Does Paradox Sail Upwind?

This is the question that is often asked on various boating forums and blogs about Paradox and the efficacy of the chine runner concept in general. I took a leap of faith in deciding to build Johanna, going on the word of not only the designer, but on the experience of previous builders.
Well after last weeks sea trials of "Johanna", I came away with a GPS tracklog covering 40 nautical miles, quite a few of which were upwind sailing. Armed with that I can present the unadulterated data for your perusal and for you to answer the question above for yourselves.
Firstly, what does the designer himself say? Well Matt Layden provided the following info many years ago.

I took Paradox out to a part of the Indian River ... where there is practically no current, uniformly deep water and little boat traffic, and did a series of timed runs to try to quantify her windward performance . I tried to do what I reasonably could to keep inaccuracy from accumulating, but I don't really know how good the results are.  I laid out a triangular course with legs over a mile long, using fixed channel markers and a point of land that were all on my chart, but I don't know for sure how accurately they are plotted on the chart.  I took compass bearings and did some timed runs to other charted features nearby; everything checked out reasonably well, so I'll use the data as they stand, but I would really like to double check the mark positions by GPS or survey from shore some time.

I did 3 circuits of the triangle, timing the downwind leg as well as the 2 upwind ones, as a control for changing wind strength, etc.  At the end of each leg I measured (by timing the run) any offset distance to windward or leeward of the finshing mark, to use in calculating distance made good to windward.  I dipped the standing lug sail to leeward of the mast at each tack (which I don't ever normally do) to reduce any error from the otherwise asymmetric rig.  I had no way to accurately measure wind direction (compass bearings are not accurate enough to be meaningful at this level), so I used half of the total angle between the windward legs as sailed in calculating VMGW. 

The data all seemed okay, but there were fluctuations of several percent from one run to the next on each leg, and the last 2 legs were complicated by the arrival of a short raain squall that made the wind a little uneven for about 30 minutes, so I would say that there is a margin of error of at least 1/10 knot in the results, I think doing much better would probably require a towing tank.I will include a copy of the raw data in case you are interested, but the short results, averaging the 3 runs for each leg, go like this:

Downwind leg VMG: 4.53 kt.Starboard tack VMG: 3.63 kt.
Starboard tack angle made good to wind: 51 deg.
Starboard tack VMGW: 2.29 kt.Port tack VMG: 3.49 kt.
Port tack angle made good to wind: 51 deg.
Port tack VMGW: 2.21 kt.(VMG = Velocity Made Good along course; VMGW = Velocity Made Good to Windward)

These results pretty much bear out what I have observed before, so I'm willing to believe them.  What this mostly does is tell me in numerical terms that yes, she really does go to windward passably well without a centerboard.  It supports my experience that in real world conditions, the rare few times Paradox has been near other small shallow-draft cruising boats, she has usually sailed right past them, upwind or downwind.  This of course isn't because Paradox is a particularly fast boat but because, by and large, other other small shoal cruisers are not particularly fast either.

Back to the present. On my first days outing in Poole Harbour, I had the opportunity to beat up the gradually narrowing Wareham Channel. The wind was fairly steady, straight up the channel at about Force 4, occasionally stronger, and I was well pressed. I'd put in a reef of about 18" but was still overpowered. I felt that a bit more reefing would improve windward performance, but with a temperamental furling system, I continued as I was (more info about that later!). I'll call this track Beat 1.

Here's the GPS track of the first beat (Beat 1)

  The purple track is the raw GPS tracklog data, the red is the smoothed data eliminating track points between tacks. The dark brown is the velocity made good upwind (VMGW) between start and finish points. Some boat traffic and a tangled mainsheet (more about that later!) caused a few issues between D and F and I'll take the liberty of removing them as not representative of the overall upwind passage.

This is what the data looks like numerically:

Looking at the data we can see that on the first four tacks we were tacking through the following angles:

114°, 116°, 115°, 119°, and finally after the fluffed tack 111°.
This equates to an average of 115°.

The average speed was 3kts with leg three at 4kts.

 Below is the data for velocity made good to windward (VMGW)

 

As you can see the VMGW was 2Kts over the 0.7nm leg, taking 27mins, 37 secs. As this was my first time sailing a totally unfamiliar boat I am more that happy with the results under those conditions.

The next day brought about much more benign conditions with light winds predominating throughout. The intervening hours had given me a much clearer grasp and understanding of the boat's handling, and in particular how to work her upwind. Almost exactly 24 hours after the upwind leg above, I had the opportunity to repeat it in a variable wind of between 0 and force 2. I'll call this Beat 2. This track is a bit further up stream than the previous.

 Here's the GPS track of the second day's beat (Beat2)

This time I was concentrating on two things, initially racing a group of dinghys to the windward mark of their race course, and making progress in a narrow channel which had only inches of water outside. I'll be honest I worked the boat hard on this occasion and pinched every bit out of her that I could muster. I was getting to understand the visual clues from the luff of the sail, particularly on port tack, which I'd incorrectly read the day before.

This is what the data looks like numerically:

 Looking at this data we can see that we were tacking through the following angles:

88°, 72°, 92°, 84°, 70°, 79°, 72°, 89°, 92°, 57°, 50°, 79° and finally 79°.
This equates to an average of 78°.
The average speed was 2kts with several legs of 3kts.

Below is the data for velocity made good to windward (VMGW)

As you can see the VMGW was again 2Kts over the 1.0nm leg, taking 34 mins, 52 secs.


I was actually quite surprised at these results, although at the time I couldn't believe how well we were progressing. As mentioned above I got tangled up in a dinghy race comprising Toppers, Lasers and Enterprises. I literally left these boats standing, Johanna cut through the fleet like a knife through butter. I'll admit that it looked like most of the sailors were pretty novice, but I'd have expected to at least be given a run for my money.

On both days we had a very slack current, Poole harbour tides have a long stand and we were currently on neaps. Any current would have been assisting which probably accounts for the better than expected performance on track 2, but so far, counting experiences elsewhere within the Harbour where I was beating, the chine runner concept appears to work beyond my expectations.
You decide.