PvP Done Right, by Scott Jennings
The designer formerly known as Lum the Mad has a wonderful volcanic eruption of a post up called How To Make A Game With ‘PvP Done Right.’ He makes a lot of great points, and I pretty much agree with all of what he says in there. It’s a great insight for many people, I’m sure, showing how any little decision you make will have repercussions, and designing a PvP system is like walking on a thin line that you might violently crash off into some endless void of suck. Come to think of it, designing anything for a game is pretty much like that. 
Anyway, check out his post. The main lines from each of his points are after the fold.
- PvP should not be the focus of your entire game.
- PvP should not be a random afterthought.
- PvP players hate classes.
- PvP players need classes.
- PvP players detest grinding.
- PvP players need some grinding.
- PvP should not screw new players over.
- PvP should screw over someone.
- “You gotta keep ‘em separated.”
- But not too separated.
- In the endless player skill argument, you should assume your players don’t have any.
- That doesn’t mean you shouldn’t reward those players who do.
- PvP players are angry and bitter, and will hate you.
http://brokentoys.org/2007/12/10/how-to-make-a-game-with-pvp-done-right/
It always cracks me up that although IANAGD (I am not a game designer) the solution to pVp predates the computer by a long shot (hint it goes back to the early 1700s). let me explain:
e = lim(n->inf) of (1 + 1/n)^n
where n = fun
Euler hypothisized that the upper limits of fun (towards infinity) would need to be applied to 1 + and ever decreasing amount of fun (in most cases lets call it grind) but if you raise grind by fun in the end (his definition of the “end game”) then you can find a balance to the limit of fun and in most cases produce a numerical constant for a particular game no matter where you fall on the plot (mathmatical graph “plot”, not “plot” of the actual game — if this were so Mass Effect would cause a tear in this forumla and a huge black hole would open up sucking us all in).
for example:
WOW
n is a low number to start ie very easy entry, not as much treadmill early in the game, attanment of weapons for “achievement emotions” ( sometimes known as the Koster-Mascheroni Constant or whatever it says on MS/XBLA patent) is present in starter zones.
but as the curve progresses (even after mutliple characters) the peak is still rising until approximately 3 weeks after expansion peak 1. Most people in the WASDmatics (study of gaming euler based mathmatics) attribute this to the so-i-waited-all-this-time-for-one-expansion-and-this-is-all-I-got decline discovered back in 1812.
the final value of e ends up being 37.
EQ2
n starts out much higher in the equation because we automatically have to add (hwb) to it. Where has we had just n previously, this (hwb) constant comes into effect with eq2 because all of the yearly (h)ard (w)are (b)udget must be spent in order to begin. Once invested though this addition will be mitigated over time. Although with a 12 month cycle of peaks and valleys that was recently moved to the so-i-waited-all-this-time-for-one-expansion-and-this-is-all-I-got is currently projected to be hit around Feb/march of 2008. This was somewhat flattened by the “GU/fresh Content approximation”, but the current wave is being disrupted by a slight detection in the Oh-God-please-dont-tell-me-scott-is-leaving abnormaility post recent expansion peak.
the final value of e ends up being 86.26 .
EvE
There is no game in history that has ever began with such a high n. It completely destroys all theory and messes the any calculators boot section that tries to computer it. it seems to be so much that it cannot be determined.
A team of seven three-foot-high market analysts that almost died of asphyxication mixed with surprise tried to estimate it back in July of 06 and they surmised it was around 234e26.
Vanguard
The “vision” of n starts out quite high but then slowly decreases as nearer to the origin. The forumla was slighty changed to accomidate to the lack of the final power raise (for endgame), but ends up not existing. Although the line pretty much surrenders to the WKB2SDO approximation (welcome back to the san diego office) it a lot of WASDmatic professors are keenly watching this calculation play out because it just might surprise everyone around 3 ticks out (or atleast somewhere in certain scientists hearts they really want it to).
e = -3 projected e in 2012 = 6
other calcs that need further development but that can be estimated:
Pirates of the Burning Sea : e = 19 (must apply “instancing denominator” but gets “but-its-fuking-pirates-man!” bonus increment each tick.
Fury : e = 6 (this line seems at the beginning of the graph to be a/e(ffected) by the “Red Shirt” differential whereas it was invited along one path/launch, then abruptly killed off to allow others to survive. This has not been confirmed though and is still subject to dispute)
I would love to see any other findings by other WASDmatic scientists.
In Euler we trust,
CB
Wow. You are either a genius or on something. Or both.
neither, just a procrastinator and once i started I couldn’t stop.
someone, somewhere, is reading this and is writing a nerdcore song about it.