PDA

View Full Version : Elemental confluence movement roundtime tracking (script included)



Goat
08-17-2016, 11:46 PM
Agility (AGI): 100 (30) ... 100 (30)
Intuition (INT): 100 (30) ... 100 (30)
Climbing...........................| 134 37
Survival...........................| Ain't got none.

>;countconfluence
... hunt in the conflence ...
>showcon
total moves: 200
total RT moves: 116
0 sec: 84
1 sec: 33
2 sec: 30
3 sec: 31
4 sec: 22

0 sec: 42%
1 sec: 16.5%
2 sec: 15%
3 sec: 15.5%
4 sec: 11%



resetcon sets everything back to 0.
countconfluence.lic: http://pastebin.com/kx1bw00b
It's basically a bastardized version of sexual-favors. I'm not positive it works in all cases but it looked good for me.

Anyone else want to try it?

Goat
11-12-2016, 04:50 AM
These two samples will serve as my controls once I get some survival ranks. The first represents my situation when hunting with a pike, the second comes from hunting with a runestaff (I use a different set of enhancives for each).

Pike setup
Agility (AGI): 100 (30) ... 115 (37)
Intuition (INT): 100 (30) ... 100 (30)
Climbing...........................| 155 55
Survival...........................| Ain't got none.

total moves: 579
total RT moves: 351
0 sec: 228
1 sec: 84
2 sec: 76
3 sec: 82
4 sec: 109

0 sec: 39.4%
1 sec: 14.5%
2 sec: 13.1%
3 sec: 14.2%
4 sec: 18.8%
mean RT per move: 1.58 seconds

I guess the benefit of +7 AGI bonus and +18 climbing ranks (compared to OP) is not enough to overcome noise in medium-sized samples. Either that, or it matters that I spent a decent fraction of the hunt wounded.
I'll likely try again soon with 1x survival. Hopefully that will be a more noticeable benefit.

Runestaff setup
Agility (AGI): 100 (30) ... 100 (30)
Intuition (INT): 100 (30) ... 100 (30)
Climbing...........................| 145 45
Survival...........................| Ain't got none.

total moves: 328
total RT moves: 217
0 sec: 111
1 sec: 40
2 sec: 65
3 sec: 60
4 sec: 52

0 sec: 33.8%
1 sec: 12.2%
2 sec: 19.8%
3 sec: 18.3%
4 sec: 15.6%
mean RT per move: 1.70 seconds

Goat
11-18-2016, 01:32 AM
With 20 ranks survival (otherwise identical to "pike setup" above):

total moves: 250
total RT moves: 104
0 sec: 146
1 sec: 33
2 sec: 34
3 sec: 37

0 sec: 58.4%
1 sec: 13.2%
2 sec: 13.6%
3 sec: 14.8%
4 sec: 0%
mean RT per move: 0.85 seconds

Huge difference from 20 ranks. Mean RT/move cut in half, no more 4 sec RTs, and 58% RT-free moves compared to 33-42% in the trials above. Since we know it takes way more ranks in both survival and climbing to eliminate RT, it makes me wonder if the effect of survival is based on skill bonus rather than ranks.

Goat
11-19-2016, 01:33 AM
I've got some sample size issues. Things got worse with 40 ranks of survival (though they were with the lower-AGI/lower-climb runestaff setup):

total moves: 204
total RT moves: 94
0 sec: 110
1 sec: 24
2 sec: 26
3 sec: 44

Goat
12-01-2016, 03:16 AM
I've come upon the idea that confluence RT is heavily threshold-based. The hypothesis is that survival/climbing ranks (and additional AGI/INT bonus) accomplish nothing until you reach some new threshold, at which point you get a substantial benefit.

I believe the effect of hitting a threshold is to reduce your max movement RT by one second. At that point, whatever probability you used to have of hitting your max RT now becomes additional probability of no movement penality.

I imagine a system something like this:

With worst stats/skills:
1 sec 16.6%
2 sec 16.6%
3 sec 16.6%
4 sec 16.6%
5 sec 16.6%
6 sec 16.6%

First improvement:
0 sec 16.6%
1 sec 16.6%
2 sec 16.6%
3 sec 16.6%
4 sec 16.6%
5 sec 16.6%

Second improvement:
0 sec 33.3%
1 sec 16.6%
2 sec 16.6%
3 sec 16.6%
4 sec 16.6%

Third improvement:
0 sec 50%
1 sec 16.6%
2 sec 16.6%
3 sec 16.6%

Etc., until finally reaching no RT at all somewhere around 200 ranks surv/climb.

To be clear, this is only a hypothesis, but it is somewhat well supported by the data I've gathered (which covers only the cases of 4/3/2 max RT). I'm pretty confident about the max RT threshold, but less confident about the exact probabilities.

Here are some more trials:

With 60 ranks survival (runestaff):
total moves: 204
total RT moves: 99
0 sec: 105
1 sec: 30
2 sec: 31
3 sec: 38

With 80 ranks survival (runestaff)):
total moves: 212
total RT moves: 107
0 sec: 105
1 sec: 39
2 sec: 35
3 sec: 33

With 101 ranks survival (runestaff):
total moves: 214
total RT moves: 94
0 sec: 120
1 sec: 33
2 sec: 33
3 sec: 28

All the runestaff trials above were with 45 ranks climbing.

With 101 ranks survival, 60 ranks climb (runestaff):
total moves: 200
total RT moves: 61
0 sec: 139
1 sec: 33
2 sec: 28

Goat
10-04-2017, 12:48 AM
I finally found a skill/stat threshold for 1 second max RT:

Agility (AGI): 100 (30) ... 122 (41)
Intuition (INT): 100 (30) ... 102 (31)
Survival...........................| 217 117
Climbing...........................| 201 101

(Holding the other stats/skill fixed, 116 ranks survival is not enough).

total moves: 75
total RT moves: 9
0 sec: 66
1 sec: 9

I'm unlikely to ever try for 0 sec RT. The occasional one-second delay is nearly unnoticeable. I may try to fiddle at the margins, though, and get a sense of how many ranks of climbing makes one rank of survival, etc.