Facebook

Login

Support Sailonline

If you haven't already - join the SAILONLINE YACHT CLUB!

Please also consider making a donation - all amounts are greatly appreciated!

Board » Technical Discussion » Performance loss

Page: First Previous 7 8 9 10 11 12 13 Next

Posted by
kroppyer


Users avatar

Posts
347
Registered at
2010-02-05 15:25:55
2014-07-18 15:02:50 - Re: Performance loss
rod,

Starting at 100%, a loss of 8% is achieved by

With tack or gybe:
Tack to an angle that gives sog=16 kts, more speed is more performance loss. (the angle depends on polar and tws)

Without:
Turn 110 degrees. Greater angle is more performance loss.

For more complicated things (not starting with 100%, or doing more than one manouvre) please refer to the PDF I posted earlier in this thread.

--- Last Edited by kroppyer at 2014-07-18 15:04:28 ---
Posted by
dtayls


Users avatar

Posts
30
Registered at
2014-06-23 03:34:16
2014-08-15 04:30:08 - Re: Performance loss
Wow! Great thread. I've raced for almost 40 years and I'm an engineer & a computer scientist, so this one's right up my alley.

Having just read through the whole thread here are a few thoughts from me is not particular order:

1. SOL started with a realistic model and that should probably remain. Anything that breaks with realism should be explained, at the very least.

2. There are good real-world reasons for deaparting from the realistic model. Server load and reasonable sample rates need to be considered.

3. Different yachts recover differently so how you should sail each yacht should be different.

4. The world is changing quickly. Downwind-flyers come on and off the plane and their performance changes massively when this happens. If a boat is only just on the plane then a gybe will be costly.

5. And don't even talk about foiling and whether you can keep it up during a tack/gybe!!!

I suspect the SOL models are based on displacement vessels. It's probably time for an update particularly given that we're playing with some pretty impressive craft these days.
Posted by
dtayls


Users avatar

Posts
30
Registered at
2014-06-23 03:34:16
2014-08-15 06:24:08 - Re: Performance loss
I should also mention that I was impressed by some of the analytical work. Excellent stuff.

This reminds me of a marketing subject I did around 20 years ago at university. A key component was a computerized marketing strategy game.

In our group we analyzed the 'black box' of the game so well that after the results were published the lecturer bailed me up and accused me of cheating ...

Let me just say "I like your work."
Posted by
kroppyer


Users avatar

Posts
347
Registered at
2010-02-05 15:25:55
2014-08-15 09:45:59 - Re: Performance loss
I made this yesterday:
Sailonline Performance Calculator
http://sol.kroppyer.nl/performance/

It's about the same as the spreadsheet I posted earlier, but this one is more precise. Only time lost to next DC is not always accurate (because there is no clear definition for "time lost").

dtayls, funny that you read and posted on this thread just after I finished the calculator. Of course a new model would still be simplified reality, heavily simplified. But it's not hard to make a model that's more realistic than the current one, with the current model I fail to see the relation with reality.

You might see some worked out propositions for new models over here fairly soon.
Posted by
dtayls


Users avatar

Posts
30
Registered at
2014-06-23 03:34:16
2014-08-16 01:24:32 - Re: Performance loss
Kroppy,

Some SOLers are sailors while others are not.

It is probably true to say that most sailors want the model to be ultra-realistic while SOLers just want a model (aka. rules of the game) that they can understand and play against.

I personally think there is a middle ground that will probably work for everyone ... if implemented and explained properly.

The 15 second sample/update interval is an interesting aspect of the game. I am sure there are many reasons for this - server load and UI development requirements are just two that come to mind. But let's face it, this is not an online shoot-em-up game and people should probably not be trying to sail these things in real time. And if you want to play with a high performance yacht and really know how to tack/gybe on the plane/foil then get out on the water and do it IRL!

But that said, I think there is a lot of value to be had in changing some parts of the model and most of these are probably quite straight forward to implement.

Dtayls.
Posted by
HappyHour


Users avatar

Posts
27
Registered at
2009-01-08 06:48:51
2014-10-02 01:47:42 - Performance loss with high speed boats
A performance model hasnt to be realistic but it should work.
Our current model works with slow boats ( max speed 15kn ), of course far from perfect but it works.
It doesnt work well with faster boats (max speed 15+ kn).

With a AC72 it just gets crazy.
For example:
Gybe sailing downwind with 40kn speed at 92% performance, hurrah, no performance lost.
Doing the same at 94% performance, oops, performance drops to 74%.
Thats not just unrealistic. thats stupid!

Isnt the 93% rule outdated, a rule from history when the max speed was around 14kn?

Why dont we drop the 93% rule, let all performance calculations as it is and set hard limits, like max perf loss 90%.
Maybe even better:
Divide our fleet into "slow" and "fast" and set differrent max performance lost limits,
like 88% for the fast boats and 93% for the slow boats.
IMHO it should be possilbe to implement these changes to our server software.
Just an idea.
Posted by
kroppyer


Users avatar

Posts
347
Registered at
2010-02-05 15:25:55
2014-10-02 08:48:53 - Re: Performance loss
I fully agree that performance loss needs an update. To say the least

I really disagree on the idea that performance should be different across the fleet. First of all I like the beauty of everyone is exactly equal too much. Second, why would the people with less practise be faster in manoeuvres than the people with more practise?
[edit: I misunderstood the idea]

---
If it's going to happen, should we try a "do it once and to it good" approach, or should we spend some time on an intermediate performance system as "quick fix"?

--- Last Edited by kroppyer at 2014-10-02 15:55:39 ---
Posted by
HappyHour


Users avatar

Posts
27
Registered at
2009-01-08 06:48:51
2014-10-02 12:43:18 - Re: Performance loss
when i was talking about fleet, i didnt mean the boats in a single race.
I meant the diff between a "slow" yacht like first 40.7 and "Racing machines" like VO4 or our latest baby AC72.
Our fleet of available boats. From 6.5m upto 90m.

--- Last Edited by HappyHour at 2014-10-02 12:48:49 ---

--- Last Edited by HappyHour at 2014-10-02 14:04:02 ---

--- Last Edited by HappyHour at 2014-10-02 14:06:05 ---
Posted by
kroppyer


Users avatar

Posts
347
Registered at
2010-02-05 15:25:55
2014-10-02 15:54:33 - Re: Performance loss
Ah! I misunderstood. Now I fully agree: the performance system should depend on the boat type. A very simple solution would be as I described a few posts back and use the weight of a boat for the performance calculation.
Posted by
HappyHour


Users avatar

Posts
27
Registered at
2009-01-08 06:48:51
2014-10-07 15:56:46 - Re: Performance loss
"or should we spend some time on an intermediate performance system as "quick fix"?"

If some time means some hours, we should do the "quick and dirty fix". As i said in my previous posting, let all the calculations as it is and set a hard limit for the max. performance drop. that would at least eliminate the worst impacts of our current perf system.

Today i sailed 2 runs with a fast boat in the Atlantic Circle and tried to beat the performance drop system and it seems to work. A 3 step tacking, like from 45 twa to -5 twa, -20 twa and -45 twa may be faster then a normal tacking from 45 twa to -45 twa. The same with gybing. Gybe to the slowest possible speed to minimize the performance loss. Anything will be better than a 34kn speed to 34kn speed gybe with a 17% perf loss at 94% performance.

A nice tactical aspect, but im sure this isnt what the developers had in mind when they wrote our current code.


--- Last Edited by HappyHour at 2014-10-07 16:23:31 ---

Page: First Previous 7 8 9 10 11 12 13 Next

Please login to post a reply.

Races

Next Race: 00d 00h 00m

Current Races:

Roslagen TIMED Sea Race

Welcome to quite possibly the most challenging TIMED race of the year. This month finds us in Sweden with lots of islands and narrow passages when we sail our Seacart 30 trimaran over the 86nm from Lidingö outside Stockholm to the beautiful coastal village of Öregrund. Consider yourself a winner if you complete the course nonstop! This is a TIMEDrace so you may RE-REGISTER HEREto try again after finishing a run. You will have 13 days and 11 hours to test your skill and decision making after the race opens.
Race #1945
INFO by brainaid.de
Seacart 30 Particulars
WX Updates:
0430 / 1030 / 1630 / 2230
Ranking:
TRQ3 - TRCH - SUPSOL - SYC
Race starts: Sep 07th 12:00 Registration will open soon
▶ Flash
GO TO RACE

The Vineyard Race 2025


For the first time since first hosting a virtual Vineyard in cooperation with Long Island's Stamford Yacht Club back in 2010, due to an operational calendaring error, Sailonline has missed the start of the real life race, the USA’s Labor Day weekend this year unusually falling across the turn of August. Silly SOL! So, a week late please join us anyway for a virtual rerun of Stamford’s iconic annual overnighter from their club line out the sound to Buzzard's Bay Light near Martha's Vineyard Island and back again, leaving Block Island to starboard. This year, Sailonline is making its magic MC 38 available to competitors, a race boat with a quirky polar that will require you to stay wide awake as you speed around the c 250 nm course, in rockstar style!
Race #1946
INFO by brainaid.de
MC 38 PARTICULARS
NAM_AWIP WX Updates:
0245 / 0845 / 1445 / 2045
Ranking: SYC
Race starts: Sep 06th 17:05 Registration Open!
▶ Flash
GO TO RACE

Great Nicobar to Marion Island 2025

The Adventure Continues – Leg #6 of the RWW Series.
We’re setting sail on the next stage of the RWW Series: a 4,400-nautical-mile passage from Great Nicobar to Marion Island. This demanding leg will test endurance, strategy, and seamanship at the highest level. As part of the 2025 Ocean Championship, this marks the ninth legendary challenge in the series — and one of the most exhilarating yet. The race will be contested aboard the Archer 78’ ketch, a vessel built to push sailors to their limits across the vast Indian Ocean.
PRIZE: SMPF
Race #1942
INFO by brainaid.de
Archer 78 PARTICULARS    
WX Updates:
0430 / 1030 / 1630 / 2230    
Ranking: OCQ3 - RWW - OCCH - SUPSOL - SYC
Race starts: Sep 01st 11:00 Registration Open!
▶ Flash
GO TO RACE

Africa by Sea 2025 - Maputo to Mogadishu


The fifth leg of our exploration of Africa By Sea takes us north along the length of the Continent’s succulent sub-tropical and tropical Indian Ocean coast. Much fought over by traders and colonists both well-before and after a Portuguese expedition under the command of Vasco da Gama first sailed there from Europe back in the 16th century, it is yet to be properly explored as a yachting destination. With plenty of islands off- and inshore, small and very large, there is much to interrupt your 1900nm voyage and visit in your Swan 65, from Maputo to Mogadishu! Race #1858
INFOby brainaid.de
Swan 65 PARTICULARS
WX Updates:
0430 / 1030 / 1630 / 2230
Ranking: SYC - ABS
RACE CLOSE: Saturday,
September 6 at 2300 UTC.
Race starts: Aug 20th 17:00 Registration Closed
▶ Flash
GO TO RACE

Go to race archive

SYC Ranking

  1. Sailonline Yacht Club Member WRmirekd
  2. Sailonline Yacht Club Member CriticalHippo
  3. Sailonline Yacht Club Member vida
  4. Sailonline Yacht Club Member rafa
  5. Sailonline Yacht Club Member KaSToR
  6. Sailonline Yacht Club Member Kipper1258
  7. Sailonline Yacht Club Member BRENTGRAY
  8. Sailonline Yacht Club Member Sax747
  9. Sailonline Yacht Club Member CollegeFund
  10. Sailonline Yacht Club Member Panpyc

View full list

Series

Mobile Client

SYC members have the benefit of access to our mobile/lightweight web client!

The mobile client