Mechanics questions

[Drew]

Here's a thread to put dumb questions you never bothered asking:



Does everyone in the game receive a mana pulse at the same time or is it tied to the character?


Has anyone done any research on the exact average time a pulse can take and what's the longest and shortest time it can take to pulse?

[droit]

Mana pulses are universal. I've heard 2 minutes +/- 30 seconds, but there was a thread a while back in which there was some dispute over those numbers.

[WRoss]

jamus did a count of like 10000 pulses in a row and said he couldn't figure out the formula

[Bobmuhthol]

Unless the mana pulses are designed to have no real distribution, I'm willing to bet that a sample of 10000 pulses would resemble some sort of common density function. He probably didn't look for one.

[Askip]

2 to 2.5 minutes seems to be my average.

I do not time it but it seems never to be under 2 minutes, can anyone verify that it can actually be less?

:D

[WRoss]

someone on psinet type ?pulse

[Deathravin]

It's per area. sort of.

Lets put it like this... If I have 10 characters logged in, each one in a different city, sitting around... They'll pretty much all get it at the same time, or within a few seconds of eachother.

If I move them around with Go2 constantly, using Plat Portals, and just move them randomly, some of them will receive double pulses, some of them wont receive a pulse, they'll be out of sync... But even when they're moving around, 8 out of 10 of them will still be completely in sync.

I've heard that the Rift and the resting area outside of it has weird pulses, out of sync with the general area.



So with that information, I theorized that there is one big pulse for 80% of the world, but there are different areas with slightly changed pulse characteristics.

Pulses in general are something like between 92 and 152 seconds apart (I'll have to dig up my spreadsheet), with experience/stamina coming in at the end of each pulse and directly in the center. (so if you time exactly how long it took between exp pulses, you can know exactly when the next mana pulse will come).


Quick mention of how I came by this data... Simply had a lich script time the distance between exp pulses while I was hunting one day. I got curious and logged into every account I had a password for and made a new script to time the mana gains and infuse any mana gained to keep below 1/4th mana pool. And did a test with them sitting still for 10 hours, and flying around like bats out of hell for 4 hours with portals on, and flying around like bats out of hell for 4 hours with portals off.

Also ran all the same tests for 30 minutes in prime to ensure my data wasn't unique to plat.

Saved data, and put on server, tried to make a mana pulse timer for my uberbar, got frustrated with SF limitations and gave up LOL.

[Ryvicke]

2 to 2.5 minutes seems to be my average.

I do not time it but it seems never to be under 2 minutes, can anyone verify that it can actually be less?

:D

Just anecdotally--sometimes when I have to play on Avalon (it has a pulse timer like the wizard) there are many times when the pulse hits before the timer hits :00. So I believe there are definitely pulses under 2 minutes. I'm guessing it's just a random roll and server clock function that willy-nilly decides to add or subtract x seconds from 120 and pulse. Just like every other randomly generated "decision" in gemstone.

[Deathravin]

From what I remember it was 92-152 seconds, with a stepping of 4 seconds. So 92, 96, 100, ... etc.

[Bobmuhthol]

The ?pulse list (627 data points) has an average within 2% of 120 and a median of 120, standard deviation of 17. Min is 91 and max is 153.

I don't have STATA installed on this computer but a uniform distribution with those paramaters would have a standard deviation of approximately 17.9, so that's my best guess right now. It is definitely a symmetric distribution, anyway, so it's probably uniform or normal.

Some values do not exist, so Deathravin is probably right about the stepping and there is likely some measurement error causing odd numbers to occur in the sample. I'm calling it a discrete uniform distribution.