DPS Excel calculator for Sorcerer / Sage

[lostdummy]

(03-22-2012 04:13 AM)CaseyTheRetard Wrote:  This is the part where I say I would use my actual DPE/DPET for the same purposes that you would use your predicted DPE/DPET, and you - being a formulationist - complain that my numbers are inaccurate, and I - being a simulationist - complain that yours rest on too many assumptions. Then we either continue to argue or give up or - ideally - decide that we are both correct. Your numbers are accurate for the ideal formulation model, and mine are accurate for the scenario being simulated. Those two things will never be truly identical, and are both just different approximations of what really goes on in-game.

Well, I mentioned before that this is not about philosophy , and neither is this about simulation vs formulation. And especially it is not about me being 'formulationist' ;p

I wrote much more simulations than I did formula based solutions like excel. In fact, I worked quite a bit on simcraft too, and added quite a number of features, and most of those exists still in SimC although I bet Nate did his "rewrite" act on most of them by now. While I cant remember them all, most would probably go under 'practical' or 'simulationist':
- normalized random to be able to use 1000 runs and get stable results
- 'if=' with condition parsing and possibility to AND, OR combine conditions (before you could only use fixed precoded conditions specific for spell)
- new Buff classes that could be generically used in above '=if' conditions, and also could generically collect usage stats
- ability for simcraft to read HTML pages and parse them
- ability to parse character data (gear,talents...) from sites like armory
- default action lists (so only needed line was armory=xyz, and it recognize class and select default actions if none specified)
- writing game(wow) log files, so other parsers can use sim data
- aliases in actionlist, so you could for example use 'Force_Lightning' instead of 'Telekenetic_Throw', even if Empire names were not supported
- ability to simulate moving and breaks during fights

etc, I cant remember all, but these should be enough to show that there was nothing 'formulationist' about them ;p

So ... back to issue - I said several times that this is not about preferences but about not being able to do with 'clipped' DPE values what you would be able to do with 'full' DPE values. Here I will repeat just two such cases, and I would appreciate if you can explain how you would do those with your DPE/DPET. Both cases are related to testing/checking if SimC values are correct, by comparing them with something else:

1) you add some feature in SimC, like some new talent or new spell in rotation, and your numbers change more than you expected. now you want to check that new addition did not change spells it should not change. Easiest way to do that , as a start, is to compare DPE of those spells in old run with DPE of those spells in new run

- for this to work, you must use 'my' DPE values, since between those two runs you may have different count on spell you are comparing (for example if you added new spell in rotation), so if you use 'your' DPE you will get different value in those two runs even if (real) 'my' DPE value of that spell remained same in both runs

2) if you want to compare SimC results with other calculation , like excel. Again, when there is difference, too many factors would effect total damage or DPS, but good starting point is checking DPE of single spells. If talents/gear are interpreted correctly on both sides, and spell is not under significant influence from other spells effects, then their DPE should be same, even if number of used spells in rotations on both sides is not same

- again, for this to work you can not use 'partial' DPE, because even if both tools have exact same starting conditions (talents, gear...) and both have correct calculations and interpretations of effects, they may not have exactly same number of used spells. meaning that with 'your' DPE in practically all cases you will get different values, while with 'my' DPE you will get same results if inboth tools interpretations of talents and damage calculations are correct, even if number of used spells differ somewhat

(03-22-2012 04:13 AM)CaseyTheRetard Wrote:  Religious wars aside, I'll admit it would be really handy if Simulationcraft's Abilities table had an "Avg" column next to hit and crit and "T-Avg" near "T-Hit" and "T-Crit" because I do sometimes want to predict the damage of n ticks or Lightning Strikes or whatever and it would be nice not to have to compute ((crit-hit)*crit%+hit) in my head.

agree with this ;p