I'm going to give you a careful answer, because I know that you'll read it.
I'll agree with Bullgrit that the presence of the modules is useful when trying to discover some form of norm. It is not the only useful factor; the statements of the game designers are also useful.
In the event of extrapolating from a module -- any module, any system, any edition -- a track of "all that is possible", even modified by any arbitrary % of "what is likely to be gained" results in an arbitrary result. It is reasonable to suspect that there are factors not present in such an analysis that must be taken into account to gain worthwhile results from the module. These factors may not be present in all modules; where they are present, they may not always be the same factors. Only a deeper analysis can determine this.
<snip>
I do believe that a deeper analysis of the G series modules would be more helpful than Bullgrit's shallow analysis. Obviously, a deeper analysis of a wider range of modules would be more helpful still.
So, no, the levelling rates in the G modules, as per Bullgrit's analysis, cannot provide levelling rates in AD&D 1e by their extrapolation alone. They can, however, be useful as part of a larger picture.
<snip>
I am making a specific claim as to the relationship between the treasure hauls in various modules. Specifically, I am claiming that the analysis B/Q does on ToH, esp. in light of the deeper context supplied by other thread, paints a "truer" picture of the expectations of the designer. It also, I believe, highlights the problem with the shallower analysis of other modules B/Q has done.
<snip>
I was there, and involved in playing 1e. I know what leveling was like at my table. I played 1e in several US States, with over 100 individual players and DMs. I know what leveling was like at those tables. By extension, since those players were involved with other games, I know what they said about those other games as well.
My experience is that, at least within that dataset, approximately 1/3 of all treasure in a module was missed within the areas explored and that, most frequently, some portion of a module was not explored. Even recently, running KotB using 3e Search rules, we had the same rough ratio. The group explored less than 1/6th of the caverns, and left 1/3 of the treasure unfound in the areas they did explore.
<snip>
If there is no way to determine the norm or intention of the adventures, any % you choose, from the maximum, to 75%, to 2% -- it doesn't matter -- perforce will "massage the data".
The rational conclusion I draw is that a deeper analysis may well demonstrate Bullgrit's conclusion to be correct. Bullgrit's analysis does not.
Fair enough.
I agree that any multiplier is to some extent arbitrary. I think that some multipliers are probably more tenable than others. 2% would be extreme, because a module like ToH which locates all the treasure in a single hard-to-access point is fairly atypical.
The shape of the XP charts in AD&D is roughly geometric for at least some of the classes up until name level - fighters are the closest, I think, and from memory MUs and druids have the largest departure from the geometric patter at their mid-levels (4th to 9th or so).
I think this geometric doubling XP makes the question of the multiplier less significant - a 50% figure will cost many classes only one level, for example (and its potentially heavier effect on those classes whose mid-levels are more linear will tend to bring down the greater level increases that B/Q's data attributes to them).
Your KotB experience, on the other hand, suggests a weighting of a bit less than 1/6 by 2/3 = around about 10%. And even on a geometric XP chart with doubling, 10% of B/Q's numbers will cost about 3 levels (1/8 = 12.5%). On this
handy complilation of the analysis, B/Q gives the level gain from KotB for a Basic D&D party as about +2 levels (a bit more for a thief, unsurprisingly). On your KotB experience, this would suggest that a party could easily move on from that module without having gained a level!
(Although a group of skilled players might use treasure finding, rumours etc to increase their XP count per unit of module dealt with.)
My own feeling is that KotB is probably more likely to produce a 10% experience than the G-modules. As Melan indicates
here, for example, G2 has a much more linear map than does B2.
At posts 339 and 341 of the
original thread you mention a 1 to 1.3 ratio of AD&D to 3E levelling, and comment
this is at some variance with Q's work, I note, which demonstrates in the AD&D modules that the 1e character rate of advancement was slightly higher than that of the 3e rate of advancement.....IOW, opposite of what the linked statements show was expected.
Looking over the B/Q compilation I linked to earlier, they seem fairly consistent with a 1 to 1.3 ratio. B/Q has the ratio after level 1 of ToEE at 4/5 (only the low-XP cleric and thief get ahead of this in AD&D), after level 2 its 6/7 (only the high XP AD&D paladin is below this), after level 3 its 7/8 (again, the AD&D thief pulls ahead) and only after level 4 does some sort of parity set in, as the 3E party hits 9th while the AD&D party gets split by their varied XP charts (MUs and Illusionist benefit from the stretching of their charts to create a higher name level). So until we hit those upper-mid-levels, the AD&D party is about 1 level behind.
G1, by the B/Q numbers, takes an AD&D party to level 8 to 11 (with only the low-XP Illusionist and Thief passing 9th) while the 3E party is at 11th. G2 gives AD&D ranges of 9 to 11, while the 3E party reach 13 - there's the 1 to 1.3 ratio almost exactly. G3 adds one level to most AD&D PCs while adding two to 3E PCs. D1 adds a level only to some AD&D PCs without helping the 3E ones, and at that point B/Q's comparison stops.
So the overall 1 to 1.3 ratio, with some wonkiness particularly in the middle levels where some AD&D classes with higher-than-9th name levels get the benefits of linear advancement, seems prety right to me. Even a 50% multiplier on the modules that B/Q is discussing isn't going to change that very much - it would make it .9 to 1.3, or a bit more than 1 to 1.45. Given the radically different experiences any two groups might have in their own play through a particular module, I'm not too fussy about +/- 10% on the ratio.
(B/Q's KotB study also suggests a ratio less than 1 to 1.3: +2 levels for the typical Basic D&D PC (less for an elf, more for a thief) while the 3E PCs gain 4 levels. This is 2 for 1. Although obviously a single module is hard to generalise from.)
In the original thread, at post #62, you also said (if I've read properly) that:
Oh, yeah, IME, we never used the gp = xp rule, and PCs stopped to train when they were done with what they were doing. So, they'd clear the moathouse, tally XP, and then train. As the above QFT quote shows, YMMV, and probably does.
This seems potentially relevant to your own experiences with levelling rates in AD&D. When I played AD&D I always used the gp=xp rule, plus the magic item=xp rule, and as best as I recall my experiences the B/Q numbers seem pretty reasonable - only moderately slower level gain until name level, and then a major decline in advancement rates.
EDITED TO ADD:
Bullgrit also misses that, in the other published modules, it is not expected that the PCs can recover all of the treasure. (It is actually stated in module B1 that the PCs will not get all the treasure in a good dungeon.) In ToH, though, the majority of the treasure is found in one area; if the PCs succeed, they should get it all.
The difference is not an illustration of how stingy the ToH is; it is an illustration of how off-base Bullgrit/Quasqueton's assumptions were about treasure recovery in other modules
B/Q's analysis has G1 to G3 providing an AD&D party with 380,420 XP, 948,640 XP and 1,618,746 XP respectively. ToH, on the other hand, provides 152,895 XP, or about one-sixth of the G-module average. On the reasoning I've quoted, this suggests someting like a 15% multiplier being expected in play of the G modules. I personally find this very hard to believe - it would mean that the typical name level AD&D PC would not gain any levels over the course of the G series, gaining only 80,000 or so XP (a one-sixth share of one sixth of the total XP available). I can't imagine it being very common for a group to play through the G series in any serious fashion with the PCs gaining less than half the XP that a thief needs to gain a level.
(If you include the 100,000 XP for defeating the demilich than the implied multiplier becomes about 25%, suggesting that a single PC should gain 125,000 XP or so from the G modules - still not enough for even a thief to gain a level, although it might help a 9th level Illusionist or a 10th level MU.)
I think it more likely that ToH is in this respect, as in many others, an outlier.
