Jengu - Round The Island Race 2019 - Performance Analysis
In any sport its important to do post match analysis. Sailing is no exception.
With all the races that we do on Jengu we try to assess what we did right, what went wrong and what we could do better next time.
The sources of information on any sailing boat are vast. There are literally dozens of tools that we can use to record and review our performance.
To make the most of that information you need to have a fairly consistent analysis methodology.
For Round The Island Race we work backwards from the finish. From the results spreadsheet we can identify how the other boats in our class did. This gives us boats to compare against when replaying the race on the tracker.
This year Jengu was in class IRC 2B starting at 8.20am. This years race had about 1300 starters divided into a large number of starts and classes. The boats are split with IRC (boats with formal handicaps ratings) and ISC those that don’t normally race that have been given a handicap by the Island Sailing Club.
Overall in the IRC classes there were 351 boats with 19 J109’s. Jengu was 89th overall; a significant improvement on previous years performance.
The other interesting split for us is between the boats running the older J109 one design sail plan with large overlapping head sails; the boats in the image above highlighted in green and the rest of the J109 fleet running the IRC non overlapping head sail. The numbers would suggest that Jukebox and Jubilee had a significant advantage in what was mostly a race dominated by light winds.
So what else can we learn.
If we pick up Jigsaw, Jengu and Jambo, three very similar J109’s, on the Round The Island race tracker (http://ngtrack2.ngresults.co.uk/webviewer/) we can follow their course around the island. Well; we can where there is mobile phone coverage anyway. The tracker is an app loaded on a mobile phone carried by the crew on each boat. No signal means no track data recorded.
Jengu is the orange boat, Jambo is pale blue and Jigsaw is black
The start line is the blue line annotated onto the image.
We use Rockit to track our own boats progress around the race track. Rockit records and aggregates the information on Jengu’s instrument and navigation system. It creates a series of sample periods during a sailing session making it possible for us to both plot information on a time series style chart and spatially on a basic map.
For the Round The Island Race start we can see from the Rockit track that the wind is from the South East force 3-4; 12-17 knots True.
We make extensive use of Predict Wind (https://www.predictwind.com/) and Windguru (https://www.windguru.cz/) to try to understand how the conditions will develop during the day. We knew from the weather forecasts the wind was likely to drop turning to an Easterly then dropping off late afternoon before coming back from the West in the early evening.
Based on that forecast there was little prospect of a sea breeze developing due to the mainly onshore nature of the existing breeze. It would be cancelled out by any sea breeze that developed as the land warmed up. There is a very good article by Matthew Sheahan that talks about Sea Breezes in the Solent in the context of Cowes week but a lot of the logic also applies to Round The Island Race. (https://www.yachtingworld.com/cowes-week/cowes-week-sea-breeze-65633)
Jengu started the race under Elvstrom C1(light headsail) and an Elvstrom EPEX Mainsail hoisting the A5 Asymmetric just after the start. The hoist was good but we got hit by a gust and broached briefly before we could drop the Jib.
So the pink segments of the Rockit track line, on the left, is us hitting the line flat out. Pink implies this is a personal best for the boat in the given true wind angle/true wind speed combination.
Once over the line we had two goals; to keep the boat in the deep water and to stay away from other boats dirty air.
We are going West down the Solent with the tide. The tide generally runs fastest where the water is deepest. We can see from the Navionics Web App (https://webapp.navionics.com/) that the deepest part of the Solent is mostly on the Southern island shore especially between Yarmouth and Hurst Castle. When you get to Hurst Castle the fastest current is closer to the main land shore and spit that the Castle stands on.
Peter Bruces book on Solent Tides (http://www.peter-bruce.com/books.html) and Winning Tides (http://www.winningtides.co.uk/pages/open.htm) are both good sources if you want to get into the real detail of where all the tidal back eddies are.
The fleet tends to bunch up at the Cowes end of the start line. Staying in the pack is slow as you are sailing in the dirty air of several hundred boats.
So we want to do a dog leg down the Solent; starting somewhere central off Cowes at the start then aiming for an imaginary point off Yarmouth pier then out to the Castle, then South West to the Needles.
The only other considerations we have are staying in clear aim and sailing the boat fast.
So we’ve cleared the line, having started close to the line on the gun.
We were close to the start time without being in danger of being over which would have incurred a time penalty.
We have hoisted the correct Spinnaker for the conditions and are moving into space away from the other boats.
Because we are to leeward of the pack we are also sailing a slightly higher apparent wind angle which makes us slightly faster.
Fast forward to 9am.
The wind is still in the South East and but has now dropped to 6-8 knots True. We have the wrong Asymmetric up.
We are visibly losing ground on the rest of the fleet.
We could have gone with the ultra light North A1.5 asymmetric but there was always the possibility that the wind would come back. So we went with the North A2 which has the added advantage that it really likes to sail deep. A very quick change without hoisting the Jib got us back up to speed,
The short section of the track between 9:15am and 9:25am reflects the process of working out we have the wrong sail up, preparing to hoist the new sail, dropping the old one, hoisting the new one and getting back into the groove and the wind coming back up a bit.
Consistent speed, coming up in the lulls, bearing away in the gusts, saw us pull ahead of Jambo on the run down to Yarmouth. In the very light gusts the A2 doesn’t give us the optimum boat speed, but, we are keeping the boat in the deep water sailing a good course and OK apparent wind angle.
At this point we have crossed ahead of a big group of boats and cleared the group of slower boats that started at the same time as us.
We are now reaching the Island shore and rapidly running out of Solent. The deep water goes in close but once you are sailing through moored boats its time to Gybe back across the Solent.
We pick our way around a couple of larger ISC class boats being careful to pass upwind and astern of the them to ensure we don’t get stuck in their dirty air.
We then Gybe. We have to get the Gybe right, we have our club chief instructor on the big boat next to us and don’t want to look rubbish - a wine glass at that point would have lead to ridicule in the club bar for months to come.
Gybe executed we head across to Hurst Castle making the most of the fact that we are crossing the fleet on starboard tack. We Gybe again into clear air to follow the fleet out down past the Shingles Bank.
We are being pushed out noticeably here by the tide which is running at 2.8 knots by the time we clear the Castle on the spit.
The photo above taken from Hurst Castle shows us crossing the Solent. We are the boat with the blue spinnaker foreground right.
As we progress with the tide towards the Needles the wind starts to get very light and fickle.
We follow the gusts on the water sailing away from the other boats closer in to the Island shore crossing the channel until we are almost in Totland Bay.
We are very much aware there is the risk that we will be carried out down the Needles channel by the tide if the wind goes very light. We can already see a lot of boats further out than we want to be.
Finally we reach the needles.
In motor racing you try to hit the Apex of the corner. Sailing is not really that much different. We try to drive the shortest distance and carry as much speed as possible through the corner. The Needles is the first real corner on the Round The Island course.
However, unlike a normal mark rounding where the “corner” is well defined the Needles has a number of options.
go all the way around the rocks, wreck and outside the cardinal marker;
go between the cardinal marker and the wreck,
or the shortest route between the rocks and the wreck.
Cue much discussion; between the Skipper, Navigator and Helm, eventually we agree we want to go between the wreck and the Needles rocks.
By now the wind has dropped right down and we are ghosting through the water on a Starboard tack. We have rights on the boats that have been carried further down the Channel and have had to tack over to make the rounding of the Needles.
The Round The Island Race Sailing instructions are pretty clear on where the wreck of the Varvassi is even so crossing between the two sets of obstacles when surrounded by a lot of other boats is nerve wracking. It would however have been a very slow speed accident.
We ferry glide serenely between the rocks and the totally submerged wreck. More worried about the increasingly close proximity of the boats astern. The wind shadow of the headland means that there is more or less no wind to push us away from the rocks and the rest of the fleet.
As can be seen from Paul Wyeths photo the right the boats rounding the Needles are hardly moving you can also see the Swell coming in from the East.
So whilst there is little or no wind the waves are still running; the effect is to knock the wind out of the boats sails.
We do our best to stop the flapping; leaning on the boom to stop it swinging, putting the crew weight forward to leeward to heel the boat and dampen the movement.
Referring back to the race tracker we can see that Jengu is the first to round the Needles, Jambo rounds later and further out. Jigsaw’s tracker stops tracking somewhere near the Needles.. Presumably because the cliffs are blocking the mobile phone signal.
The waves that are disturbing the boat are coming in at an at such a angle as to make tacking onto starboard to go into the cliffs under Tenyson Sound a very slow route so we tack over onto port and attempt again to gain clearer air and flatter water.
Our boat speed picks up and we start to make good boat speed relative to our personal best. As we make ground to the south we clear the headland and pick up the easterly breeze again which this side of the Island is gusting to 15 knots True. We sail South on port for roughly 20 minutes then in clear air and smoother water we tack over onto starboard and start to close the South close of the Isle or Wight again aiming to reach the shore towards Brightstone Bay.
We are making good speed to windward. Our boat speed is in the top 10% relative to the boats previous performance.
The tide has now turned and is running East to West along the Island shore. In the deeper water (26m) we are stemming 1.8 knots of tide; further in much less (11m of water) .75 knots.
So the sooner we are into the shallows the better.
We also notice that our performance on starboard tack is better than port. Its unclear whether this is an issue with the rig setup or something else. For instance; the way that the waves are hitting the boat.
We need to tack up the shoreline to St Catherines Point; short tacking to stay out of the deep water and the tidal current. We follow this strategy for the next hour or so crossing tacks with the Sunsail Fleet which we are now starting to overhaul.
We keep our tacks out of sync with the other boats so that we are not following anyone and sailing in dirty aim.
We are well ahead of Jambo and sailing close to Jigsaw.
Its a slow process but each sequence of tacks to and from the shore gives us a little bit more ground on the rest of the fleet. Our boat speed is still good staying in the top 10% and the port versus starboard tack performance is relatively consistent now.
Once around St Catherines we continue the short tacks along the cost until it becomes obvious that the tide has finally turned in our favour. At 14.40 we are seeing .3 knots of adverse tide by 15:06 we are starting to see a knot or so of tide in our favour. At 15:20 we start to carry our tacks away from the shore further out into the deep water to get the most of the East going tide.
Our SOG immediately gains a knot and a half. We are now level with Sandown.
At this point we change helms and the boats performance drops off.
We go from greens, top 90% to blues bottom 70%. We also see a lot more variation in the boats angle to the wind.
Looking at the relative wind time series graph from Rockit, the angle that the wind is coming over the boat, the profile becomes much spikier.
The area in question is marked “Slow” - between about 16:20 and 17:25.
The question for the post match analysis is why do we go slow at that point.
We can see from the wind direction plot relative to North the wind starts to fluctuate much more noticeably during this phase. Making it more difficult to steer a fast efficient course.
The wind speed plot below shows the wind dying away after 15:45.
Whilst the obvious explanation for the slowing of the boat is the change of helm, the reality is the change to a more shifty lighter wind is just as much of an influence.
So on this occasion we can give the helm the benefit of the doubt.
So its now getting on for 1800. We’ve been racing for nearly 12 hours. We’ve done less than 50 miles. Its safe to say the crew are getting quite tired. We’ve been alternating the helm every couple of hours to keep them from getting stale. We’ve eaten the sandwiches we made before the start…. (top tip: East Cowes marina to Round The Island start line should be enough time to make a days worth of sandwiches for the whole crew) .
Once more the wind has shutdown.
The boats are between Sandown and Bembridge in what can only be described as a large wind hole with occasional gusts of 2-4 knots. The helms can see the wind gusts on the water but there is no way of getting to them. The wind is now circling around; first from the East then the South East.. We attempt to fly the light A1.5 spinnaker. That and the tide which is 1-2 knots in behind us carries us down towards Bembridge and the Eastern End of the Solent.
Talking to people afterwards; Bembridge at 1800 was the breaking point for a lot of crews and many boats retired not expecting to get back to Cowes before the race cutoff at 10.30pm. This was particularly obvious to the slower boats.
One Sigma 33 crew that I spoke to actually took their sails down at Sandown. The slapping of the sails was slowing the boat down; the tide carried them down to Bembridge and out of the wind hole; they eventually finished in the 50’s which shows its worth sticking it out to the bitter end … you don’t know what the wind is going to do.
About 1800 the wind started to fill in for us, building to nearly 7 knots and we were able to sail East up the Solent towards Ryde and the forts with the A1.5 up. The tide in the main Solent channel is flowing East against us but there is an eddy on the Island shore which we follow with what remains of the fleet. Our relative boat speed is still good.
Sadly that was not to last. As predicted the wind dies at Ryde; just before we reach the No Mans Land Fort.
We begin drifting, trying to find the wind, metres from Jigsaw. So after 12 hours the whole race resets again.
Our forecast; from 12 hours before had suggested the final shift would be the wind from the East dropping off before coming back from the West in the early evening.
You can see pretty clearly on the Rockit track when that happens. The wind does a complete 180 degree change between 19:20 and 19:35.
Jengu’s zigs and zags were matched by the other boats trying to work out what the wind was doing and how to stay out of the tide; and off the mud at Ryde.
With the wind coming back from the West we headed off down the Solent.
As we get into the early evening we need a morale boost. There’s not a lot of food left just a packet of sausages that were supposed to be breakfast, some cheese, some falafels and some rolls. It makes for an interesting snack, but enough to keep us going for the final push to the finish.
As with most years the question of where to go in the Eastern Solent arises. Is it better to stick to the shallows along the Island shore into Osbourne bay or should we just go for clear air up the middle in the slightly deeper water with the potential for more adverse tide ?
With hindsight I’m not sure there is a right answer, By the time we got beyond No Mans Land Fort it was approaching high water Cowes. In theory the tide starts to flow West in the Solent an hour before high water so we should have had the tide behind us from Ryde. The reality, as can be seen from the Rockit track, is that we had a little bit of tide against us. Less than a knot but not the couple of knots behind that we might have expected. We were in clear air and the tack lifted us most of the way to Cowes and the finish line.
Swapping helms around again kept the performance of the boat above the baseline numbers; though perhaps not surprisingly not as consistent as earlier in the day.
We crossed the finish line at 21:24.
We matched Jigsaws course fairly closely up the Solent crossing a couple of times in the tacks. Ultimately they beat us by three minutes after 13 hours of racing. Jambo whose tracker stopped working finished about half an hour later.
So what does the overall boat performance for the day look like ?
The session polars below from Rockit show that while we had reasonable performance in 10 knots plus wind range our light airs boat speed wasn’t great.
Each cell in the table below represents a True Wind Angle/True Wind speed combination. If the cell value is 0.00 we didn’t encounter that value during our recording session. The number in brackets is the best performance for the boat and drives the colour of the values in the cell.
So for 10 knots wind at 150 degrees wind angle we have a value 4.90 (6.48) 4.90 is the session average 6.48 is the best session average value for the boat. The number is in Red to reflect the session being in the 70-79% of boats best performance.
With Rockit we build a profile of the individual sails that we use.
Jengu has a set of 2 headsails and 3 spinnakers that we use on a regular basis. We also have a limited amount of data from earlier seasons about the performance of the older overlapping J109 One Design head sails.
The images below show the data for the Elvstrom C1 Head sail that we ran during the race.
The image on the left is the boat level profile that aggregates the sail data to the boat level. The image on the right is the data plotted specific to the Round The Island Race Session.
There are two ways that we can plot the data.
We can use a linear regression type model that assumes that the relationship between wind speed and boat speed is a straight line.
The second lower line at the boat level plots the average boat speed for each given wind speed. So if we have 30 sample periods for a given wind speed the value displayed is the average of those 30 samples.
The C1 Headsail has a recommended wind range of 0-14 knots, at 14 knots we cross over to the C3 Headsail.
What we see on the graph pretty clearly is that the boat speed peaks at approximately 7.5knots. Which is reasonably consistent with, if a little bit above, the J109’s hull speed. Where maximum hull speed (in knots) = 1.34 x the square root of the waterline length (in feet). A J109 is 35ft long.
We can also see that our actual average performance, the lower line at the boat level, shows a lower level of performance.
When we get the sail performance right we can get close to the hull speed but there is a level of inconsistency in the performance. Understanding what causes that inconsistency allows us to gradually move the bottom line upwards.
The more data we collect the more complete the profile of the sail and its performance we build.
Sampling the data from Round The Island and comparing to the boat level data using the linear regression line.
So at 7 knots wind we are seeing 4 knots at the boat level and 3 knots at the session level which suggests our performance is actually below the longer term average.
At 10 knots wind we are seeing 6 knots at the boat level and 4.5 knots at the session level again below the long term average.
At 14 knots wind we are seeing almost 8 knots at the boat level and 7.0 knots for the race level.
So potentially we have an issue with the boats performance declining rather than improving over time.
It might just be the crew having a bad day or it could be an issue with the head sail.
We can compare back to the JOG Portland race from last year which was one of the first outings with the sail and compare the performance.
For that race we have 4.5 knots at 7 knots of wind, at 10 knots we are seeing 7 knots of boat speed.
So we need to think about that and what might be causing the issue.
We have made a couple of obvious changes.
In 2018 we ran the sail with vertical battens.
For the Round the Island race we ran with horizon battens. We’re also running the sail without the furler so it sits slightly higher relative to the fore deck.
Clearly we need to do some more analysis of the problem…. and realistically check the data to make sure we haven’t done anything stupid with the Rockit data.
So to summarise;
What we did right;
The routing around the race course looks good.
The boat performance has improved relative to the other J109’s this year.
We’ve climbed a couple of places overall and are now fairly consistently not last.
We have a much more consistent crew and are building experience.
What went wrong
Broaching after the start was not cool.
We lost 3 minutes to Jigsaw between Ryde and the finish
We still don’t have the upwind speed to match Jybe Talking. We were within shouting distance at Sandown but somehow they got away from us between Bembridge and Cowes.
What we could do better next time;
We need to look at how we build our boat speed.
We need to consistently find another couple of minutes in each hour on the rest of the fleet.
We could probably have sailed with the lighter spinnaker earlier in the race.
Doubtless we will be back next year to try to improve our result further.
Bill/June 2019