Zezo forum:

Post by zezo | 2016-12-13 | 12:49:02

I've replaced the long obsolete "Avoid tacks" checkbox with one named "6% extra speed"

It does what it says - multiplies the wind and polar derived speed with 1.06

Leaving the track tool without multiplier so we can visualize the difference.

65 Comments

Post by guigui | 2016-12-13 | 22:07:50

That's a great addition ! Thank you.
And I like your explanation about the track tool not multiplied : it's not a bug, it's a feature ! Hahaha ! Good job !

Post by jfloviou | 2016-12-15 | 01:06:20

I think there is a bug with the 6% applied: the TWA/Heading displayed is wrong, as if the wind is not correct.
To reproduce: compare first step Heading&TWA with and without 6% ticked. it should give the same wind direction but it does not.

Post by zezo | 2016-12-15 | 01:17:30

This is a quirk of the router. It likes to introduce 5 degrees of noise in the first step sometimes (quite often, in fact), no matter if there is a speed multiplier.

There is a heuristic that tries to filter the noise and straighten the track somehow - that's the black line. The red line is the actual result produced. But it's hard to straighten a curved path in constantly changing winds.

I have some ideas for improvement, but have to find the time.

The current logic is that if the change of direction from one step to the next is small they can be considered the same direction. It was good with the wind squares years ago, but does not work very well with fixed TWA now.

The actual output usually looks like this (Considering optimum TWA to be 145)

140
145
145
145
...
So when you check the black dot it says 140, but when you overlay the green line it is close to 145 down the road

Post by jfloviou | 2016-12-15 | 01:32:54

hummm i'm not sure i understand. But it does not happen anymore :-)
And i understand better that using green line is important

Post by zezo | 2016-12-15 | 02:03:36

Yes, the green line is important, and checking the best VMG in the polar chart.

Trying to explain one more time ... At some wind speeds the VMG changes very little between let's say 140 and 145 degrees - the polars are somehow flat. If the difference is very small - like say 0.1% it may get lost in the calculation error, so the router won't have strong preference for the perfect TWA and will chose random headings withing 5 degrees.

That won't matter much in the real world where the wind is not coming from the GRIB direction and is not constant, and the prefect point of sail is not exactly 140, and there are waves to avoid and use, but in the virtual world you notice that the numbers don't add to 360.

I have the idea of adding some heuristic like "if the TWA for the first step is 5 degrees off best TWA and next step is at TWA then adjust the first step" but those kind of things need testing and tweaking.

Post by dede | 2016-12-15 | 07:17:50

Hi,
Perhaps a stupid idea but isn't there a link between the 6% speedboost and the 3' ZEA warning zone?

By the way CVetan, could you explain us when and how to use this new feature?

Post by jfloviou | 2016-12-15 | 15:46:26

Hello,

It seems the routing does not work after 3 days right now. The route and isochrons stop.

Any clue?

Post by zezo | 2016-12-15 | 15:54:25

You should be able to select up to 8 days in the drop-down in the upper right corner. Or 10 days when the server is not too busy, but the routing time is never forced below 8 days.

Edit: OK, no forcing of a limit now. You can choose up to 10 days whenever you like, but I may return the limit if the server overloads.

Post by Inicio | 2016-12-15 | 16:19:20

As you say the green line is very important. Also the red one.
But if they not indicate -as now- the 6% speed bonus, they are not useful now.
We realized the difference.
Perhaps it's time to include de 6% bonus in the track lines? I need it! :)

Thanks a lot another time for your great work!

( for me, the 6% bonus is present every time. At times, more than 6%...)

Post by zezo | 2016-12-15 | 16:29:17

You can't always get what you want ... but I'll make an exception this time ;-)

Post by Inicio | 2016-12-15 | 17:55:27

:))))
Are you Santa Claus, isn't it?
I have other Christmas wishes!!!!

Post by jfloviou | 2016-12-15 | 20:41:50

What is your strategy to check the 6%?

Post by Inicio | 2016-12-16 | 00:36:16

Locate the wind measurer in front of the boat. See the advance in 10 minutes. Calculate.

Post by whereami | 2016-12-16 | 08:40:08

I think it's more obvious that that: Compare the speed shown to the speed the polars say you should be going. If it shows you're going faster than the polars by 6%, then there you go. I'm in that state right now, at least. Says I'm going 20.1 kt in 24.5 kt winds at 150 TWA, whereas the polars say I should be going 19.5 kts. Sure enough, that's 6%.

Post by Beaware | 2016-12-16 | 14:41:34

Avec quelle voile, parce que avec le Grand Gnk. c'est bien 20.1 noeuds ... Voiles pro ou pas voiles pro ?

Post by Inicio | 2016-12-16 | 23:14:12

The speed shown in the game is exactly the speed the polars say.
But in fact the boat is a +- 6% faster

Post by zezo | 2016-12-17 | 00:05:23

I've seen the polars differ, but that was probably a bad bug and was fixed.

What differs client-side and sever-side seems to be the wind. I is very obvious if you compare the info about another boat with the wind information at the cursor. Part of the difference can be explained by the 10 minute time difference and possible boat position difference, but not all of it.

One way to see what's going on is watching your boat from another boat/account. Another is to watch if the projected track moves ahead in time by placing a scheduled point and comparing it against the red line or against the wind meter as Inicio suggested above.

Post by Isatis | 2016-12-17 | 12:22:54

Hi and many thanks for this great soft.
Actually, i don't see any chackbox named 6 % extra speed. i just see a link wich point on this forum.
I don't see also any green line between two points in the router. I see only red and black line.
Did i make any mistake or bad understanding ?

Post by zezo | 2016-12-17 | 13:24:21

The checkbox should be in the right mouse button popup menu.

The green line we talk about is the dynamic track projection linked to the mouse cursor. Also the blue line.

Post by Isatis | 2016-12-17 | 15:23:29

OK, understood.. Many thanks again ;)

Post by marcusbelgicus | 2016-12-17 | 12:55:23

That's a while I have not seen extra speed on the boats in the neighbourhood. Thompson is alwost behind the 1st ones, maybe VR has made some adjustments and they don't apply the extra speed now. Did you still notice extra speed?

Post by zezo | 2016-12-17 | 13:29:56

Right now for my boat wind speed 22.7 boat speed 12.9. From another boat it says wind 23 boat 13 kts.

One step later it's 22.9/13 vs 23.1/13

It could be just a difference in the rounding of the values or could be something else. No way to tell what's going on server-side.

Post by LJadra | 2016-12-17 | 19:34:09

Since I read about 6% speed adjustment on this forum a few days ago I calculate distance (between one or more 10 min periods) a few times per day (with great circle calculator using my boat coordinates) and results are truly strange. Differences in displayed and calculated boat speed are somewhere between 2% and 13%. In different wind speeds and TWA. Yesterday it was more than 1,0 kt difference in two consecutive 10 min. cycles (without maneuvering).

Btw what is the penalty for gybing or tacking?

Thank you Cvetan for all the good work.

Post by zezo | 2016-12-17 | 19:51:29

The last official information I'm aware with is 10% speed penalty within one 10-minute cycle. Actually they say it's 50% penalty for two minutes which may matter with sub-cycle updates.

Please correct me if there is new info posted in the official game forum.

Post by guigui | 2016-12-18 | 17:50:01

Hi Cvetan,
Is it possible to make the 6% option available on the mobile version of your fantastic tool ?

Post by zezo | 2016-12-20 | 12:52:13

It's there since VOR in 2013 or so.

Post by marcusbelgicus | 2016-12-19 | 22:55:54

Cvetan, just realized that you made a version for smartphone. Waw! That's great!
Regarding the bonus, sometimes, the routing with the 6 % bonus can be completely different than without, which is quite interesting information as well since it gives some information if the route is risky or not. If both routing are different, it is probably better to take the less risky of both.

Post by jfloviou | 2016-12-20 | 04:24:36

Damn, in the days to come, if nothing change, it will important to decide if we should use 6% or not. The routing is really different...

I still do not manage to find a good way to know :-(. Globally i have the impression that we are faster than polars. Also we are wining some ground to the real skippers, where in the beginning they were WAY faster: another sign that a boost is applied.

Post by Beaware | 2016-12-20 | 12:18:25

A mobile version ? Android ? Where is it ?

Post by zezo | 2016-12-20 | 12:55:13

No. Just the chart should be mostly usable on a mobile device. There are three main differences:

1) Touch scroll
2) A cross in the middle of the screen, with coordinates pointing to the cross
3) A Menu drop down instead of right click. There is also a menu entry called "show first info box" so you don't have to tap the first black dot to see the numbers.

Post by Beaware | 2016-12-20 | 13:27:49

Excellent, thank you very mutch

Post by zezo | 2016-12-20 | 12:59:21

But it changed ;-)

The decision point was (and still is) is about 24 hours ahead, and probably there will be no second option at the time we get there.

That's what I mean when I say that differences in weather are bigger than the differences in boat speed.

Post by jfloviou | 2016-12-20 | 04:52:00

As of now, i can see an 8,2% bonus...

Post by karriv | 2016-12-20 | 11:20:54

I see a delta of 10,6 %.

Post by karriv | 2016-12-20 | 11:48:00

And now it is to 5,6 %...

This sucks big time.

Post by zezo | 2016-12-20 | 13:01:17

It would not be that bad if the process is transparent. In the real world your boat would not follow the grib model anyway. But I'm not sure it's intentional. Anyone got an official VR position on that issue?

Post by jfloviou | 2016-12-20 | 13:09:02

cvetan,

I think i've discovered something... From my recent calculations, i do confirm that 6% is exactly the correction. The distortion is coming from the way VR calculates bearing and distances: not the same than it used to be, and not the right one! (approximation valid in short dist, not long)

I could drop you a private message if you want.

Cheers.

Post by zezo | 2016-12-20 | 13:33:08

That would be nice (zezo@zezo.org). I won't fix the routing to match every VR quirk, but would like to know more what they are doing (and doing my own research, but it's not top priority).

Post by jfloviou | 2016-12-20 | 13:12:14

BTW, my routing today is proposing to go into Magellan strait!! It means your algorithm is almost perfectly handling obstacles! This is amazing! I will compare with professional routing system: i'm not sure they handle it as good!

Post by zezo | 2016-12-20 | 19:32:39

At that distance it tends to jump over quite large land masses - the obstructions are checked only for calculated points, and those are 1 hour apart in the 3-day routing, so the boat can jump over 20+ nm of land.

And the algorithm does not handle the 'behind the corner' routing very well. But it's usually good enough because you get to the corner sooner or later and it corrects itself.

Post by marcusbelgicus | 2016-12-20 | 13:48:45

jfloviou, did you notice that there were variations in the speed bonus, I have noticed sometimes 10 % but other times 0 %, the average might be 6 %

Post by jfloviou | 2016-12-20 | 14:24:24

Marcus: i was thinking the same, but now i think it is steadily 6%. I need to discuss with Cvetan to confirm my assumptions.

Post by karriv | 2016-12-20 | 17:41:46

cvetan, marcus, jfloviou,

I was following the speed bonus as follows:

I put the wind indicator at the 1 hr mark to get the distance. Then for each iteration I recorded the display speed on the UI (equals to what the UI polar gives) and new distance to the wind indicator. After that I calculated travelled distance per iteration and consequently speed per each iteration. I got the following series:

10,61 %
10,61 %
10,61 %
7,26 %
5,59 %
5,59 %
2,23 %
2,23 %

Now, those numbers seem systematic, don't they? Well, they follow a formula:

SpeedBonus = 0,55866 % + n * 1,67598 %

What determines n, god only knows...

It is slightly disturbing that the whole series is decreasing, I would like to understand why. I will collect more samples.

Edit: I corrected the numbers, I had copy/pasted the wrong number to be multiplied by n.

Edit 2: OK forget about the numbers, turns out my speed during the observations was a constant 17,9 kn and 1/17,9 = 0,055865922, so each variation of 0,1 nm in the traveled distance results in a speed delta of 1,68 % (=3*0,557)...

Post by karriv | 2016-12-20 | 17:52:16

On a different note, I think it would be important that game rules are clear to everyone. On the other hand, one of the rules can well be "you need to figure out the rules", which would be fine with me (actually I would like that very much). The problem is that the organisers don't say anything clear about the rules.

Post by marcusbelgicus | 2016-12-20 | 20:36:25

@ jfloviou, thank you.
@ karriv, they talk about the 'rafales', which I believe is turning into this speed bonus, so in some way they are talking about something not clear which impacts the speed.

Post by guigui | 2016-12-20 | 21:43:41

I come back to my request : Any hope to have the 6% speed increase in the mobile version ?
I will be away for the holidays with no access to a PC and I will be lost without the 6% thing.

On the 6% itself :
On VRTool, to get a good match inbetween the planned routes and reality, I have had to change the polar files by mulitplying every values by 6.8% (value found by somebody on the VR forum when the phenomenon started)
Now I have a very good match. Almost perfect when the winds are the same in both game and VRTool.
I think the differences I get come from the wind interpolation when the gribs are updated, for like an hour or so and when the winds are changing a lot in time and position.

TLDR : 6.8% is good if the winds are good.

Post by guigui | 2016-12-21 | 20:07:53

Arf, forget what I wrote : I was at 0% for the last 5.5 hours. Like perfectly 0% : if I plan a route from where I was to where I am now, with the starting time adjusted to match when I was, with the original polars, I get my position now.
So I don't know...
At least, I know that I don't know. It's something :)

Post by marcusbelgicus | 2016-12-21 | 20:56:44

I am still convinced that sometimes this is 0 %, sometimes 10 %, sometimes other speeds.

Post by nicolasdb | 2016-12-21 | 23:55:27

Yes it changes a lot.

Last night 0%, this morning 10% then back to 3% up again to 6/7%.
This afternoon it was back to 9% and now it's 3%...

Average was about 6%, it seems not to be the case anymore.

Post by jfloviou | 2016-12-22 | 03:23:26

Well I thought i had understood something, mainly that VR is not using Haversine / Great circle distances, but instead some Rhumline distances, which changes depending on the latitude...

But in fact it still changes too much.

Also, globally nothing explains why, not even speaking about speed and distances, the bearing is wrong. For example right now, i'm bearing 106�, not using the regulator, and whatever is the method i use, the bearing is 106,5� (approx).
This cannot be explained by a change in wind as well.

=> Simply a massive bug.

Post by nicolasdb | 2016-12-22 | 15:45:01

Coef is 0,967 for me now... Careful, they are slowing the race !

Post by ArchivisteM7 | 2016-12-23 | 17:28:12

Bonjour,

Thank you very much for this work.

J'ai tent� de lire, mais je ne comprends pas bien le fonctionnement de cette option "6% speed", je ne vois pas de diff�rence dans les vitesses affich�es dans les rectangles blancs.
Quelqu'un pourrait-il m'expliquer en fran�ais, merci d'avance.

Post by jfloviou | 2016-12-24 | 00:11:41

En fait lorsqu'on regarde deux positions successives, et qu'on calcule la distance parcourue en 10 min, on trouve que la vitesse affich�e sur le jeu est inf�rieure � la vitesse "r�elle" du bateau. Par exemple si le jeu affiche 19,6 noeuds sur l'interface, en r�alit� si tu regarde la position apr�s 10min, tu verras que la distance est sup�rieure � 19,6 / 60 * 10. (je divise par 60 pour trouver la dist / minute, puis je multiplie par 10 pour avoir sur 10min).
Comble de la difficult�, ce coefficient multiplicateur varie... En moyenne il semble qu'on va environ 6% plus vite.

Post by ArchivisteM7 | 2016-12-27 | 12:53:01

Merci jfloviou pour ta r�ponse mais je voulais surtout savoir comment fonctionnait cette option sur ZEZO car avec ou sans option la vitesse affich�e est la m�me. Et je me demande sur les lignes vertes et rouges tiennent bien compte du bonus ?
Si quelqu'un peut me r�pondre merci. (Sinon je suis d'accord pour le coeff j'ai aussi remarqu� des �carts importants m�me sur une seule it�ration de 4% � 8% m�me plus et le 6% est une bonne moyenne)

Post by Manu-PB | 2016-12-27 | 13:26:40

Cocher / decocher 6% modifie les isochrones... ainsi que la route optimale (y compris le cap a suivre les premieres heures...).
Les ortho et loxo tiennent compte du cap, donc sans doute aussi de la vitesse.

Post by Beaware | 2016-12-28 | 17:35:16

Do we always increase by 6% ? Is it always necessary to check this option ?

---------------------

Avons nous toujours l'augmentation de 6 % ? Est ce toujours n�cessaire de cocher cette option ?

Post by mak08 | 2016-12-28 | 18:20:53

The increase varies.
If the wind tool method is telling the truth, it's currently at 13% (and 30min ago it was 9%).

Edit: Looks like waiting only one cycle is inaccurate. The current increase is more likely 10%.

Post by zezo | 2016-12-28 | 18:24:51

It's not a fixed value. I suspect the real difference is in the wind forecast used by the game engine and the interface, and can be positive for some boat and negative for another at the same time.

Post by jfloviou | 2017-01-01 | 21:55:20

Hello Cvetan,

To me it looks like VR has now reduced the speed coeff since a few days. I assume it is because we are now ahead of Le Cleach.
I'm now steadily underperforming compare to the polars/router. -3% in average.

=> might be worth proposing a new coeff if others confirm the same tendancy ?

Post by jfloviou | 2017-01-02 | 19:19:12

Damn, I'm (again) wrong: it is back to 6%.

Maybe in light winds VR winds are lighter, and in medium they are stronger?

Post by Isatis | 2017-01-06 | 14:26:40

Hello all, and thanks again to Zezo.
It seems to me, that the 6% box change the router, but doesn't change the boat speed the router shows..So the only way to know if the 6% is in use in VR is to compare distances in VR and in the router.
Is it right ?

Post by PalaEksa | 2017-01-06 | 17:42:32

Hi ! Thanks a lot for such a great job!
Is there a way to get the 6% speed vonus option when using zezo on an android smartphone ?
Thank you !

Post by guayacan | 2017-07-30 | 17:32:26

Hello,
on the current Atlantic record, WITHOUT the "6% extra speed", zezo estimate the arrival in 66.5 hours, trying WITH the "6% extra speed", the remaining time is 68 hours !!! 1.5 hour more ; can someone explain this strange behavior ??

Post by zezo | 2017-08-03 | 17:43:23

Thanks. The configuration for the atlrecord was wrong and the 6% speed option did something else. Fixed.

While it's possible not to get faster crossing with faster boat if you are riding a low pressure system and can't get ahead of it, the faster boat should not take more time.

Post by GeGaX | 2017-08-03 | 09:00:07

Hi,
Surely because you go faster than the wind of the blow you lose the advantage of being constant and you must therefore wait again the wind to advance