Full transcript
0:00Today we are going to answer two very
0:01common questions about the carpet . The
0:04first question is : how good is the
0:05carpet ? The second is how do we compare
0:09builds with different carpets , whether
0:12it's a carpet two versus a carpet three
0:14, or a carpet three with a 24 - second
0:17standard versus a carpet three with 18
0:19seconds using a cooldown gem . And , at
0:23the end of the video , I will show you
0:25how you can make a comparison , for
0:27example , of these wheels here , where
0:30one has the carpet cooldown gem but has
0:32less total strength on the wheel , and
0:34another that doesn't have the cooldown
0:36gem , so it's the standard 24 - second
0:38level 3 carpet , but it has more damage ,
0:41right ? It has ballistics and , anyway ,
0:43it has more total wheel points as well .
0:46But before going straight to the
0:48calculator , we need to do a mental
0:50exercise first , because if we are going
0:52to model the carpet , we need to model
0:54how we manage to calculate this damage
0:56bonus it has . However , the bonus we
1:00know , which is 8 , 10 , or 12 % , is damage
1:04that only works when you are active on
1:07top of the carpet . But that is not
1:10enough for us to be able to model it ,
1:12because this is only a part of the hunt
1:13. There is another part of the hunt
1:15where you don't have the carpet active ,
1:16right ? So we need , besides knowing the
1:19damage it gives , to know the ratio in
1:21which it is active during the hunt . And
1:24you know something similar to this , for
1:26you to understand what we will be doing
1:28here today , is critical damage . It also
1:30has the same behavior . It has a part
1:32that is how much extra damage we are
1:34gaining and it has a part that is how
1:36often it is activating , right ? Here it
1:39is a probability ; for the carpet , we
1:40are talking about a ratio . So , when we
1:43come here to Tibia Tools and compare a
1:45build that has critical damage with a
1:4710 % chance and a 50 % critical damage
1:49increase , compared to a build that
1:51doesn't , we see that the build that has
1:53this critical damage has an average
1:55damage increase of 5 % here , which is
1:57precisely the multiplication of the
1:59chance by the increase . If I make it 60
2:02here , it will increase to six here . If
2:05I make it 100 here , it increases to 10 .
2:08So it's simple , it's just a
2:09multiplication of the two factors . So
2:11for the carpet , we're going to do the
2:13same thing . We already have the damage ,
2:15so what are we going to do ? Find out
2:17what the activation is . We also need to
2:20model one more thing , which is that the
2:22carpet isn't used at any time ; it's
2:24used at the best possible moment , which
2:26is when you're in a box or a double box
2:28in a party . And what happens at this
2:31moment , when the carpet is active , is
2:33precisely the moment with the highest
2:35density of monsters on the screen . So
2:38the targets when the carpet is active
2:40tend to be higher than the average
2:42targets during the entire hunt . So we
2:44need to model that as well . And why do
2:46we have to model this increase in the
2:48carpet's average damage ? Because the
2:50way we're going to make the comparison
2:51is by using the Tibia Tools calculator .
2:53And Tibia Tools doesn't have a place
2:55for you to put , " ah , I have a level X
2:56carpet and this is my frequency . "
2:59That's all very subjective . I
3:00completely understand why they don't
3:02include anything for the carpet here in
3:04the calculator . However , with our
3:06knowledge , we can adjust some things in
3:09the calculator to be able to simulate
3:11builds with different carpets . And how
3:14are we going to do that ? Well , first of
3:15all , I'm going to equalize the builds
3:16here , right ? So , now they are at zero .
3:18And let's go through a thought process
3:20that is as follows . Let's assume that I
3:23have the carpet 100 % active . Let's
3:26pretend I have a carpet with zero
3:28cooldown . I will always activate the
3:31carpet and my carpet has a 12 % bonus .
3:35What I hope I can do in the calculator
3:37is for build B , which is the build I
3:39will choose here as the one that has
3:40the carpet 100 % active in every turn , to
3:42show 12 % here . And how do I do that ? And
3:46the easiest way to make this appear is
3:48by changing the rotation here . So , if I
3:51take the targets here , which is the
3:53average number of hits per spell per
3:56turn , and increase it by 12 % , I'll get
3:58this number here of 8.96 . And if I
4:02replace it in all of them , what will
4:04happen here ? The number 12 % appears ,
4:06right ? Build B is 12 % stronger than
4:08build A. And that's what we're going to
4:09do , right ? The math is the same .
4:11Dealing 12 % more damage is
4:13mathematically the same thing as
4:16hitting 12 % more mobs on average per
4:20turn . So that's how we're going to
4:22model builds with different carpets .
4:24But now we're going to the classifier
4:25because we need to know the uptime of
4:27the spells . And what did the classifier
4:29say here ? First of all , it found that
4:31the carpet bonus is 12 % . OK ? I already
4:34knew this , but the classifier managed
4:36to find it on its own . Uh , it also says
4:39that the global carpet uptime is 28 % ,
4:41right ? So on average , 28 % of the turns
4:44have a carpet active . But we can't do
4:47calculations with this global number
4:49here . We need to do calculations at the
4:51component level because they have
4:53completely different behaviors from one
4:55another . Look , for example , at this
4:56table . This table here is the number of
4:59components where the carpet was active .
5:03And when these components are active ,
5:05they have these averages that are much
5:07higher than the session averages . So
5:09you look at the attack ; on average ,
5:11it's landing 6.39 hits per turn .
5:13However , when the carpet is active ,
5:15that average goes up to 9.66 , right ?
5:17This makes a lot of sense because the
5:19carpet is being used at the ideal
5:20moment , which is the double box moment .
5:22So there are many more hits , many more
5:24mobs on screen for it to hit . That's
5:26why the average goes up . And the
5:27component uptime is also different . If
5:30I take the number of auto-attacks here
5:32that have the carpet active , 35 , and
5:34divide by the total number of
5:36auto-attacks in the session , we see
5:38that the auto-attack uptime is 23 % . So
5:42only 23 % of the auto-attacks received
5:44the carpet bonus . But the grenade is
5:46different . If we take the grenades here
5:49with the carpet active , it's 12 , and
5:50divided by the grenades in the session
5:52here , which is 18 , it gives 66 % . This
5:54also makes sense , right ? So we went
5:56from 20 - something to 60 . So the grenade
5:59should receive a higher carpet bonus
6:01than the auto-attack because the
6:03grenade's uptime is higher . And the
6:05calculation we're going to do is the
6:07following ; let's use the auto-attack
6:08example . First of all , we need the
6:10uptime . So that's what I already told
6:11you , 35 divided by 152 . We also need to
6:15calculate the ratio of what I mentioned
6:16earlier , which is how much higher the
6:18average hits are when the carpet is
6:19active compared to the total average
6:20hits for the session . In this
6:23auto-attack case , I have to multiply by
6:259.66 and I have to divide by 6.89 . And
6:30then , at the very end , this is just
6:32like the little calculation we did for
6:34critical damage , we have to multiply by
6:36the carpet damage bonus , which in this
6:38session is 12 % . And there we go , we
6:41found how much we need to adjust . the
6:45auto-attack hits to say , ah , this is
6:47numerically equal to a carpet used x
6:50times with a 12 % bonus . So I just do
6:54this here , look , add one more . And now
6:56I take this 3.87 % here and multiply it
6:59by the original targets here , right , of
7:016.89 . And the number that gives us ,
7:047.16 , right , rounding up , is the number
7:07in this column . So the classifier is
7:09already doing everything for us . I just
7:11showed you what the reasoning behind it
7:13is . And if we take this number here ,
7:15right , the original 6.89 and the
7:17adjusted 7.1569 here , we will see that
7:20this generates an average increase of
7:233.87 % . and 3.87 is exactly what we
7:25wanted to appear here . It's the
7:27difference the carpet makes for this
7:29component . And this number here helps
7:31us already answer the first question
7:32from the beginning of the video , right ,
7:34which is how strong is the carpet . For
7:36this component here , which wasn't a
7:38component with a very high uptime ,
7:39right , we saw that the grenade had an
7:41even higher uptime , the auto-attack is
7:43receiving an average damage increase of
7:453.87 % . 3.87 is exactly a tier 8 Fatal .
7:50That Highwin log , the fact that he is
7:53using the carpet at the frequency he
7:55uses it in his build with the 12 % bonus
7:58on the auto-attack component is the
8:00same as if he had a wheel with no
8:02carpet , but had a tier 8 . So yes , the
8:06carpet is very strong . And this is the
8:08thinking we will use from now on so we
8:10can make these comparisons . So we will
8:12always use the classifier , look at the
8:14column of average hits adjusted by the
8:16Gravis Sun , which is the carpet , and
8:18put that into the calculator . Simple as
8:20that . And coming here to the comparison
8:22, uh , of the two Highwin wheels .
8:24Remembering that they are these here ,
8:26which is the wheel with the carpet ,
8:28which in this case was build A , and the
8:30wheel with more damage , but without the
8:32carpet cooldown gem here . This here
8:35became build B. And to show beforehand
8:37that build B does indeed have more
8:39damage than build A when we aren't
8:41thinking about adjusted targets , I'm
8:44just going to take the simulation here
8:46and equalize the targets , right ? And
8:49regardless of whether I copy from A to
8:50B or B to A , it doesn't really matter ,
8:52right ? So , what do we have here ? Same
8:55monsters , same rotation , different
8:57build obviously , and different strength
8:59here . We see that build B is indeed
9:02stronger in raw damage , right ? But then
9:06, when we are making the comparison
9:08thinking about the carpet and making
9:11build A have a slightly higher target
9:13adjustment than build B , because build
9:16A has the carpet gem , this is enough to
9:18make a weaker build become stronger ,
9:21because the Pete bonus is very good ,
9:23and this explains a lot why the carpet
9:26became the favorite perk for high
9:28levels . Velor , for example , uses two
9:31cooldown gems . Most Royal Paladins ,
9:34when they reach a very high level ,
9:36around 2300 more or less , most of them
9:39are using a cooldown gem on the carpet .
9:42But even so , this is for everyone , for
9:44any type of comparison . So if you have
9:47a doubt about a wheel that is wheel A
9:49with carpet one versus a wheel B that
9:50is carpet two , you can do that nowadays
9:52. You just need to get the login , throw
9:55it into the classifier , and then pass
9:57it to Tibatools to make the comparison .
9:59The classifier is public here , right ?
10:01So you can use it here via this URL ,
10:04it's 99 % working . What I mean is that
10:08there are some turns that I still can't
10:11classify , but I continue working to
10:14manage to make all turns of all
10:16vocations of any build possible to be
10:19classified . And that was what I wanted
10:20to talk about in today's video . So ,
10:21thank you very much . Cheers . No.