Zezo forum: What point or line is being used to determine the rankings?

What point or line is being used to determine the rankings?

Post by BooBill | 2020-11-19 | 03:56:52

Dashboard reveals an invisible timing gate running between Natal and Harper, Liberia.

In their latest commentary though, VR says proximity to Cape Good Hope is determining the rankings. That does seem to be the case. The guys hugging the African coast are still listed as leading, because their eastern position makes them closest to the Cape. However, in distance to the gate, they are miles back and of course, we all know (maybe they don't) they are out of this thing.

The invisible gate would appear to me to actually be a fair gauge of progress. So, if it's not being used, what's its purpose?

60 Comments

Post by Scup | 2020-11-19 | 06:37:11

Good question! I've been thinking the same thing.

Post by zezo | 2020-11-19 | 11:17:20

It's a big mystery. Ranking is definitely not distance to the gate line. Maybe a distance to the main buoy. At least now the rankings don't jump around like they used to.

Post by B210 | 2020-11-19 | 13:50:04

A few days ago I also tried to find out more information how VR calculates the ranking. By selecting a bunch of boats which are close togetter in the ranking (say positions 5000 to 5020, 10.000 to 10.020 and 50.000 to 50.020), on the VR display these sets form nicely (all parallel) straight lines, roughly from SW to NE. To me this sort of ranking rather seems to be based on a equidistance to a (another parallel) line (gate) accross the Atlantic (I didn't know about the Natal/Harper reference yet then). But if the Natal/Harper line would be the reference, then the boats near Africa would not figure in the first 10000 anymore by now. Neither does the Natal/Harper line seems parallel with the lines I found for boats with similer ranking.

Anyway, whether such a long gate line is a good approximation for equal distance to the finish is another question. As this certainly doesn't correspond to the same distance-over-the-globe to Cape of Good Hope which technically/theoretically should be a better reference point, but this would rather give curved equidistance lines (I guess... as on a flat projection map of a globe that is tricky to judge).

So I do believe that current rankings are based on a reference gate line over the Atlantic (and not to a reference point like the CGH), but very unsure it would be the Natal/Harper line, and furthermore also not sure if such a line-based ranking is a good reference for the real equidistance to the finish.

But anyway, current rankings, whatever they are based on are of limited value, as really the wind conditions that each of the boats will find on their path could easily compensate (for now) for 100's of miles difference in the current ranking (as the boats in the east will most probably experience very soon).

I sent my observations also to the VR Support. Haven't received a reply yet though. Maybe they want to keep that a secret anyway?

Enough for now.

Wouter
B210

Post by zezo | 2020-11-19 | 14:10:17

It's complicated. How do you compare a boat that has passed a gate against one next to it that has not passed yet? I guess the calculation is something distance to the gate + predefined distance from the gate to the finish (let's say along the white line). If the calculated distance is way off there will be huge difference in rankings around it (we've seen that before)

But even if it's in the ballpark the distance to the next gate comes into play.

Post by B210 | 2020-11-19 | 14:46:08

I guess everybody that passed this (or any other) gate will be ranked according the next criteria/gate, and by default they will be all ranked higher than the ones that did not pass the previous reference/gate.

Post by BooBill | 2020-11-19 | 14:54:46

Not the case. MidnightFoiler has now crossed the line, ranked 36000, while Stevoo has yet to cross, ranked 30,000.

Post by B210 | 2020-11-19 | 16:11:45

I judge that this fact may rather comfirm that the Natal/Harper line is not the reference for the ranking (anymore), not for boats that crossed it already, and neither for boats that didn't cross it yet (as I suggested already in my initial reply).

In fact when I select again a bunch of boats with similar ranking as Stevoo (i.e. same (VR) calculated distance to finish), then this results again in a straight line from SW to NE, and Midnightflyer is still above this line, which then also explains his (slightly) worse ranking. So I guess calculations are overall still based on a line accross the Atlantic. Not sure it's still the same line as in my initial calculation, but the bearing angle looks similar to what I see in my previous screenprints.

Post by BooBill | 2020-11-19 | 14:07:33

If you want remaining distance in the race, then Cape Good Hope Bouy is probably the best approximation. However, that's not a good gauge of race progress. Each race will differ a bit depending upon conditions, but the Natal-Harper line wasn't that bad at reflecting the angle the fleet was spread across.

I did some calculations last night before I did the original post. The Natal to Harper line is at a bearing of 71degs. The bearing from me to YourMomSA was 87degs and at the time we had almost the same routing ETA to 45S waypoint. Yet, YMSA was ranked at least 70,000 positions ahead, owing to his more easterly position. I found Stevoo was a bearing of exactly 71 degs from me, meaning we were equidistant to the line, yet, his ranking was about half the difference between YMSA and myself.

Post by Les Boineaux | 2020-11-19 | 22:28:12

What about St. Helena, the island itself, for the moment?

Post by BooBill | 2020-11-21 | 19:05:27

Any more intelligence gathered on this question?

I think we have 8,900 reasons to figure this out definitively before Pt Nemo.

Post by YourMomSA | 2020-11-21 | 20:26:47

Pat-Rock posted something about the Point Nemo challenge in the VR forum a couple weeks ago, that was both more specific than we'd been given before and more vague than a quality "terms and conditions for a contest" should state... He said his understanding is that at some fixed point in the future, which he theorized to be when the leaders pass Cape Leeuwin, a target area for the challenge will be shown on the screen. My guess is that it'll be some kind of finishing circle, and that perhaps they're intentionally waiting so that they can position it in a way that fits their goal relative to the wind forecast.

There's actually a buoy there now, if you zoom in far enough.

Hopefully whatever their plan is, it'll show up as something Hardtack can decrypt and show on Dashboard, so that there won't be anything mysterious about it.

Post by zezo | 2020-11-21 | 21:47:37

You can't stop amazing me. Never zoomed that far myself. But is it a cardinal or lateral mark or something else? The X suggest a "special" one, but those are supposed to be yellow. Color suggests lateral ;-)

To me announcing a second award looks like making you register and paying for second boat, so you could race for it.

Therefore the goal should be somehow outside of the fleet path.

Post by Hardtack | 2020-11-21 | 22:53:43

I followed the discussion in the VR forum, and I found a Point Nemo entry in the list of POIs (response of the Meta_GetMapInfo request):
{
"lat" : -48.876 ,
"lon" : -123.393 ,
"zoomOn" : 4 , "zoomOff" : 26 ,
"type" : "POINT" ,
"name" : "nemopoint"
}
The entry was already there when Pat-Rock said Point Nemo would eventually appear on the map, but I don't remember the buoy being shown at that time.

There are more entries of type POINT which carry the names of capes and straits. They are rendered differently, but I didn't see anything in the messages the UI could use to differentiate between them.

Post by YourMomSA | 2020-11-22 | 01:26:55

I suppose if Cvetan were to put a destination option on that spot, it would be a preliminary routing option for people who want to win the watch. On the other hand, it also could result in people thinking there's reason (other than the watch) to go there.

Post by BooBill | 2020-11-21 | 21:37:11

Okay, that seems pretty clear. Thanks YourMom

Post by YourMomSA | 2020-11-24 | 04:28:34

Hmm... Now it has whatyousea in first, Starbuck in second, and Spektakel in third, over by Brazil. I wonder how it's calculating that...

Post by zezo | 2020-11-24 | 05:56:15

Noticed it too last night. That breaks all "scientific" theories. No gates and white lines can explain it, unless there are "hidden hidden" buoys that are not visible in the client-side data at all.

Post by BooBill | 2020-11-24 | 04:41:05

Well, it appears the grib they are using since the last update came from a random number generator, so anything is possible.

Post by karriv | 2020-11-24 | 06:39:17

Perhaps it could be possible to solve this little challenge VR has kindly put in front of us by developing an algorithm that would do the following:

Take DTF for a number of boats in different positions
Subtract a constant, e.g. DTF from one of the capes to get closer to a possible solution (this constant could be varied to see different options)
Draw circles fore each boat with a radius of what remains after the subtraction
See what comes out: a line, a crossing point, something else, nonsense

Unfortunately I'm not capable of doing the coding or calculations. Well perhaps the calculations but not the coding.

Post by B210 | 2020-11-24 | 08:26:38

Even more strange is that the new leaders (top-9, all very near Brasilian coast) are all of a sudden +1700NM(!!) in front of all the others. That can only be explained as a radical change in the way the ranking (i.e. DTF) is calculated. Or they made a position error or calculation error for those boats.

When I check the position of boats around rank 2000, I still get something that looks like a straight line, or with very slight curve which may indeed be an arc from Cape of Good Hope. Something similar for boats around position 10000. So that seems reasonable. But those 9 leaders are way "behind" those (both) lines/arcs!?

And just behind these 9 leaders (ranking-wise), are still the boats that were leading earlier, but obviously those can't all have lost 1700NM in just few hours/days. And visually judging, they seem still to be closest to the Cape anyway from all boats, and much closer then the new top-9.

Apart from VR position/calculation errors, the only way this could be explained is that those boats near Brasilian coast are in an area that VR has a very different (and probably not very representative) algoritm to calculate the DTF.

Interesting enough, a few of the earlier leaders near the African coast are also still in top1000. Which is quite surprisingly by now, knowing that they met already quite some unfavorable wind conditions. But their position at least still fits on the line/arc of boats with similar ranking.

BTW, on zezo map, I could not find any dots for that top-9!? Not sure if that means something.

Post by B210 | 2020-11-24 | 09:52:50

Things change fast, surely much faster then the real virtual reality could do ;-). Just few hours later and those 9 leaders near Brasil have been given a much worse ranking by now. Whatyousea is now ranked +2000th, about 250NM behind the VR leaders (the same as yesterday). This is still much better then I would assume where these Brasilian boats belong in the ranking, but obviously VR is doing strange things with (at least) those boats. Not sure what VR is doing there behind the scenes. Are they updating their DTF calculation algoritms on the fly!?

Post by BooBill | 2020-11-25 | 01:44:23

I think I may have solved this. Earlier;

Post by YourMomSA | 2020-11-24 | 04:28:34
Hmm... Now it has whatyousea in first, Starbuck in second, and Spektakel in third, over by Brazil. I wonder how it's calculating that...

ReplyReply to this postPost by iconPost by zezo | 2020-11-24 | 05:56:15
Noticed it too last night. That breaks all "scientific" theories. No gates and white lines can explain it, unless there are "hidden hidden" buoys that are not visible in the client-side data at all.

Noticed that the rankings has Spektakel and myself virtually tied, despite him being 800 miles to the east. Well, I started drawing circles in Google Earth to find points that we were equidistant from. Basically, it had to be points south of both of us. South pole was pretty close. And, if you extended that line through the pole you got to Invisible Checkpoint 5.8, south of Tasmania.

Love to see if someone else can verify this, with other equally ranked boat pairs.

Oh, and btw, running the routings to 45S, Spektakel, despite what it looks like, isn't actually that far behind.

Post by MidnightFoiler | 2020-11-25 | 07:40:08

It certainly does seem to be distance to somewhere very south of us as I'm rapidly climbing the rankings despite not really sailing toward our destination. DTF is reducing by about 70% of my distance travelled in 10 mins and my heading is 192!

Does seem strange that the actual leaders are nowhere near as south - suggesting different algorithms are being used depending on where you are - maybe your nearest checkpoint? or shortest of distance for any checkpoint of (distance to you + its "distance to the finish")

It'll be interesting what they do in the daily report when the "leader" is clearly not nearer the finish .... since they love talking about the irrelevant people at the top of the rankings ... Might make them fix it! ... or they might continue reporting rubbish driven by facebook posts ...

Post by karriv | 2020-11-25 | 07:45:35

This actually reflects the difficulty of calculating a meaningful DTF. Sticking with a single point can give you ridiculous numbers. Moving the point according the progress of the race gives you a different set of interesting symptoms as does using gates.

Post by gavitellogolos0 | 2020-11-25 | 12:38:04

The same happened to me: two days of sailing to SW at around 190/195 and I moved from 40.000th position to 1974th right now...

Post by BooBill | 2020-11-25 | 09:44:41

Waiting to see your name in lights Midnightfoiler? VR may not be heralding it yet, but by my reconning, you've been the leader of this thing for several days now ;) On routing at least.

Post by MidnightFoiler | 2020-11-25 | 14:16:37

In our bunch maybe, there are a few close as I have discovered now. GAVIB69 could be in the best spot though

Post by zezo | 2020-11-25 | 14:30:23

You had a nice lead before the front double crossing.

Now there are few boats in similar position or slightly ahead. koky, Arilla, taberly-TPN on my screen.

I'd say koky is currently leading our group

Post by karriv | 2020-11-25 | 16:28:13

Armada Jet.

Post by BeggarBee | 2020-11-25 | 17:11:57

What about latortue2 ?

Post by B210 | 2020-11-25 | 15:09:10

New developments in the rankings, once again... It looks like the top boats moving SW now (which is theoretically a longer route (!), are moving up in the rankings, with the line of similar rankings now running rather NW to SE for these boats, something I haven't seen before here. I don't think any logical gate or reference point can explain that (unless they plan to circumnavigate the south pole in the wrong direction ;-) ). Boats with similar ranking further behind show still a rather SW to NE equidistant DTF line. Strange?!

What does seem more obvious, is that top boats on that NW to SE ranking line, are all on their way to round the St Helens high on the left (i.e. counterclock, so enjoying winds that will carry them fast in the right direction soon). So distance-wise they are certainly not sailing ideally, but routing-wise they clearly look forward to a big advantage in neae future. So that made me start to think that VR is maybe also basing the ranking on the estimated optimal route according the current wind forecast, which could be representative also (as that is theoretically the best route to follow if forecast would be static, and also why people use routers to set their course).

OTOH, that would rather give a ranking based on ETA and not based on DTF. So not sure how VR would calculate that ETA back into some "virtual" DTF to show in the rankings.

Any thoughts on all that?

@gavitellogolos0, any idea if your ranking improvement could be linked to getting into a much more favorable position routing-wise? More favorable and stronger winds in the latest forecast for your future optimal route? Instead of just moving faster to a particular reference point/gate/line? I mean, it seems near impossible that one can make up the difference in DTF between place 40000 and 2000 (which represents in VR a difference of about 300NM) in just two days of sailing.

Post by gavitellogolos0 | 2020-11-25 | 17:02:16

I am 10/15 miles NW to machaon and Tipapacheri, so close to koky and MidnightFoiler. We could ask them if they experienced the same change in ranking over the last 2 dyas. Our routes were very similar in the last 2 to 3 days.
By the way, now my position has stabilized around 1960 while I reached a low 1820 this morning

Post by YourMomSA | 2020-11-25 | 16:41:06

Hopefully once we pass Gate 1, the rankings will make sense.

Post by BooBill | 2020-11-25 | 17:09:45

You would think. However, at the moment, there is no point on earth that is consistent with the rankings. Those in the main pack are being ranked by how far south they are. The small pack currently in the lead ae clearing being measured to a different point.

Even once we are all heading east in the southern ocean if the point they picked is too far away, the orthodromic distance is going to favour whoever is most south, even though they might be pressed up against the ice limit and can not realize that path.

Like in any race, closest to the finish, does not really mean leading.

Post by karriv | 2020-11-25 | 18:23:14

OK, I've been tracking in Excel my rankings, DTF and DTL. Interestingly enough, my VMG has been 31.5 kn on average since the morning (about 400 miles in about 13 hours).

I think any effort in putting to understand what is the point or line to what VR calculates the DTF is, to put it mildly, wasted.

Also, harmonie85, who's max 5 nm pretty much directly ahead of me has 11 nm smaller DTF.

Post by B210 | 2020-11-25 | 20:11:25

karriv wrote:
"I think any effort in putting to understand what is the point or line to what VR calculates the DTF is, to put it mildly, wasted."

I agree with that by now!
But surely VR uses some sort of preprogrammed/automated algoritm to create this pretty dynamic ranking. So the quest to find that algoritm, still continues.

Just really very curious to figure out how they do it...

Btw, how are the real VG competitors ranked officially, and how is their DTF calculated? Is there any formal information on that?

Post by karriv | 2020-11-25 | 22:19:37

It could be that they are actually moving (that is it is moving gradually) the point/gate they calculate the DTF to in order to avoid the sudden jumps we have seen before.

Post by BooBill | 2020-11-26 | 02:01:51

Yeah, it's definitely a moving target. I was steadily improving in the rankings, up to 284. That made sense, I'm on the outside of main pack moving faster and slightly lower bearing than anyone else. Then, from one 10 minute update to the next, I jumped to 606, and 888 on the next update.

So, clearly they just moved the mark. It's still mostly south of us, but they've moved it a bit east.

Post by Scup | 2020-11-26 | 03:02:21

Hi BooBill,
I noticed the same thing happening over the last 20-30minutes.
I'm trying to remain focussed on sailing my boat and sticking to my strategy but find that the fluctuations in the rankings are very distracting!

Cheers,
Scup (WishfulThinking)

PS - I'm just behind you and chasing your stern.

Post by BooBill | 2020-11-26 | 03:15:54

Stay there ;)

Look, at the end of the day, this all doesn't matter till the finish line. Even within sight of the finish line, the guy closest to the line is not the guy most likely to win in most cases. This is sailing, that's the whole point of the game.

But, I admit, it sure is nice to see validation of your strategy in numbers.

Post by Furious | 2020-11-26 | 03:27:23

Hey Scup,
BooBill likes to be in the front possie mate.....just stay where you are try and get passed him when it counts....lol :-) maybe do it when he is in sleep mode.....

Only Joking BooBill

Post by BooBill | 2020-11-26 | 09:58:54

4:57am. Catching me sleeping is not that easy.

BTW, I have a new theory. They are using the current position of the leading real boat. Right now using distance to Apivia would sort of work out.

Post by karriv | 2020-11-26 | 14:42:53

Nope. Calculated distances to Apivia and they don't correlate with the DTF's.

Looks like for a subset of boats (apparently below 35S) distance to 56S60E correlates quite well with DTF.

Post by YourMomSA | 2020-11-26 | 16:59:27

Looks like the rankings are moving toward where they should be today.

Post by MidnightFoiler | 2020-11-27 | 00:47:52

Yesterday I saw the top 3 were in in a ⌜ arrangement and quite close together. 1st was in the top left corner and 2 and 3 were on the sides - one of the 2nd or 3rd should have been ahead no matter where it was calculating the distance to!

Post by karriv | 2020-11-28 | 09:33:31

Something interesting going on with the rankings again...

Post by BooBill | 2020-11-28 | 10:12:28

You are not kidding. I jumped from 800ish to over 50,000 over night. Now seems who ever is farthest south is ahead in the rankings. I've a teammate ranked around 500 who is 50 miles south west of me, lol.

Post by karriv | 2020-11-28 | 10:19:57

We all know the rankings this "early" in the race don't matter much, but it would be nice if the relative rankings of boats roughly in the same area made some sense.

Post by polo218 | 2020-11-28 | 11:58:16

Look at this "funny" change in my ranking this morning.
I jumped from rank 6499 to 4258 and my Distance to Finish increased from 18142 to 18568 !

28/11/2020 � 11:50:00 UTC+1 4258 - 18566.1
28/11/2020 � 11:40:00 UTC+1 4258 - 18568.4
28/11/2020 � 10:30:00 UTC+1 6499 - 18142.5
28/11/2020 � 10:20:00 UTC+1 6289 - 18142.6

Post by karriv | 2020-11-28 | 16:14:17

Yes, average VMG since this morning -35 kn.

Post by MidnightFoiler | 2020-11-30 | 07:57:30

philippej in our team suggested it could be based on distance to nearest ice gate point. Given above I think he could be right ... would explain increasing DTF as you sail east of it but are not yet closer to the next one and would explain boats in different regions clearly based on distance to a different point.

Ugh ... no, still would not explain a VMG greater than boat speed (positive or negative) ... unless the distance to finish of each of the points on the ice gate is calculated stupidly and does not actually decrease as we move around giving us sudden increases in DTF

Post by Turux | 2020-12-08 | 12:25:11

I'm getting more and more confused on the rankings as the race goes on... At this stage I'm closing up with the leaders (I can see in the fleet tab in VR Dashboard that my DTU is decreasing) and yet I am constantly loosing positions.

I think that the best thing to do is generally ignore it and focus on your race plan.

Post by Hardtack | 2020-12-08 | 13:23:33

DTU is "Distance To Us" - your distance to the other boat. Of all the distance numbers it has the weakest relation to the "actual" ranking.

I recently reenabled DTL but this is just the DTF difference between you and the boat with minimal DTF of all the boats you clicked in the game UI.

You already came to the right conclusion - navigation decisions should not be based on ranking at all, but others can elaborate on that :-)

Post by YourMomSA | 2020-12-08 | 14:23:51

The fleet is heavily compressing now, and will continue to do so for a couple of days, until we can clear the flat ice limit and turn south again. The margin between 1st and 8,000th will probably drop below 10 nm. When the fleet is that dense, the rankings will be more unstable than when it's spread out.

Post by BooBill | 2020-12-08 | 15:33:37

It's always a mystery where they are measuring the DTF to, but, if it's anywhere close to my target destination (50S/150W), then our current heading is 70 degs off the orthodrome. That means those on the right side of the pack are favoured for ranking. Doesn't mean they make it to the corner of the ice line first though.

As in any race, between marks or other choke points, determining fair ranking is true alchemy.

Post by Turux | 2020-12-09 | 07:52:51

Thanks for you replies! That was my first message on the forum! What a great platform and community!

Have a nice day and keep safe!

Post by KatanaSG | 2020-12-15 | 00:28:46

Utter confusion, now, sailing along the ice limit, this should be quite an easy call as we are basically rounding a peninsula on the ice limit. That said, the further south seems to give the best ranking, ignorant of the GC course to the next corner...

Post by BooBill | 2020-12-15 | 01:45:49

It would appear that currently the leaderboard is being measured against orthodromic distance to Cape Horn. That line is 75 degrees south of our current heading and passes through Antarctica and the Ice Limit. You know, places you can't go.

Therefore, it strongly favours whoever is most south, when any sensible person is looking at who is the furthest east.

As in any race, it also ignores routing. If you analyze routing, then the current leaders are going to be hours behind when they tuck in behind the main pack up on ice limit in a few days.

Post by MidnightFoiler | 2020-12-23 | 02:22:45

... and now the current leader Giordano is apparently 12nm closer than me to the finish according to VR and he is way north but route to cape horn in zezo and it has me 19nm closer (presumably GC distance), therefore VR must be calculating to somewhere north of Cape Horn.

Orca on our team just said to me "dont understant why they cant use the same system as the real boats which is basically a moving waypoint along the GC, a couple of days ahead of the leader (I think). its simple to impliment, easy to understand and as accurate as any method."

Eureka ... that would make sense - 2 days ahead on the GC would be around the corner north of Cape Horn at the moment and would have been cape horn direction a day ago and would have favoured boats south before that as we we were seeing ... another possibility at least.

Would not explain different boats in different locations appearing to have different measurement points like before Good Hope though ... perhaps they changed it since then.

Post by BooBill | 2020-12-23 | 03:58:46

I think they have been using that system since about Good Hope, as you suggest. Most of the time that favoured boats on the south side of the fleet, since the Great Circle would be a more southerly bearing that we are sailing. They probably had it too far ahead, as that shortest distance took you through prohibited areas.

I noticed a few hours ago that my position jumped significantly in the right direction and concluded them moving the measuring point significantly around Cape Horn was the only explanation. In fact, I suspect it's now the finish line. That would explain why Giordano, who is at a significantly more westerly latitude, but way north, is shown in the lead.