Zezo forum:

Post by zezo | 2009-08-29 | 14:41:19

This thread is open for technical discussion related to the Istanbul race.

I'm working on it, but full support is impossible before the official game start, so we have to wait few more hours.

40 Comments

Post by zezo | 2009-08-29 | 14:42:06

You can study the polars in the meantime.

Quite complex dependencies at certain wind speeds - 5 different sails at 18 Kt.

Ugly crossover points too.

Post by BennyLava | 2009-08-29 | 18:06:51

Winds on the charts seems not to be in sinc with the game?
Don't currents in this leg, yet?

Post by BennyLava | 2009-08-29 | 23:37:38

The chart seems to indicate an extra wind box? This may cause a wrong optimalisation?

Post by zezo | 2009-08-29 | 23:44:23

It's unrelated. The chart is leftover from the Figaro race, but the optimizer uses only 1 degree winds. I'll get rid of that, but it takes time to get in sync and re-render.

Post by BennyLava | 2009-08-29 | 23:54:12

Thanks! Just that it is a little confusing for my simple mind.

Post by BennyLava | 2009-09-01 | 15:52:03

Cvetan, is it correct to say that the forcasted winds are still not in sync with those of the VR site? Therefore we cannot use the optimizer?

For instance in my sector the wind for next 12h is forecasted as 3,2/87degr, whilst on the website it's forcasted as 4,3/94degr.

cheers

Post by zezo | 2009-09-01 | 16:35:27

In this case the game is wrong. See the posts below.

Post by BennyLava | 2009-09-01 | 20:16:37

So we cannot use the optimizer. Not sure in this case what's wrong, as the game determine which wind to use. Isn't it possible to get it in sync?

Post by zezo | 2009-09-01 | 20:51:53

Read more carefully. The optimizer is more useful this way than if it was in sync with the game forecasts. I would not make 4 hour advantage on Frogeater & co if this was not the case ;)

Post by BennyLava | 2009-08-29 | 14:47:28

Thanks for the ongoing support Cvetan. Keep you posted about my experiences.

Post by zezo | 2009-08-29 | 14:50:24

It seems so far that the wind will be in the good old 1-degree grid so we may get the accuracy back without much effort ;)

P.S. But there may be currents that will bring back some uncertainty before everything gets in sync (mostly before the game get in sync with itself). I've seen currents in the Fastnet only once and that was well after most players have finished.

At least currents don't affect the long range forecasts - they are relevant mostly to the present optimum course.

Post by Fuga9 | 2009-08-29 | 16:38:54

Could you give the file .csv for the polars?

Post by zezo | 2009-08-29 | 16:55:18

Yes, but as you know they are not exactly in the format required by VRTool.

http://zezo.org/ier/classic.csv
http://zezo.org/ier/pro.csv

Post by zezo | 2009-08-29 | 17:16:24

The new polars are included in the optimizer. Winds are assumed to be with 1-degree resolution, but the display is still 0.5 degree - will fix it after the game starts.

Post by Fuga9 | 2009-08-29 | 17:17:44

ok thanks my friend...

Post by zezo | 2009-08-29 | 18:16:00

The winds are totally different. Remains to be seen if it's the usual screwup at the start of the race or something deliberate.

Let's wait for the next wind update.

Post by BennyLava | 2009-08-29 | 19:40:57

the wind boxes seem to be much bigger than before

Post by zezo | 2009-08-29 | 22:29:43

It's been this way in most races. High-res data (0.25 degree) is not available on global basis and it was used only for races around France/UK.

Post by Fuga9 | 2009-08-29 | 20:10:58

Cvetan,
...the velocity of the boat is not right for the polar pro.csv...

Post by zezo | 2009-08-29 | 22:31:04

Yes, way off. At least the winds are back in sync with the reality - so it's the usual stuff after all.

~~I'm a bit tired now but will get the polars in sync tomorrow.~~

Stupid error on my side. The GUI was using old polars (TDF) but that did not affect the optimizer. Should be OK now.

Post by Fuga9 | 2009-08-29 | 23:01:43

ok thanks a lot...
and the winds are great...!!!

Post by zezo | 2009-08-29 | 23:10:29

Just a side note: MP have forgotten the Black sea on the map. Good thingt the race did dot start at Atakoy marina ;)

Post by zezo | 2009-08-30 | 09:25:18

Wind puzzle solved: The game uses GFS 0Z+9h instead of 0Z+6h again. That was the case during the TDF, and was fixed for the Figaro. But that one probably used different data source / processing scripts.

The optimizer uses the 9h data right now, but the displayed winds are the 6h ones.

Post by pcayrol | 2009-08-30 | 11:32:01

Hello,

I think that the winds are wrong on +12H. Someone can confirm ?

Paco

Post by zezo | 2009-08-30 | 12:16:56

Yep. That's related to the morning issue.

What is displayed in the game is the 21h UTC forecast, but what will come up tonight will be the 18h UTC forecast.

So it should be a bit more probable to see the weather from my +12h info (it will be updated one more time in 2 hours with the next GFS too)

Post by BennyLava | 2009-08-30 | 13:36:12

Cvetan, I guess the program has some difficulties to find the most optimum route, probably due to the many islands?

Post by zezo | 2009-08-30 | 13:44:57

Mostly the islands, helped by the weak winds expected along the shortest route (37N /22-23E)

Setting the destination somewhere around Kythira seems to work most of the time.

With more than 1-day forecast duration the algorithm starts sailing over the islands too.

Post by pcayrol | 2009-08-30 | 13:44:43

Sorry but my english is poor and I didn't read all the topic. :-(

So do you can fix this problem ?

Paco

Post by zezo | 2009-08-30 | 13:55:11

Current (0h) winds should get fully in sync with the next update tonight.

12h forecast will be different because the game displays incorrect information - players are complaining in the forum all the time but nobody cares.

Post by pcayrol | 2009-08-30 | 14:41:50

You are thinking that the official site show wrong winds ?

Post by zezo | 2009-08-30 | 14:56:45

I'm sure about that - have spent a lot of time studying the problem. It's wrong if you compare it with UGRIB (the original GFS forecasts used by the game).

Post by pcayrol | 2009-08-30 | 15:07:58

So the winds I am using with VRTool are wrong too after 12H ?

Post by zezo | 2009-08-30 | 15:14:59

It's a bit confusung, but I'll try to explain it again:

If you download the GRIB files after the morning update you can see that the current winds correspond to the 9h GMT forecast and the +12h winds are the 21h GMT forecast.

But if you look at the GRIB files after the evening update you can see that the current wind corresponds to the 18h UTC forecast instead.

3 hours may not matter too much in the middle of the ocean, but with breeze circulation as it is now in the Aegean 3 hours at 20h usually make difference between 10kt and no wind at all, or a huge shift if the meltemi is blowing.

I've experienced that first-hand three weeks ago while sailing roughly in the same area ;)

Post by pcayrol | 2009-08-30 | 15:17:37

Ok I got the 18:00 UTC updates and it didn't look like the game prevision...

Thanks for detect this.

Paco

Post by Belladonna | 2009-08-31 | 12:22:46

Cvetan,
it looks like you have done a very good operation during last night ! Do you have a secret weapon ? :-)))
Bernard

Post by zezo | 2009-08-31 | 14:16:19

Local knowledge combined with proper forecast always gives advandage.

Now the gamble is going South or North of the next low pressure system. Someone with local Ionian sea experience? ;)

Post by Belladonna | 2009-08-31 | 22:35:35

Yeah, but the big bird is gonna be eaten by the frog (I hope legs) eater. Nothing to do against pro sails. Real sportsmen make it raw !!!!

Try hard, Cvetan.

Post by zezo | 2009-08-31 | 23:15:57

It's too tiring to be in the lead and that's the reason for lack of pro sails. He is going to pass me sooner or later and it makes no sense waking at odd hours to steer the boat. Much more relaxing experience this way - becoming 2nd in the Tour was enough for me ;)

Not to mention that maintaining this site as public service is incompatible with playing the game myself. No fun being second to one of your users ;)

Post by Fuga9 | 2009-08-31 | 13:30:55

Good Job Cvetan,...
Everything is working well...
The prediction of 12hs ( one hour before) too..1,5k..i did not trust...now i will trust!!! Good winds...its yet a long way...!!!
Regards from Brazil.