Why Logarithmic?
It's been asked why use a logarithmic approach rather than a linear one. There are several reasons for this.
- Skills that TMd too much should be reduced most
Some skills have TMd way too much while others haven't. This approach reduces the way high skills far more than the lower ones.
- Some skills have very high levels that don't TM.
This, however, is largely caused by a knock-on effect. Because many skills were TMing very fast people had a lot more leftover XP to spend on those skills. When that is combined with the next reason it's clear that these skills need reducing as well.
- Skills cost more at higher levels
This is how things are intended to be. Right now once you're over level 60 the cost of learning and the increase in bonus is linear. This is one of the areas that just isn't coping with the really high skills.
However, when it's corrected 20 levels at level 400 will be worth far more than 20 levels at level 100. - Lower level characters (especially newer players) don't
advance fast
With a linear approach they take a much bigger hit. Many newbies take 5 or more days to simply gain 50 levels. Chop them by 60% or 70% and as a percentage of their Discworld time you've taken away a huge amount.
Other solutions
Some other solutions have been suggested apart from the linear approach.
- Player purge
We could do this. Some muds do it on a regular basis. This has not been Discworld's approach and, unless there is a lot of player support for the idea, I and the other creators would rather not take this option.
- Skill specific modification
This option gets very complicated. For a start it can lead to a situation where after the change you are lower level than someone who you were higher than before the change.
It also raises a lot of complexity about exactly which skills should be modified more and which less. This is especially true for skills that at one point TMd too much but recently have been fixed. Some players will have taken advantage of this while others wouldn't have. Even players who were around during the extremely generous period may not have taken advantage of it.
The final problem is that as mentioned before skills are relative. If you reduce fighting.combat.xxx and other.health differently you get an imbalance. If you modify other.perception and covert.xxx differently you get an imbalance. The same kinds of problems arise for many different skill comparisons.
- Why not just change bonuses
This approach falls down in a number of ways.
Firstly we would still have the problem whereby newbies wouldn't be able to reach the levels of oldbies. Unless of course we picked some arbitrary level to start seriously reducing the rate of TMs.
Plus the system needs to have head room (Yukk, Jakka et al will still be wanting to advance a bit at least) so we'll be looking at even higher skills to which newbies will need to be able to advance.
It would also make the differential between players and NPCs even worse since all the NPCs would be _downgraded_ with respect to the players rather than upgraded.
Unless of course we completely changed how skills are dealt with and made NPCs use a different algorithm to players which will lead to further confusion in the future (and make comparison of levels between players & npcs impossible).
Players however would still have "lost" everything they'd lose through a skill reduction.
In other words, this approach adds complexity, doesn't fully solve the problem but still downgrades the players. The only positive it has is that players can feel as though their numbers haven't gone down.
If that's what we want to achieve we could always write an 'oldskills' command or something to show you what your levels would have been. Then you could alias 'skills' to 'oldskills' so you feel as though nothing has changed.
