you a lot of what you need to know

I simply do not believe that. There's got to be something wrong with your method. Or a bug in money stacking, or something.

In 3.1, I got close to squelching money even playing non-ironman. I observed that money I picked up was insignificant compared to what I got from selling. I only left it unsquelched for flavor, and stopped going more than a couple of keystrokes out of my way to pick it up.

When I played 3.0 to test how much harder it would be with no egos, so I was level clearing, I was reminded over and over how much more money was available.

[Timo Pietilä]

I simply do not believe that. There's got to be something wrong with your method. Or a bug in money stacking, or something.

Probably the fact that deep levels contain a lot of monsters with ONLY_ITEM counts a lot in that. Money drops are insignificant because you don't kill monsters with money drops.

[PowerDiver]

Probably the fact that deep levels contain a lot of monsters with ONLY_ITEM counts a lot in that. Money drops are insignificant because you don't kill monsters with money drops.

I'm thinking about the second trip of the game, say from DL15 to DL30. It wasn't quite so bad in my current 3.2 trial. Perhaps the egos or ego quality have been toned down or it could be variance.

[Magnate]

I'm thinking about the second trip of the game, say from DL15 to DL30. It wasn't quite so bad in my current 3.2 trial. Perhaps the egos or ego quality have been toned down or it could be variance.

Does the reference to egos mean that your current trial is with selling?

[Magnate]

This was the average before the infinite variance stuff was put in, which accounts for the difference, I think.

Yes I suspect so - so as soon as d_m sorts that out we're good to go, I think.

[d_m]

I think part of the problem is unlimited range plus skew (http://en.wikipedia.org/wiki/Skewness). The problem (anecdotally, I haven't looked at the code) is that the mean and variance are both fine, but that the data has such a positive skew that over a relatively small number of samples you're very likely to be under the mean. It only averages out if you play long enough that you're quite likely to hit one or more "lotteries". Having a median below the mean is a typical indication of skew.

Similarly, the average/mean return on actual real-life lotteries can be as high as 60%, but you're very unlikely to reach the mean or even close to it, unless you play thousands of times (or, depending on the lottery, millions).

Colbey! Good to see you on here again.

Thanks for the link, it looks interesting... I may try to compute the skew for the current method and see what I can do to improve it.

[PowerDiver]

Does the reference to egos mean that your current trial is with selling?

I haven't tried 3.2 with the no selling option yet. At less than a 10x multiplier [as mentioned before, you can build up to it gradually, not 10x town drops], I cannot imagine it would be anything but aggravating. If I want a challenge game, I'll play ironman or no artifacts or something like that.

I wonder if the folks collecting stats are extracting chest drops. Early on, those tend to be money. Not that they matter much.

I tested a 3.0 warrior clearing DLs 1 to 10. He collected a bit over 4K, grabbing all gold drops, but missed part of DL8 to a trapdoor. That surprised me, since my 3.0 mage who did not collect all money got near 6K. Perhaps it is related to spawned monsters, with the mage needing much more rest so killing more spawned monsters.

The 3.0 warrior found too many good [no ego patch] weapons and wands and staves to keep track of in that time, as well as average stuff I paid no attention to. At a guess, 15 to 20 each good+ weapons and magic shop sales. Playing with selling, I'd have made 3 trips to sell the stuff if I was clearing levels.

[PowerDiver]

So I just played a warrior clearing levels in 3.2, down to DL10 and then recall.

He collected 914 gold. I lost 20 or 30 to Smeagol the first DL I saw him. Compare that to the 4K to 6K range I saw in 3.0.

Then, I sold all of the stuff I would normally sell. I had to buy 3 ?id to do it, and ended with 13981. So for the same buying power, no selling would have needed a factor of 15 to break even in 3.2 pricing scheme on this particular experiment. I think the comparison number goes significantly higher in the depths, where for example a mage picks up all magely dungeon spellbooks without noticing or worrying about slots.

Now perhaps there is too much money in the game. Perhaps all sales prices should go down or something. IMO that should be a separate discussion.

For now, I'd suggest a multiplier of around x12. Money would be a little tighter, but not too much IMO. Then adjust it if people think the gameplay needs a different value. If you slowly scale up the multiplier, it might be better to just go with x15.

[buzzkill]

So I just played a warrior clearing levels in 3.2, down to DL10 and then recall.

He collected 914 gold. I lost 20 or 30 to Smeagol the first DL I saw him. Compare that to the 4K to 6K range I saw in 3.0.

Then, I sold all of the stuff I would normally sell. I had to buy 3 ?id to do it, and ended with 13981. So for the same buying power, no selling would have needed a factor of 15 to break even in 3.2 pricing scheme on this particular experiment. I think the comparison number goes significantly higher in the depths, where for example a mage picks up all magely dungeon spellbooks without noticing or worrying about slots.

Now perhaps there is too much money in the game. Perhaps all sales prices should go down or something. IMO that should be a separate discussion.

For now, I'd suggest a multiplier of around x12. Money would be a little tighter, but not too much IMO. Then adjust it if people think the game play needs a different value. If you slowly scale up the multiplier, it might be better to just go with x15.

x15, or really anything beyond x4, seems ridiculous to me. Consider the following if you haven't already.

I've had a few drinks, so bear with me.

Up till now, most have always been playing with selling. If you don't change your play style at all of course you'll end up with less gold. Players need to change too, not just the game.

Imagine if up to this point selling was never an option to sell and suddenly the ability to sell was incorporated. If you didn't change your tactics you would also end up with less gold, because you wouldn't lug stuff back to town (that would be a change in behavior) and GP drops would be less.

Don't put the full impetus of change on the game. Play style will change over time, esp if no_selling is the only option (as I understand it is the ultimate goal). If you do, then the over-correction (too high multiplier) will in effect be an exploit. Furthermore, smaller drops to start with will encourage the change to play style that is required to be successful in the early stages of a no selling game, rather than just doing what you always did with selling.

When I play with selling, I sell stuff. When I play without it I'm fully aware of the fact that there are some differences in the two games. To ignore them is to be a fool.

Admittedly, I haven't played V since 3.1.x. I just don't see how there could be that great of a variance V and the variants I have played. Maybe gold with selling_on was always too abundant, at least that's how I've felt ,so maybe I have a bias.

[Derakon]

The question is, how much do we want players to be able to buy? With selling, players can effectively buy everything they might possibly want except for black market goodies after two dungeon trips. Without selling, even basic consumables are hard to afford. Presumably the desired balance is somewhere in the middle, especially if we expect no-selling to eventually be a newbie-friendly option.

[PowerDiver]

x15, or really anything beyond x4, seems ridiculous to me. Consider the following if you haven't already.

If you are willing to do x4 on 3.0 gold drops, you ought to be willing to do x20 on 3.2 drops. Read my numbers. If anything, that 914 total was above average for 3.2 money drops in my experience. I remember that a previous char was under 800, but didn't have a dump.

[ewert]

Colbey! Good to see you on here again.

Thanks for the link, it looks interesting... I may try to compute the skew for the current method and see what I can do to improve it.

All I can say is ... why? =P

I just can't see any sensible game design reason to have an option for near-infinite gold drop to happen in a game. That just breaks the buying game altogether for that game ... Probably ranks as #1 game design blunder currently in Vanilla. =P Sure, it almost never happens. Except I got 5k from a town drop with a new char, which sort of meant I could splurge on ccw etc. and broke the first levels of gameplay totally for that char.

[Twilight]

All I can say is ... why? =P

I just can't see any sensible game design reason to have an option for near-infinite gold drop to happen in a game. That just breaks the buying game altogether for that game ... Probably ranks as #1 game design blunder currently in Vanilla. =P Sure, it almost never happens. Except I got 5k from a town drop with a new char, which sort of meant I could splurge on ccw etc. and broke the first levels of gameplay totally for that char.

How is this in principle any different from the slight chance that you can get an incredible drop (say Ringil) at 50' if it passes OOD and rarity checks? A longstanding tradition, and one of the good parts of Angband IMO, is that there is a chance, however small, that you can get an awesome itemdrop early. The only slight change I'd make is to make this impossible in town to make it comparable to the itemdrops. Another example is an early potion of experience - sure you bypass the early game due to that, but you soon "catch up" with some diving to make the game challenging.

Those wow-feelings you can get from time to time more than compensates for bypassing some stages of the game. The game *is* self-regulating. If you are too powerful you can just dive some, even with oldschool playstyles.

Just my 50 öre