A Few Questions/Observations From an Old Player

[Oramin]

So after a certain point, additional Strength is meaningless for carrying purposes?

I'm 34.1 lbs overweight both at 18:110 and while wearing a RoS +5 (18:160). No changes in Speed either.

[kaypy]

The higher the score, the more grinding. So for a winning game, score is backwards as a measure of performance (Winning underlevelled is a better display of skill).

worst        lowest losing score
  .                  .
  .                  .
  .                  .
  .          highest losing score
  .          highest winning score
  .                  .
  .                  .
  .                  .
best         lowest winning score

[DaviddesJ]

What about keeping the scores based on experience but adding a penalty for dying and an extra reward for winning, so most of the time, winning scores will beat losing scores.

That still means that the winner who grinds or milks their results is going to score higher than the one who doesn't. Is that really what anyone wants? At least, the reward for winning should be higher the faster you achieve it, or there should be some other factor to compensate for just pointlessly increasing your score.

[Monkey Face]

What about keeping the scores based on experience but adding a penalty for dying and an extra reward for winning, so most of the time, winning scores will beat losing scores.

[DaviddesJ]

Yes, the law of averages does apply here, but the turn-count argument (I think) is not as critical to scoring as say gold-count, Max XP, mob kill count x Avg. Mob Level, .. if such existed.

I don't see it. The latter things are all easily manipulated by (e.g.) farming high-XP or high-drop monsters. And you could do that forever to get as high a value as you want if you aren't counting how many turns you play. It seems to me that measuring how much stuff you accumulate without measuring how long you take to do it is completely unrelated to how well you're doing in the game.

[Spacebux]

I wasn't thinking so much of a slowness penalty as a reward for efficient slaughter. For example, if 2 players rack up the same amount of experience, shouldn't the one who finishes in, for example, 5 million turns get a higher score than the one who finishes in 10 million turns?

So, basically, I was expecting a formula which factored in experience per unit time.

I don't think turn count necessarily equals "efficiency". One player might have had an extremeley lucky run with the RNG, got 3-4x as many great items / drops, or that many fewer run ins with great mobs throwing mana bolts &/or summoning pals. Yes, the law of averages does apply here, but the turn-count argument (I think) is not as critical to scoring as say gold-count, Max XP, mob kill count x Avg. Mob Level, .. if such existed.

[Oramin]

Two differents sources of x3 multiplyer can logically produce x3 if the dont stack, x5 if bonuses add or x9 if they multiply, I cannot imagine any way to get x6. Just blindly adding values is definitely wrong, since it will result in x1 slays adding up.

I made a thread specifically for this discussion; please take it over there.

[MattB]

Two differents sources of x3 multiplyer can logically produce x3 if the dont stack, x5 if bonuses add or x9 if they multiply, I cannot imagine any way to get x6. Just blindly adding values is definitely wrong, since it will result in x1 slays adding up.

Ooooh yay! I was wondering when someone was going to reignite the launcher multiplier debate!

[LostTemplar]

Two differents sources of x3 multiplyer can logically produce x3 if the dont stack, x5 if bonuses add or x9 if they multiply, I cannot imagine any way to get x6. Just blindly adding values is definitely wrong, since it will result in x1 slays adding up.

[DaviddesJ]

Once upon a time I think Angband did have gold scoring, presumably inherited from Rogue where it made (slightly) more sense. I suspect I hold the all-time world-record score in Rogue 5.3, although the servers are long gone.

[Derakon]

Also, don't Winners get a bonus? With that formula, somebody could kill a bunch of monsters and die to Sauron/Morgoth but still have a higher score than somebody who actually wins the game.

No bonus for winners. I've had exactly that scenario happen to me.

Honestly the takeaway here is "Don't pay too much attention to scores."

[Oramin]

I wasn't thinking so much of a slowness penalty as a reward for efficient slaughter. For example, if 2 players rack up the same amount of experience, shouldn't the one who finishes in, for example, 5 million turns get a higher score than the one who finishes in 10 million turns?

So, basically, I was expecting a formula which factored in experience per unit time.

Also, don't Winners get a bonus? With that formula, somebody could kill a bunch of monsters and die to Sauron/Morgoth but still have a higher score than somebody who actually wins the game.

[Raxmei]

I'd have been surprised if speed did play a factor. The scoring just tells you how much dungeon you've murdered your way through. Experience as a measure of challenges overcome and depth as a general indicator of progress does that very simply. The practice of speed play wasn't ascendant back when I started playing to the point that penalizing score for slow cautious playing would have raised complaints.

[Oramin]

Hmm, thanks. Guess if I want to run up the score I should Grond several levels of monsters.

I'm a bit surprised that speed doesn't play a factor.

[Derakon]

Hm, looks like points are actually pretty simple; in score.c the function total_points() just returns (max experience + 100 * max depth).

So there you go. Experience and max depth, that's it.