Difference between revisions of "Research:Blue water"

From Discworld MUD Wiki
Jump to: navigation, search
(Conclusion: very interesting)
(Conclusion: comments)
Line 81: Line 81:
  
 
:--[[User:Ilde|Ilde]] 06:10, 29 October 2009 (UTC)
 
:--[[User:Ilde|Ilde]] 06:10, 29 October 2009 (UTC)
 +
 +
::My suggestion would be to put the information back together again before analysing it:
 +
 +
::{| class="wikitable" border="1"
 +
!Start
 +
!End
 +
!Total
 +
!Regen
 +
!Regain
 +
!Start as <br />% Total
 +
|-
 +
| 755
 +
| 961
 +
| 1873
 +
| 6
 +
| 200
 +
| 40.3
 +
|-
 +
| 770
 +
| 976
 +
| 1903
 +
| 6
 +
| 200
 +
| 40.5
 +
|-
 +
| 872
 +
| 1078
 +
| 1873
 +
| 6
 +
| 200
 +
| 46.6
 +
|-
 +
| 1042
 +
| 1248
 +
| 1903
 +
| 6
 +
| 200
 +
| 54.8
 +
|-
 +
| 1075
 +
| 1281
 +
| 1873
 +
| 6
 +
| 200
 +
| 57.4
 +
|-
 +
| 1286
 +
| 1504
 +
| 1903
 +
| 6
 +
| 212
 +
| 67.6
 +
|-
 +
| 1270
 +
| 1476
 +
| 1873
 +
| 6
 +
| 200
 +
| 67.8
 +
|-
 +
| 1394
 +
| 1557
 +
| 1903
 +
| 6
 +
| 157
 +
| 73.3
 +
|-
 +
| 1400
 +
| 1599
 +
| 1873
 +
| 6
 +
| 193
 +
| 74.7
 +
|-
 +
| 1419
 +
| 1575
 +
| 1873
 +
| 6
 +
| 150
 +
| 75.8
 +
|-
 +
| 1511
 +
| 1712
 +
| 1903
 +
| 6
 +
| 195
 +
| 79.4
 +
|-
 +
| 1588
 +
| 1832
 +
| 1903
 +
| 6
 +
| 238
 +
| 83.4
 +
|-
 +
| 1620
 +
| 1751
 +
| 1873
 +
| 6
 +
| 125
 +
| 86.5
 +
|}
 +
::My interpretation of the raw data would be that if your starting HP is less than 70% of your total HP you get a 200 HP heal if your HP is greater than 70% of your total HP you get a reduction in the heal.  Assumptions: 
 +
::*the 212 regen encompasses 3 heartbeats of regen (it may of course be a coincidence that the surplus is a multiple of the regen) 
 +
::*the 238 heal could be a data transcription error (just to fitsa nice easy worldview where the heal has a 200 baseline and starts to randomly reduces at a proportion of HP as % of total HP)
 +
::My recommendation if you wanted to research it further would be to collect the data to test a hypothesis, for e.g., heal depends on proportion of HP to Total HP.  And then conduct 30 trials for each proportion that you want to test, for eg. under 50%, 50-70%, over 70%.  You could test with greater sensitivity than that of course, but 30 trials in range increments of 10% could take a while :-).  If you did do 30 trials in each 10% increment there is a possibility that more patterns might start to emerge, for e.g. are you more likely to get a higher/lower result than that expected at 40 than 60%, at 70% than at 80%?

Revision as of 05:00, 29 October 2009

Cyst's fiddling

I created a new alt to figure out the average return of HP by drinking blue water. The results confused me. Disclaimer: I am easily confused.

Attempt #1

While drinking the blue water from a small clear glass phial which holds 20 drops of liquid, and while having a HP regeneration rate of 4 with 13 points in each stat:

  1. HP before consumption: 829(1421), HP after consumption: 883(1421), 883 - 829 = 54 hp - 4 regen hp = gain of 50 hp
  2. HP before consumption: 955(1421), HP after consumption: 1009(1421), 1009 - 955 = 54 - 4 regen hp = gain of 50 hp
  3. HP before consumption: 1025(1421, HP after consumption: 1079(1421), 1079 - 1025 = 54 - 4 regen hp = gain of 50 hp
  4. HP before consumption: 1099(1421), HP after consumption: 1153(1421), 1153 - 1099 = 54 - 4 regen hp = gain of 50 hp
  5. HP before consumption: 1173(1421), HP after consumption: 1227(1421), 1227 - 1173 = 54 - 4 regen hp = gain of 50 hp
  6. HP before consumption: 649(1421), HP after consumption: 703(1421), 703 - 649 = 54 - 4 regen hp = gain of 50 hp
  7. HP before consumption: 719(1421), HP after consumption: 773(1421), 773 - 719 = 54 - 4 regen hp = gain of 50 hp
  8. HP before consumption: 793(1421), HP after consumption: 847(1421), 847 - 793 = 54 - 4 regen hp = gain of 50 hp
  9. HP before consumption: 863(1421), HP after consumption: 917(1421), 917 - 863 = 54 - 4 regen hp = gain of 50 hp
  10. HP before consumption: 933(1421), HP after consumption: 987(1421), 987 - 933 = 54 - 4 regen hp = gain of 50 hp

All ten results are consistent, indicating that 20 drops of blue water result in gaining 50 HP.

Attempt #2

While drinking the blue water from a clear shot glass which has been filled from a small clear glass phial four times so that it contains 80 drops of liquid, and while having a HP regeneration rate of 4 with 13 points in each stat:

  1. HP before consumption: 852(1421), HP after consumption: 1056(1421), 1056 - 852 = 204 - 4 regen hp = gain of 200 hp
  2. HP before consumption: 558(1421), HP after consumption: 762(1421), 762 - 558 = 204 - 4 regen hp = gain of 200 hp
  3. HP before consumption: 806(1421), HP after consumption: 1010(1421), 1010 - 806 = 204 - 4 regen hp = gain of 200 hp
  4. HP before consumption: 1058(1421), HP after consumption: 1262(1421), 1262 - 1058 = 204 - 4 regen hp = gain of 200 hp
  5. HP before consumption: 1035(1421, HP after consumption: 1239(1421), 1239 - 1035 = 204 - 4 regen hp = gain of 200 hp

Like before, the results seem to indicate 20 drops of blue water result in gaining 50 HP.

Attempt #3

However, I then rearranged to have 21 con, 20 str and 8 in each other stat, resulting in a boost in HP with a regeneration rate of 6, and began to see different results while still using a clear shot glass which has been filled from a small clear glass phial four times so that it contains 80 drops of liquid. I also taught myself one level of adventuring.health midway through, which was probably a bad idea since I was doing research, but I'm not a very smart man.

Resulting HP below 1500

  1. HP before consumption: 872(1873), HP after consumption: 1078(1873), 1078 - 872 = 206 - 6 regen hp = gain of 200 hp
  2. HP before consumption: 1270(1873), HP after consumption: 1476(1873), 1476 - 1270 = 206 - 6 regen hp = gain of 200 hp
  3. HP before consumption: 755(1873), HP after consumption: 961(1873), 961 - 775 = 206 - 6 regen hp = gain of 200 hp
  4. HP before consumption: 1075(1873), HP after consumption: 1281(1873), 1281 - 1075 = 206 - 6 regen hp =gain of 200 hp
  5. HP before consumption: 770(1903), HP after consumption: 976(1903), 976 - 770 = 206 - 6 regen hp = gain of 200 hp
  6. HP before consumption: 1042(1903), HP after consumption: 1248(1903), 1248 - 1042 = 206 - 6 regen hp = gain of 200 hp

Resulting HP above 1500

  1. HP before consumption: 1620(1873), HP after consumption: 1751(1873), 1751 - 1620 = 131 - 6 regen hp = gain of 125 hp
  2. HP before consumption: 1400(1873), HP after consumption: 1599(1873), 1599 - 1400 = 199 - 6 regen hp = gain of 193 hp
  3. HP before consumption: 1419(1873), HP after consumption: 1575(1873), 1575 - 1419 = 156 - 6 regen hp = gain of 150 hp
  4. HP before consumption: 1394(1903), HP after consumption: 1557(1903), 1557 - 1394 = 163 - 6 regen hp = gain of 157 hp
  5. HP before consumption: 1286(1903), HP after consumption: 1504(1903), 1504 - 1286 = 218 - 6 regen hp = gain of 212 hp
  6. HP before consumption: 1588(1903), HP after consumption: 1056(1421), 1832 - 1588 = 244 - 6 regen hp = gain of 238 hp
  7. HP before consumption: 1511(1903), HP after consumption: 1712(1903), 1712 - 1511 = 204 - 6 regen hp = gain of 198 hp

I split the results above in two since the inconsistent results seemed to appear as soon as it reached the 1500 threshold, though it's more likely that the controlling formula bases its result on the original HP and not the, well, result.

Attempt #4

I then tried drinking from the small clear glass phial again:

  1. HP before consumption: 1014(1903), HP after consumption: 1070(1903), 1070 - 1014 = 56 hp - 6 regen hp = gain of 50 hp
  2. HP before consumption: 1100(1903), HP after consumption: 1156(1903), 1156 - 1100 = 56 hp - 6 regen hp = gain of 50 hp
  3. HP before consumption: 1186(1903), HP after consumption: 1242(1903), 1242 - 1186 = 56 hp - 6 regen hp = gain of 50 hp
  4. HP before consumption: 1284(1903), HP after consumption: 1340(1903), 1340 - 1284 = 56 hp - 6 regen hp = gain of 50 hp
  5. HP before consumption: 1376(1903), HP after consumption: 1432(1903), 1432 - 1376 = 56 hp - 6 regen hp = gain of 50 hp
  6. HP before consumption: 1468(1903), HP after consumption: 1524(1903), 1524 - 1468 = 56 hp - 6 regen hp = gain of 50 hp
  7. HP before consumption: 1554(1903), HP after consumption: 1610(1903), 1610 - 1554 = 56 hp - 6 regen hp = gain of 50 hp
  8. HP before consumption: 1640(1903), HP after consumption: 1696(1903), 1696 - 1640 = 56 hp - 6 regen hp = gain of 50 hp
  9. HP before consumption: 1726(1903), HP after consumption: 1782(1903), 1782 - 1726 = 56 hp - 6 regen hp = gain of 50 hp
  10. HP before consumption: 1806(1903), HP after consumption: 1862(1903), 1862 - 1806 = 56 hp - 6 regen hp = gain of 50 hp

Consistency returned.

Conclusion

The last attempt indicates that if there is any cap to a consistent 20 drops = 50 HP rule, then it also involves the amount of total drops consumed and not merely a specific amount of HP. I'm not sure how to continue from this point on, though, so if anyone has any ideas...

Cyst 00:17, 29 October 2009 (UTC)

Speculation: Maybe after a certain point the formula incorporates randomness? I suggest this because the fact that some of them were apparently over 200 seems inconsistent with a cap, and also the amount of hp gained doesn't increase or decrease with more or less starting hp.
Hmmmm... I agree that it would be strange for the formula to depend on ending hp. Maybe the cutoff point is whether the starting hp is more than or less than some number between 1270 and 1286? (It doesn't look like it's either percentage of hp (since the 1270(1873) one in the 200 area has a slightly higher percentage than the other) or remaining hp (ditto, since that one had fewer hp left)--not unless that one did have some randomness but the randomness worked out to 200 by coincidence.)
Also, there's a typo (I assume?) in attempt #3, Resulting HP above 1500, #6... the hp after consumption looks like it was copied from an earlier one.
--Ilde 06:10, 29 October 2009 (UTC)
My suggestion would be to put the information back together again before analysing it:
Start End Total Regen Regain Start as
% Total
755 961 1873 6 200 40.3
770 976 1903 6 200 40.5
872 1078 1873 6 200 46.6
1042 1248 1903 6 200 54.8
1075 1281 1873 6 200 57.4
1286 1504 1903 6 212 67.6
1270 1476 1873 6 200 67.8
1394 1557 1903 6 157 73.3
1400 1599 1873 6 193 74.7
1419 1575 1873 6 150 75.8
1511 1712 1903 6 195 79.4
1588 1832 1903 6 238 83.4
1620 1751 1873 6 125 86.5
My interpretation of the raw data would be that if your starting HP is less than 70% of your total HP you get a 200 HP heal if your HP is greater than 70% of your total HP you get a reduction in the heal. Assumptions:
  • the 212 regen encompasses 3 heartbeats of regen (it may of course be a coincidence that the surplus is a multiple of the regen)
  • the 238 heal could be a data transcription error (just to fitsa nice easy worldview where the heal has a 200 baseline and starts to randomly reduces at a proportion of HP as % of total HP)
My recommendation if you wanted to research it further would be to collect the data to test a hypothesis, for e.g., heal depends on proportion of HP to Total HP. And then conduct 30 trials for each proportion that you want to test, for eg. under 50%, 50-70%, over 70%. You could test with greater sensitivity than that of course, but 30 trials in range increments of 10% could take a while :-). If you did do 30 trials in each 10% increment there is a possibility that more patterns might start to emerge, for e.g. are you more likely to get a higher/lower result than that expected at 40 than 60%, at 70% than at 80%?