Scent Mechanics Analysis

[Timber Wolf]

I dove into the scent mechanics in an effort to understand them better and to find out how far away the wolves can pick up on the scent. 

To start off, I wanted to know how much each type of scented item counted on the scent meter.  So, I put together a collection of stinky stuff and rolled through the different combinations and recorded the results.  As an example, 1 Rabbit + 3 Fresh Guts = 3 on the scent meter.

I then estimated what the value might be for each of the items and then ran those estimates through the different combinations, adjusting the item values as necessary.  Eventually, I was able to pin down the value of each item.  The values for the items are not exact, but they are very close.  Fish work the same way as the other meats (1 raw fish = 1 fresh piece of meat).

Once I had the scent values for each item, I set out to determine the distance that the wolves could detect me while I was carrying scented items.  While trying to figure out the best way to go about this, I realized that I needed to first find out if the wind would factor into the equation.  If it does, then it will be almost impossible to get accurate information about the distances. 

And unfortunately, the wind does make a difference in how far away the wolves to detect the scent.  I was able to determine this by picking up and dropping a single piece of fresh venison.  I would remain crouched with the venison in front, and then right-click it as if to re-position it.  When I was in range of the wolf, it would turn to me as soon as I touched the venison.  And as soon as I let go of the venison, the wolf would return to its regular patrol.  I moved away from the wolf until I found a spot where it wouldn't react when I touched the venison.  I marked this location with a screenshot.  Then I ran to the spot where the wolf was, as close as I could approximate it.  Then I marked that location with a screenshot. 

I did this twice, once with a fairly strong side wind and once with a fairly strong head wind.  I plugged the coordinates into the following distance formula.

Test 1 - Wind blowing mostly sideways: √ (945 - 875)^2 + (483 - 530)^2 = 84.3

Test 2 - Wind blowing mostly toward me: √ (873 - 931)^2 + (485 - 482)^2 = 58.1

As you can see, wind appears to make a significant difference.  This means if I was to run more tests trying to determine the distances, I can never be sure how much the wind is affecting the results.

I'm not sure if I'll pursue this any further, but I wanted to share what I was able to find out.  Thanks to @cekivi and @Drifter Man for the advice about how to tackle this mechanic!

 

[earthy]

Really neat analysis.  Though I'm curious if a piece of 100% condition cooked meat has the same scent as a piece of ruined cooked meat. Like you were saying though, the wind does seem to have a pretty profound impact on a wolf's scent range. I have played around with the scent mechanic, both while crouched and out of sight range, and there is something interesting to note. As you already mentioned, while staying out of sight range the wolf will lock onto your scent if you have the decoy, and return to patrol when you drop it.

The interesting thing here is the wolf doesn't seem to be locking on to you,  he seems to lock onto the location he first smelled you. If you were to drop the decoy, pick it back up, and then travel 15-20 paces backwards(all within scent range), the results are unexpected. The wolf will first investigate the location you originally dropped the decoy. When he arrives to the original location and can't spot you, he will turn around and attempt to return to his patrol. Within a second or two of returning to patrol, he will catch your scent once more and lock onto your new location.  It seems possible to greatly influence a wolf's position to your favor this way, though I haven't really used it myself outside of testing.

[Timber Wolf]

1 minute ago, earthy said:

The interesting thing here is the wolf doesn't seem to be locking on to you,  he seems to lock onto the location he first smelled you. If you were to drop the decoy, pick it back up, and then travel 15-20 paces backwards(all within scent range), the results are unexpected. The wolf will first investigate the location you originally dropped the decoy. When he arrives to the original location and can't spot you, he will turn around and attempt to return to his patrol. Within a second or two of returning to patrol, he will catch your scent once more and lock onto your new location.  It seems possible to greatly influence a wolf's position to your favor this way, though I haven't really used it myself outside of testing.

This is exactly what I have observed and it does make moving the wolves around pretty easy. 

Welcome to the forums, @earthy!

[Ruruwawa]

Another great test.  Which experience mode did you use?  I've noticed wolf detection range carrying a single fresh gut seems much larger in Interloper than in Pilgrim.  I didn't note the wind direction, so perhaps that's enough to account for the difference.  But it's a strong impression nonetheless, after 50-100 days in each of the sandboxes (created in v110 and v112).

[Timber Wolf]

Just now, Ruruwawa said:

Which experience mode did you use?

Stalker.

 

3 minutes ago, Ruruwawa said:

I've noticed wolf detection range carrying a single fresh gut seems much larger in Interloper than in Pilgrim.  I didn't note the wind direction, so perhaps that's enough to account for the difference.  But it's a strong impression nonetheless, after 50-100 days in each of the sandboxes (created in v110 and v112).

That's very interesting!  I could definitely see there being a difference in the detection range between the difficulties.

[cekivi]

Happy to help @Timber Wolf and welcome to the forums @earthy 

[Cray]

Great job @Timber Wolf! I've been wondering about this. So the best decoy for minimum scent would be a piece of cooked meat, if I'm reading this right. That makes me more inclined to use it as protection from wolves without attracting every predator in the vicinity. Good to know.

[Timber Wolf]

2 hours ago, Cray said:

Great job @Timber Wolf! I've been wondering about this. So the best decoy for minimum scent would be a piece of cooked meat, if I'm reading this right.

Thanks!  That's correct, cooked meat has the lowest scent value.

[JAFO]

Many thanks for the work you've put into this, @Timber Wolf! I'd say the most valuable information of all is the 'stink values' you've been able to calculate for each item. This brings up a question in my mind.. am I correct in inferring that a fresh rabbit actually ranks then, at 1.1 bars?

What I don't see in your tests is any data for rotten meat, cooked or raw. I think that would be valuable and possibly useful information to have also.

The tests I'm most interested in, (besides the distance tests you haven't yet performed), is the detection ranges for when you are at less than 1 bar. How much does 1 - 5 pieces of cooked meat add to the base detection range, in other words?

 

On 9/19/2017 at 8:43 AM, earthy said:

The interesting thing here is the wolf doesn't seem to be locking on to you,  he seems to lock onto the location he first smelled you. If you were to drop the decoy, pick it back up, and then travel 15-20 paces backwards(all within scent range), the results are unexpected. The wolf will first investigate the location you originally dropped the decoy. When he arrives to the original location and can't spot you, he will turn around and attempt to return to his patrol. Within a second or two of returning to patrol, he will catch your scent once more and lock onto your new location.  It seems possible to greatly influence a wolf's position to your favor this way, though I haven't really used it myself outside of testing.

On 9/19/2017 at 8:46 AM, Timber Wolf said:

This is exactly what I have observed and it does make moving the wolves around pretty easy.

An excellent example of leading wolves around by the nose like this can be seen in @GELtaz' recent video where he experiments with the scent and new bow mechanics. I've set the video to begin playing at the 16 minute mark, where he begins his tests. (see below)

 

On 9/19/2017 at 8:09 AM, Timber Wolf said:

As you can see, wind appears to make a significant difference.  This means if I was to run more tests trying to determine the distances, I can never be sure how much the wind is affecting the results.

I'm not sure if I'll pursue this any further, but I wanted to share what I was able to find out. 

If you do pursue it further (and I hope you do), maybe the best method is to try the approach that GELtaz took.. out on the CH ice, on as clear and windless a day as he could find. Watching the video, it's amazing the distance that 1 scent bar will be detected at.

 

 

[Timber Wolf]

5 hours ago, JAFO said:

Many thanks for the work you've put into this, @Timber Wolf! I'd say the most valuable information of all is the 'stink values' you've been able to calculate for each item. This brings up a question in my mind.. am I correct in inferring that a fresh rabbit actually ranks then, at 1.1 bars?

It's my pleasure.   That's correct, 1 fresh rabbit has a 1.1 scent value.

 

5 hours ago, JAFO said:

What I don't see in your tests is any data for rotten meat, cooked or raw. I think that would be valuable and possibly useful information to have also.

Thanks for reminding me!  I had meant to pass some time and see if it made any difference as the condition of the meat dropped.

I just tested it down to 0% condition and all of the numbers were the same, so meat doesn't get any smellier.

 

5 hours ago, JAFO said:

If you do pursue it further (and I hope you do), maybe the best method is to try the approach that GELtaz took.. out on the CH ice, on as clear and windless a day as he could find.

This is exactly how and where I intended to run the tests, but even if it seems to be a windless day, it might not be.  It could be blowing 2 or 3 mph, and changing directions, and I would never know it.  It wouldn't matter much if this was the only thing factoring into the margin of error, but there is already another large factor - the difficulty of accurately determining the wolf's position at the time it picks up the scent.  I haven't been able to come up with any real good way to do this.  On top of these two things, there could still be other factor(s) I am not aware of that could make a difference.  Tests with this many blind spots are more likely to be misleading than they are likely to be helpful.

One of the biggest takeaways from what we know of how this works, is that the wolves do not follow the scent trail and instead go directly to the spot you were in when they detected you.  So, if you're carrying meat it will probably be a good idea to incorporate some zigs and zags into the route your taking.

[JAFO]

9 hours ago, Timber Wolf said:

there is already another large factor - the difficulty of accurately determining the wolf's position at the time it picks up the scent.  I haven't been able to come up with any real good way to do this.

Indeed.. there's also the matter that the detection range for even just one scent bar is so far that I'm guessing 2 bars would put the wolf on the edge of visibility anyway..

 

9 hours ago, Timber Wolf said:

One of the biggest takeaways from what we know of how this works, is that the wolves do not follow the scent trail and instead go directly to the spot you were in when they detected you.  So, if you're carrying meat it will probably be a good idea to incorporate some zigs and zags into the route your taking.

Agreed. This is part of what's behind my usual advice on the forums that once you've harvested a kill, make haste to be somewhere else, because wolves will be making a beeline for the carcass.

[Timber Wolf]

5 hours ago, JAFO said:

Indeed.. there's also the matter that the detection range for even just one scent bar is so far that I'm guessing 2 bars would put the wolf on the edge of visibility anyway..

Very good point. They are already pretty far away when they pick of the scent of less than 1 bar.

[JAFO]

8 hours ago, Timber Wolf said:

Very good point. They are already pretty far away when they pick of the scent of less than 1 bar.

That's the area of research I'd most like to see attempted.. the various ranges for less than one bar. As you say, wind differences might lessen the accuracy, but surely some kind of average could be obtained.

[WanderingPalm]

Thank you very much Timber Wolf! This information is incredibly helpful. I'm curious, however, if fresh fish counts the same as any other fresh meat in terms of how many it takes to fill up 1, 2, or 3 bars, then why does the fish drop first as decoy? Perhaps there is slightly more range in each bar's total value, and if that is true, then having a "low" 1 bar might have a lower stink radius than having a "high" 1 bar. Which could also mean that condition might affect the smell, but in that same small way that doesn't ultimately show through measuring from the smell bars. Or perhaps I am overcomplicating it.

I wonder if you could use some sort of console command to remove wind for these tests? If you can't from the base game itself, I seem to recall Stormwolf using some sort of unofficial mod that allowed them access to the console recently.

[JAFO]

5 hours ago, WanderingPalm said:

I wonder if you could use some sort of console command to remove wind for these tests? If you can't from the base game itself, I seem to recall Stormwolf using some sort of unofficial mod that allowed them access to the console recently.

The mod does allow you to change the wind. Not sure how much, if any, control it gives you though.

[Timber Wolf]

16 hours ago, WanderingPalm said:

I'm curious, however, if fresh fish counts the same as any other fresh meat in terms of how many it takes to fill up 1, 2, or 3 bars, then why does the fish drop first as decoy?

I have no idea, you'll have to ask someone at HS this question.
 

16 hours ago, WanderingPalm said:

Perhaps there is slightly more range in each bar's total value, and if that is true, then having a "low" 1 bar might have a lower stink radius than having a "high" 1 bar.

This is likely true, considering having just 1 piece of cooked meat gets the attention of wolves, even though that won't light up the first scent bar.  If the scent value is used in the calculation that determines how far away the wolves can pick up the scent, then this is absolutely true.

 

16 hours ago, WanderingPalm said:

Which could also mean that condition might affect the smell, but in that same small way that doesn't ultimately show through measuring from the smell bars.

I tested many combinations with spoiled meat and it made no difference - everything still added up the same.  So, I really don't think the condition of the meat matters at all.

 

16 hours ago, WanderingPalm said:

I wonder if you could use some sort of console command to remove wind for these tests? If you can't from the base game itself, I seem to recall Stormwolf using some sort of unofficial mod that allowed them access to the console recently.

I'm not going to use mods to perform any of my tests, because that would mean introducing another uncontrolled variable.  Even though a mod might target only one particular piece of the code, there's no way to know if it has affected the game in some other way.

[Mroz4k]

On 9/28/2017 at 1:59 PM, Timber Wolf said:

I'm not going to use mods to perform any of my tests, because that would mean introducing another uncontrolled variable.  Even though a mod might target only one particular piece of the code, there's no way to know if it has affected the game in some other way.

Good point. Have you considered marking the distances putting some items on the ground, and then luring the wolf into specific distances with stones? I think that could work pretty well, since I believe the wolf first goes investigate the first sound of a stone hitting the ground, and from that point resets its path, maybe by picking up your scent. But that's a theory, not sure how it actually works in-game.

[WanderingPalm]

On 9/28/2017 at 7:59 AM, Timber Wolf said:

I'm not going to use mods to perform any of my tests, because that would mean introducing another uncontrolled variable.  Even though a mod might target only one particular piece of the code, there's no way to know if it has affected the game in some other way.

True, that is a valid concern.

As always, Timber Wolf, you and your tests are amazing.

and hooray for invisible and/or untestable mechanics! >.>

[cekivi]

11 hours ago, Mroz4k said:

Good point. Have you considered marking the distances putting some items on the ground, and then luring the wolf into specific distances with stones? I think that could work pretty well, since I believe the wolf first goes investigate the first sound of a stone hitting the ground, and from that point resets its path, maybe by picking up your scent. But that's a theory, not sure how it actually works in-game.

The wind would still be an uncontrolled variable though. 

[Mroz4k]

4 hours ago, cekivi said:

The wind would still be an uncontrolled variable though. 

I wonder... in wind-shielded areas, is the wind completely negated, or only somewhat? If it is completely negated, then some narrow place where wind is shielded from most sides, it may be possible to test there.

Another option is the wolf den in DP. Since that is inside, the wind plays no role there. However, the darkness might make it difficult to control the distances, same with the cave corridors.

[JAFO]

22 minutes ago, Mroz4k said:

Another option is the wolf den in DP. Since that is inside, the wind plays no role there. However, the darkness might make it difficult to control the distances, same with the cave corridors.

The problem with that cave is, it's so small that it would be useless for checking detection ranges..

[Mroz4k]

13 minutes ago, JAFO said:

The problem with that cave is, it's so small that it would be useless for checking detection ranges..

That's what I meant with the corridors. Kind of a shame Fluffy is gone, the Dam would be better for testing.

Still, the cave test could have some interesting clues - perhaps the absence of wind will prevent the smelliness from having an effect at all. If that happens to be the case, we will know that the wind is necessary for the tests, and the tests will have to be performed in different areas with different wind conditions, and results will be standardized.

[WanderingPalm]

On 10/1/2017 at 3:15 AM, Mroz4k said:

That's what I meant with the corridors. Kind of a shame Fluffy is gone, the Dam would be better for testing.

Still, the cave test could have some interesting clues - perhaps the absence of wind will prevent the smelliness from having an effect at all. If that happens to be the case, we will know that the wind is necessary for the tests, and the tests will have to be performed in different areas with different wind conditions, and results will be standardized.

I like the idea of using an indoor location for this test, but that might introduce a new variable: sound. Realistically, noises would be amplified in a cave due to the echo, but I am unsure if this is a mechanic implemented in the game. It certainly feels like it is, just from the ambient noise... Besides which, I don't know if anyone has determined how walls or tunnels affect the smell radius, so that might add another variable.

I'm also unsure of what exactly Scruffy does in that cave. I assume he's just eating one of those deer carcasses constantly rather than patrolling. Does anyone know if a wolf that's already eating something will leave it to follow a new smell? 

I wonder if you could simply wait until a time that the windchill is as close to 0 as possible. Is there a weather condition that guarantees low to no breeze? I know some maps are windier than others, too, so perhaps this test might best be done on Coastal Highway (for the visibility and flat land on the lake) in Pilgrim (for even milder weather) or at least Voyageur.

I'm just full of questions today.

[JAFO]

On 04/10/2017 at 5:32 AM, WanderingPalm said:

 perhaps this test might best be done on Coastal Highway (for the visibility and flat land on the lake)

That's not a lake.. that's the ocean. That's why the location is called "Coastal".. lake edges are always referred to as shores, never as coasts.

Here's an overview map, showing how CH is just located on an ocean bay:

[WanderingPalm]

3 hours ago, JAFO said:

That's not a lake.. that's the ocean. That's why the location is called "Coastal".. lake edges are always referred to as shores, never as coasts.

Here's an overview map, showing how CH is just located on an ocean bay:

Haha, yeah, you’re right. I guess I just assumed since it was frozen solid that it was a lake, despite the name of the place and the tides... >.> Hey, don’t judge. We all have our dumb moments.