Wind Scoring Matrix Updated for 2020

January 23, 2020

In 2019, the Bermuda Race Organizing Committee commissioned an analysis to review and propose revisions to the wind matrix used to score the Newport Bermuda Race. Jim Teeters led the project and explains the recommendations, which should level out the playing field for racing in lighter winds. 

The Newport Bermuda Race has been handicapped for many years using Offshore Racing Rule (ORR) ratings and scored with a system called Performance Curve Scoring (PCS), which automatically calculates the corrected time for each boat. (For more on ORR and PCS, read “Bermuda Race FAQs: The ORR.”)

PCS uses a curve that fits handicaps at the seven wind speeds employed by the ORR—6, 8, 10, 12, 16, 20, 24 knots. For each wind speed, a blend of Velocity Prediction Program (VPP) speed predictions are used for a set of assumed wind angles. The wind-angle mixes used are different for each wind speed, reflecting assumptions about what happens on the racecourse during slow, intermediate and fast races.

Performance Curve Scoring: The graph shows ratings and interpolating curves for each boat's different ratings at different wind speeds. Depending on each boat's elapsed time calculated as Avg. SPM (seconds per mile), PCS scores a boat by automatically sliding along that boat's interpolating curve of ratings to the value of that Avg. SPM.  The wind at that point is the "implied wind".  The boat that sails to the highest implied wind is the winner (Boat A at 11.9 knots vs. Boat B at 10.4 knots). A scratch boat curve is used to convert to Corrected SPM times. Read more about PCS.

The intent of the 2019 study by the Bermuda Race Organizing Committee (BROC) was to check the validity of those wind-angle assumptions.

Five boats were chosen to represent the fleet across a range of ratings: Flyer (Cal 40), Carina (McCurdy & Rhodes 48), Apollo (J/121), Temptation – Oakcliff (Ker 50), and Proteus (Maxi 72).

The key elements of the analysis were as follows:

  1. Create a small fleet of boats representing the Newport Bermuda Race fleet and run it through optimal routing using GRB (Gridded Binary) wind files recorded for the last 15 years. Five boats were “sailed” to Bermuda for five June start dates in each year to generate a total of 375 optimal routes. 
  2. For each route, the frequency of occurrence for each combination of wind speed and wind angle was recorded. For each boat, this generated a graph of the percent of race distance sailed at each wind angle (5 degree increments) from dead upwind to dead downwind, and for each of the seven ORR wind speeds.
  3. These curves were very similar for each wind speed and for each boat. The data for all five boats at all seven wind speeds was combined into a single curve of probabilities to be used in the ORR VPP. Of particular interest, this curve showed very little time spent sailing dead upwind or dead downwind.
  4. Also, a not surprising finding was that the fleet can spend considerable time sailing in very low wind speeds—less than 6 knots, which is the lowest wind used by PCS. The GRB files confirmed the existence of numerous periods in recent races where some or most boats were virtually becalmed.
  5. To represent this in Newport Bermuda Race scoring, the lighter-wind ratings were redefined to include the assumption of sailing in little or no wind.
  6. A number of races were re-scored using handicaps derived from the modified wind matrix.
The chart shows the optimal tracks for the 5 test boats in the 2018 race. Pink: Proteus, blue: Temptation, red: Apollo, green: Carina, black: Flyer. The isochrones in blue are for Temptation and illustrate the period of very slow sailing halfway through the race. (Proteus was fast enough that she missed that “dead” zone.) Along each track are displayed wind vectors that were collected to determine the percent occurrence, by race distance, of the all combinations of wind speed and sailing angle.

The conclusions from this analysis:

  1. The wind-angle distribution does not change significantly with wind speed, unlike current assumptions.
  2. The current assumption of mostly windward/leeward sailing in the low-wind ratings used by PCS in slow races is incorrect. It is rare that boats sail equal distances on port and starboard tacks or jibes, and while there are periods where the boats are hard on the wind, they are generally pointed to where they want to go and seldom need to tack. This is much faster than sailing dead upwind or downwind.
  3. In previous years this has created a bias favoring boats that do not perform well sailing upwind and downwind in lighter conditions.   
  4. Adopting a modified matrix of wind strengths and course mix to remove this bias should help to make lighter-wind races fairer.
Actual wind conditions (GRB files) were used to plot the percentage of wind angles experienced by racers sailing in June during the eight previous races. The curves in the graph show the percent distance sailed at each wind direction with separate curves for each wind speed.  The 24 knot curve shows the most oscillation, due to the fact that there were few times the boats sailed in that much wind. 

None of this can address the reality that environmental conditions, including building or dying winds, frontal passages, or new weather patterns over the period of the race can greatly affect who does well and who does not, depending on the location of individual boats or classes of boats relative to weather patterns. 

The only way to realistically model the impact of those factors on the outcome of racing is to do post-race scoring using the actual wind conditions experienced across the racecourse over the period of the race. This would lead to uncertainty and delays in the scoring of races, and may not be desirable from either a practical or philosophical perspective.

Therefore, for 2020, the BROC has chosen to modify its wind-strength assumptions and course mix for handicapping the Newport Bermuda Race to more accurately reflect recent historical wind patterns as demonstrated by this study.

Jim Teeters is the Technical Director of the Offshore Racing Rule, which is owned and managed by the Offshore Racing Association.

Latest Bermuda Race News

June 26, 2020

VIDEO: 2020 Online Newport Bermuda Race Prize-Giving, Hosted by Goslings Rum

Join Andrew Holmes of Goslings Rum, your host for the 2020 Newport Bermuda Race Prize-Giving Ceremony

June 24, 2020

Guest Blog: Hand Steering at Night

Putting his nav prep for the real race to good use, Sloan Burns sailed a fast and conservative course and finished second in the Double-Handed Division.

June 23, 2020

Race Blog 8: “midnightexpress” Captures Double-Handed Division Honors

The fleet of water-ballasted Jeanneau Sun Fast 3300s hung close to the larger boats, finishing in less than four days.