Difference between revisions of "Empirical Evaluation of Fences"

From Wurmpedia
Jump to navigation Jump to search
m (Removed old info)
 
(9 intermediate revisions by 4 users not shown)
Line 11: Line 11:
 
Multiple characters were used for simultaneous construction. Due to the extreme difficulty the tall stone wall was completed last, but still within 10 minutes of the other fences.
 
Multiple characters were used for simultaneous construction. Due to the extreme difficulty the tall stone wall was completed last, but still within 10 minutes of the other fences.
  
Note: [[Golden Valley]] decays 3x faster than the premium server.
+
Note: [[Golden Valley]] decays 3x faster than other servers.
  
If you have any questions, please feel free to PM [[systemparadox]] on GV or on forums.
+
== Results ==
  
== Results ==
+
We have learned that:
 +
* The difference in decay between fence types is negligible.
 +
* Wurm has nonlinear decay. The more damaged an item is, the faster it decays, so repair often.
  
All times are UTC+1.
+
All times are UTC+1. See below for a graph.
  
 
{| border="1" cellspacing="0" cellpadding="2" style="text-align:center;"
 
{| border="1" cellspacing="0" cellpadding="2" style="text-align:center;"
Line 281: Line 283:
  
 
http://systemparadox.co.uk/~simon/fences/fences.png
 
http://systemparadox.co.uk/~simon/fences/fences.png
 +
 +
== A second experiment ==
 +
 +
About 3 years after this first experiment, a second experiment was set up on Deliverance. 6 different fences of 20ql were built: Woven Fence, Crude Wooden Fence, Wooden Fence, Low Stone Wall, Tall Stone Wall and Iron Fence. The damage was noted each day for 6 months, and here are the results, collected between February and July 2013 by [http://www.wurmpedia.com/index.php/Player:Yaga[Yaga]]:
 +
 +
[[Image:FenceDecayFinal.jpg|link=http://img94.imageshack.us/img94/926/zo6r.jpg|Click here for results]]
 +
 +
The previous results could be confirmed: within statistical limits all fence types show the same rate of decay. (The minor spread at higher damages was mainly due to slight ql differences.)
 +
[[Category:Fences]]

Latest revision as of 19:29, 17 July 2023

Aim: To answer once and for all the age-old question: which fence type to build?

Wooden fences and stone walls are easy to build. Palisades, while hard to obtain material for, are not noticably more difficult to build. Tall stone walls take an immediate penalty because even with 20 masonry the success rate for attaching bricks is less than 20%. Past experience shows that they are similarly hard to repair. Bricks aren't exactly easy to make either.

All 4 fences were created on Golden Valley at 20:00 09/06/2010 (UTC+1), using 20QL components and 20 carpentry/masonry skill, resulting in 21QL fences. Multiple characters were used for simultaneous construction. Due to the extreme difficulty the tall stone wall was completed last, but still within 10 minutes of the other fences.

Note: Golden Valley decays 3x faster than other servers.

Results

We have learned that:

  • The difference in decay between fence types is negligible.
  • Wurm has nonlinear decay. The more damaged an item is, the faster it decays, so repair often.

All times are UTC+1. See below for a graph.

Date Fence Stone Wall Tall Stone Wall Palisade
2010_06_09 20:00 0.0 0.0 0.0 0.0
2010_06_10 20:20 0.47619048 0.47619048 0.0 0.47619048
2010_06_10 23:50 0.47619048 0.47619048 0.47619048 0.47619048
2010_06_11 02:32 0.47619048 0.95465934 0.47619048 0.47619048
2010_06_12 00:15 1.4354396 0.95465934 0.47619048 0.95465934
2010_06_12 14:23 1.4354396 1.4354396 0.95465934 1.4354396
2010_06_13 14:08 2.4040701 1.918565 1.4354396 1.918565
2010_06_13 21:45 2.4040701 2.4040701 1.918565 1.918565
2010_06_14 23:40 2.8919907 2.8919907 2.4040701 2.8919907
2010_06_15 15:00 2.8919907 2.8919907 2.8919907 3.3823626
2010_06_16 13:55 3.3823626 3.8752234 3.3823626 3.8752234
2010_06_17 17:15 3.8752234 4.370611 4.370611 4.370611
2010_06_18 00:01 3.8752234 4.868565 4.370611 4.370611
2010_06_19 00:50 4.370611 5.369126 4.868565 5.369126
2010_06_19 18:04 4.868565 5.872334 5.369126 5.872334
2010_06_20 21:25 5.369126 6.8868647 5.872334 6.3782325
2010_06_21 21:10 5.872334 6.8868647 6.3782325 6.8868647
2010_06_22 20:15 6.3782325 7.9125104 6.8868647 7.3982754
2010_06_23 15:40 6.8868647 8.429617 7.3982754 7.9125104
2010_06_24 02:36 6.8868647 8.949644 7.9125104 7.9125104
2010_06_25 01:05 7.9125104 9.472641 8.429617 8.429617
2010_06_26 00:50 7.9125104 9.472641 8.949644 8.949644
2010_06_26 18:10 8.429617 9.998659 9.472641 9.472641
2010_06_27 13:40 8.429617 10.527752 9.998659 9.472641
2010_07_07 18:00 14.322269 16.567505 15.437484 16.000607
2010_07_09 17:10 15.437484 17.712934 16.567505 17.138254
2010_07_16 01:43 18.87442 22.462414 19.4614 21.248388
2010_07_19 02:30 21.248388 24.319668 21.248388 23.6956
2010_07_21 03:45 22.462414 24.948881 23.076555 24.948881
2010_07_22 00:10 23.076555 26.223269 24.319668 25.58337
2010_07_23 17:24 24.319668 27.519861 24.948881 26.868717
2010_07_26 03:00 25.58337 28.839859 26.868717 28.839859
2010_07_30 01:20 28.176855 32.251175 29.50904 31.555445
2010_07_31 01:30 28.839859 32.954052 30.184574 32.251175
2010_08_02 00:25 31.555445 35.10785 32.954052 35.10785
2010_08_08 14:10 36.583878 39.642796 38.09467 39.642796
2010_08_10 16:00 38.09467 41.23115 39.642796 42.041428
2010_08_20 14:40 47.159336 49.910625 50.861305 50.861305
2010_08_26 03:10 52.81895 56.99464 55.91449 58.10192
2010_09_02 17:50 64.131584 65.45919 65.45919 71.350174
2010_09_05 16:00 69.774704 71.350174 68.273766 76.66466
2010_09_08 14:30 74.77676 78.7053 73.01228 86.31562

fences.png

A second experiment

About 3 years after this first experiment, a second experiment was set up on Deliverance. 6 different fences of 20ql were built: Woven Fence, Crude Wooden Fence, Wooden Fence, Low Stone Wall, Tall Stone Wall and Iron Fence. The damage was noted each day for 6 months, and here are the results, collected between February and July 2013 by [Yaga]:

Click here for results

The previous results could be confirmed: within statistical limits all fence types show the same rate of decay. (The minor spread at higher damages was mainly due to slight ql differences.)