Sidereal

Horse speed vs. horse color - an experiment

Recommended Posts

One of the things I love about Wurm is the way in which conjecture can rapidly become received wisdom.  After hearing all sorts of things regarding specific horse colors being faster than others, only one course of action remained: Let's do science to it!

 

 

Hypothesis:  Unlike Wurm Unlimited, Ebony horses do not have a sixth speed trait in Wurm Online.

 

Please proceed to the second post for the results of the second experiment, which are much more clear cut and reproducible.

 

(Original experiment and hypothesis included inside the spoiler tag below - more useful is the second experiment that examines specific horse speeds.  First experiment left in place for historical purposes.)

Spoiler

Hypothesis:  Horse color does not affect speed. 


Methodology:

Spoiler
  • Select one horse of each color that has all five speed traits and has never been ridden
  • Each horse should be as identical as possible (age, fatness, sex, etc.)
  • Remove all but my no-drop items for each run to both minimize rider weight and ensure a constant rider weight for all runs (6.275kg)
  • No horse should have any speed enhancing items (horseshoes, saddle, etc.)
  • Use a bridle to lead each horse to minimize rider weight
  • Lead each horse from the pens to the starting point to ensure any theoretical 'riding bonus' doesn't interfere
  • Keep each horse as straight as possible with minimal course correction
  • Feed grass to each horse prior to the first run to ensure it is not hungry
  • Attempt to feed grass to each horse prior to the start of every northbound run
  • Replenish my own food and water to 100% between each set of 8 runs
  • Start the run by pressing 'X' and 'R' at the same time (I have 'R' bound to call for the guards in local, which creates a convenient starting timestamp)
  • Press 'R' when entering the finish tile to create an ending timestamp
  • Record the difference between these two timestamps to time the run
  • Rest each horse for several seconds between runs while recording data
  • Experiment was done on weeks 2 and 3 of the starfall of the Saw

 

Experimental Setup:

A stretch of cobblestone highway running N-S is conveniently nearby.  The northbound run is downhill overall, the southbound run is uphill overall, so only runs in the same direction are an apples-to-apples comparison.  The portion between the Meerhaven waystone and the first bend to the northeast is 293 tiles (1.17 Wurm km).  The course has several flat stretches which are useful for gathering speed data without having to compensate for grade.

Location on: http://wurmonlinemaps.com/pristine
Start tile: (1082, 1601) 
Finish tile: (1082, 1308)

 

Assumptions:

  • Any drift that needs to be corrected to keep the horse on track will be minor and average out between all runs for all horses.

 

With this setup in mind, I collected data for 8 time trials for each horse - four on the northbound run, four on the southbound run.

 

Conclusions:  

  • Horse color does not appear to affect horse speed in any appreciable way  
  • Only speed boosts have any appreciable effect on travel time
  • Appearance and duration of speed boosts appear to be random


Observations

  • Each horse was able to achieve a top speed of 28.10 km/h over flat ground at least once during all 8 runs.
  • Minimum speed over flat ground appears to be 18.73 km/h 19.45 km/h for this methodology (the previous 18.73 appears to be a hunger-related penalty, something to analyze another time).
  • Speed boosts appear to be in discrete quanta.  I'll have a follow up post about this once finish analyzing the data.
  • Horses do not need to warm up to speed boosts, i.e. they may start at the maximum speed from a dead stop
  • Speed boosts appear to average out over time

 

Edited by Sidereal
Examination of the speed trait easier to see than statistical analysis - please see post #2

Share this post


Link to post
Share on other sites

Known fact: The Wurm Online and Wurm Unlimited codes bases are similar but not identical.  

 

Claim: Ebonies only have the bonus speed trait in Wurm Unlimited (!)

 

Observation:  Riding a horse in a straight line over a flat surface for a long period of time indicates that there are specific speeds that a horse will travel at - i.e. horse velocity is not an unbroken continuum.

 

Background:  Some excellent Wurmians gave me a peek at the math that's implemented in the WU code.  Personally, I'm an experimental horse breeder and like to test these things out.  Based on the math in the WU code, the Ebony speed trait is equivalent to a 2.5% boost, as compared to the core speed traits (Stronger Legs (+10%), Fleeter Movement (+10%), and Lightning Movement (+20%)).  In Wurm Online, I can find no evidence for a boost of this strength.

 

Methodology

  • I took the Venerable fat Windpie from the earlier experiment and rode her around the coast of the southern bay for two and a half hours. 
  • Whenever travelling in a straight line on cobblestone, I marked down the speed of the horse (allowing for the +/- .03 km/h that results from steering)
  • After several hours of gathering data in the initial experiment and this experiment, a five-speed horse with a rider carrying 6.275kg will only take one of these 13 values  (experimental data here)
  • Repeat with horses of each color*
  • Each horse should be as identical as possible (age, fatness, sex, etc.)*
  • Remove all but my no-drop items for each run to minimize rider weight and ensure a constant rider weight for all runs (6.275 kg in this case)*
  • No horse should have any speed enhancing items (horseshoes, saddle, etc.)*
  • Use the same bridle to lead and mount each horse to minimize rider weight*
  • Feed grass to each horse prior to each sortie to ensure it is not hungry*

* - represents methodologies consistent with the first, less interesting experiment.

 

 

Conjectures based on the data: 

  • These speeds map on to the three core speed traits - Stronger Legs (+10%), Fleeter Movement (+10%), and Lightning Movement (+20%)
  • These traits will kick in randomly at half strength, full strength, or double strength.
  • Only one trait may be double strength at any given time.
  • Wurm Online does not exhibit any evidence of a speed boost for Ebony horses at any strength.

 

Future experiment: Take horses with fewer than five speed traits and map their discrete values. 

Prediction:  Horses with one speed trait should only have four possible travel speeds.

Edited by Sidereal
Founding the field of Quantum Equine Dynamics - breaking out data into new spreadsheet, and migrating methodology to this post

Share this post


Link to post
Share on other sites

wow, thanks for doing this. I had been under the impression that the whole 'ebony blacks = 6 speed' had been some sort of official statement (I was away when the new colors came in). This is awesome info and I will be following this thread. Maybe now folks will go back to showing some love for all the other nice colors of horses :)

Share this post


Link to post
Share on other sites

Grats on the setup! I really like these mythbuster experiments!

 

My only possible concern is the relatively short course length. Ebony ones are rumoured to activate trait speed bonues more frequently than others - that may not show during a few hundred tiles test. 'mM looking forward to your analysis.

Share this post


Link to post
Share on other sites

Ebonies speed was not a rumor, it was orficial statement upon releasing them... If they are fast as others it's technically a bug

Share this post


Link to post
Share on other sites
5 hours ago, Tilda said:

I had been under the impression that the whole 'ebony blacks = 6 speed' had been some sort of official statement (I was away when the new colors came in)

 

As near as I can tell, this post may have been the start of the rumor, which may have been caused by speculation regarding this post.  However, Meerhaven Pastures does not consider Dr. I Heard That to be an authoritative source.

 

4 hours ago, kochinac said:

Ebonies speed was not a rumor, it was orficial statement upon releasing them... If they are fast as others it's technically a bug

 

Do you happen to have a source for this?  The new colors were released in February 2016 and neither the release announcement nor the patch notes from that year indicate anything to that effect.

 

5 hours ago, Jaz said:

My only possible concern is the relatively short course length. Ebony ones are rumoured to activate trait speed bonues more frequently than others - that may not show during a few hundred tiles test.


This would be a more difficult analysis to do.  Not to get my horse too far ahead of my cart (ha!), but I saw multiple speed bonuses kicking in multiple times over the run used in the experiment, which should mean the course was of sufficient length.  The second part of the analysis will hopefully demonstrate that.  The working theory is that there are five different speed bonuses - each of different sizes - that turn on and off randomly.  <---- Please note that this is pure speculation at the moment.  Turns out it was more complicated.  See post #2.

Edited by Sidereal
Response to kochinac was eaten

Share this post


Link to post
Share on other sites

With my very limited abilty I've tried to speculate something from WU code - I've found only a similarly interesting spot in code where a small trait bonus is given to all NON grey horses... which may be a misread from my side as well, I'm worse than useless reverse engineer :P

Share this post


Link to post
Share on other sites

Jet/ebony black does increase speed, but it's a fairly small boost compared to the "actual" speed traits.

 

Just now, Jaz said:

With my very limited abilty I've tried to speculate something from WU code - I've found only a similarly interesting spot in code where a small trait bonus is given to all NON grey horses... which may be a misread from my side as well, I'm worse than useless reverse engineer :P

 

Not all non-grey, just ebony, was my understanding. Could you paste the code you're referring to?

Edited by Ostentatio

Share this post


Link to post
Share on other sites
2 minutes ago, Ostentatio said:

Jet/ebony black does increase speed, but it's a fairly small boost compared to the "actual" speed traits.

 

 

Not all non-grey, just ebony, was my understanding. Could you paste the code you're referring to?

 

Server -> Creatures -> Creature.class

 }
      if ((!this.status.isTraitBitSet(15)) && (!this.status.isTraitBitSet(16)) && 
        (!this.status.isTraitBitSet(17)) && (!this.status.isTraitBitSet(18)) && 
        (!this.status.isTraitBitSet(24)) && (!this.status.isTraitBitSet(25)) && 
        (this.status.isTraitBitSet(23))) {
        if (isHorse())
        {
          if (sstrength.skillCheck(20.0D, 0.0D, !moving, 1.0F) <= 0.0D) {}
        }
        else {
          traitMod += 0.025F;
        }
      }

 

Share this post


Link to post
Share on other sites

That conditional basically means "if it's not any of the other colors, but is ebony black". Trait bit 23 is ebony, and the others are the rest of the color traits. Doesn't apply to other non-grey horses.

Edited by Ostentatio

Share this post


Link to post
Share on other sites
8 minutes ago, Ostentatio said:

That conditional basically means "if it's not any of the other colors, but is ebony black". Trait bit 23 is ebony, and the others are the rest of the color traits. Doesn't apply to other non-grey horses.

 

Yes I missed the MISSING  '!' in front of ebony...

Edited by Jaz

Share this post


Link to post
Share on other sites

For what it's worth, the reason it checks for the other color traits is because horses can actually have more than one. For the horse's "actual" color, there's basically a priority order to them... so for example, if a horse has the ebony trait but also the brown trait, it'll be brown, because ebony is last on the list. This is part of why ebony is the rarest color if you were to breed randomly.

Share this post


Link to post
Share on other sites
5 minutes ago, Ostentatio said:

Jet/ebony black does increase speed, but it's a fairly small boost compared to the "actual" speed traits.

 

Interesting!  Given that WU is approximately equal to WO and based on the data, I'll have to revise my hypothesis to "Horse color has no appreciable effect on overall horse speed."  Given that the fastest run was 168s with a Blood Bay and the slowest run was 204s with a Grey, any net effect that might provide is going to be completely swamped by the randomness of speed boost activation.

Share this post


Link to post
Share on other sites

if you really wanted to test horse speed comparisons you need to do it without speed traits, you need the speed influence to be the constant and horse color the variable. traitless horses or hell horses go the same speed so you can see if the max speed is any different

 

otherwise it's no different than throwing paper planes around with random gusts of wind and using that to determine which one flew the furthest based on the color of the paper used

Share this post


Link to post
Share on other sites
3 hours ago, kochinac said:

Ebonies speed was not a rumor, it was orficial statement upon releasing them... If they are fast as others it's technically a bug

Seems like a unfair mechanic.

 

 

Share this post


Link to post
Share on other sites

I would agree with MrGary. If you do intend to get to the bottom of whether horse colour has any effect on speed, retry the experiment with no traits on any of the horses.

Share this post


Link to post
Share on other sites

For folks following this thread, I've updated my reserved post with the proposed science of Quantum Equine Dynamics.   

 

Although the WU code indicates Ebony horses receive a minor speed boost, the additional data suggests that there is no equivalent Ebony-specific speed boost in Wurm Online

 

The suggestions are appreciated, MrGARY and McGarnicle.  If you can provide me with a zero-trait Ebony horse and a zero-trait non-Ebony horse, I'll be glad to perform the experiment.  At the moment, I believe the evidence is quite compelling as it stands.  The detailed velocity data required some time to parse and build a theory around.  I'm off to see if I have a 1-speed trait horse in stock that I can test my prediction with.

 

Update:  Miraculously, I did have an old horse with a single speed trait - Lightning Movement.  After four laps around southern bay, I've seen precisely four speeds which bears out the prediction.  Granted, they were not the speeds I expected - rather than half/normal/double, they were more of a low/medium/high range of speeds - which means the internal mechanics of how much speed each speed boost supplies a little murkier.

Edited by Sidereal
Update for new data

Share this post


Link to post
Share on other sites
1 hour ago, Sidereal said:

If you can provide me with a zero-trait Ebony horse and a zero-trait non-Ebony horse, I'll be glad to perform the experiment.

 

There's no need because many people have already tried it.  Ebony horse vs a not ebony horse has a faster top speed, but we are talking the difference of like 1-2 km/h

Share this post


Link to post
Share on other sites

Thanks for your experiment, and for publishing the results.

 

A while ago on Deliverance, some tests were made by Koroth and Awen (animal breeders) to find out if ebony black 5 speeds were faster than identical 5 speeds of other colours, given the same jockey over an identical track.

 

According to them, there was no discernible difference in speed.

 

I have not heard of any experiments on WO showing that ebony black horses are faster with the same jockey over a set track.  If anyone has done an experiment on WO recently, please publish your findings here.

 

Assuming WO code is identical to WU code would be folly, as we have no proof either way.  It is always possible the ebony speed boost was indeed put in to WO but the code could easily have been rolled back at any time since it went in.  Things happen.

 

Some breeders still want to sell Ebony Blacks at a higher price, but I think that if they do, it should only be because they personally find them more difficult to breed, and not because they are thought to be faster, as the situation stands.

Edited by Muse

Share this post


Link to post
Share on other sites

What I have found after 20 hours of riding a horse under controlled conditions and staring at the speed is:

  1. The maximum speed of an ebony horse and the maximum speed of a non-ebony horse on a flat surface under those conditions listed in the methodology are identical:  ~28.10 km/h, plus or minus a few hundredths due to the way speed is recalculated after turning.
  2. For a five speed horse, there appear to be exactly twelve 'gears' for the speed of the horse after the base speed.
  3. There is no difference in any of those gear speeds between an ebony horse and a non-ebony horse.

I do not find even a minor difference - not 2km, not 1km, not even 0.2km - under any of these circumstances.  I can only conclude that there is no appreciable difference in speed between the two in Wurm Online at this point in time.

 

I continue to look and cannot find anything to suggest that Ebony horses are different in any way - no official announcement, no forum post, no patch notes seem to support the idea that Ebony horses were intended to be faster.  If someone can find the smoking gun, I'd love to see it for my own sake.  And although I've looked at the WU code that does indicate a difference in speed,  I cannot reproduce that in WO at this point in time.

 

More interesting I think is that what I'd hoped to prove via statistics (granted, a sample set of n=4 is pretty slim pickings) has instead opened up a new line of inquiry that I hope to be able to pursue.

Edited by Sidereal
Edited for clarity

Share this post


Link to post
Share on other sites

I deleted my comments above, and am now extending a personal apology  to Sidereral, as I realized after thinking that the main reason for my opposition was simply, I did not want this to be true.

 

I especially did -not- like the suggestions  that ebony speed bonus was a myth simply invented out of thin air by horsebreeders to boost sales. I am still not convinced that the testing mechanism was 100 percent conclusive, but as i said, I realized my opposition was based on emotions and not facts. I did not want it to be true. It still may not be true, but that needs to be proven by "science" and not by wishful thinking, which I was guilty of. If it is true, I I realized it may be another case of wurm developers going back and breaking their own code, something any longtime wurm player is familiar with.

 

In any case, my own arguments were emotion based, and not reason based, and so I am deleting them as they embarrass me now.

 

 

Share this post


Link to post
Share on other sites

Brash_Endeavors - an apology is not necessary but accepted just the same and you are the bigger Wurmian than I for making it.  Your clarity of thought in hindsight is a credit to your talents.

 

I'll continue to pursue this line of inquiry until something more concrete shakes out and keep y'all posted.

Share this post


Link to post
Share on other sites

What about grooming a horse. I am convinced that grooming a horse sets a faster speed than a non groomed horse

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now