Note: I wrote a follow-up review of the Swimsense in May 2013.
The FINIS Swimsense Performance Monitor is a watch that, through various marvels of technology, monitors your pace, lap count, and stroke count as you swim.
I still maintain that for interval training, nothing beats a pace clock. Doc Counsilman’s ’50s-era invention will never go out of style. For long steady-state training, though, a watch that monitors laps, strokes, and pace might be nice. Personally, I can’t keep a good count after about 40-50 (more if the pace clock is large and digital).
In my case, it’s no idle question: I’m doing some long swims this year, and steady-state training is a regular part of the training diet.
But with niche products like this, one inevitably asks: Does it work? Counting laps and strokes is one thing – but does it count the correct number of them? This review is on the long-ish side, so for those short on time, here are the major bullet points:
- Counts laps very accurately – not even one missed lap over 50,000+ yards.
- Watch looks and feels well-made.
- User interface is simple and intuitive.
- Update-able firmware.
- Web integration – upload your data to the Swimsense web app and/or export to other popular training logs (or simply a CSV file).
- Highly configurable – can wear on either wrist, and use in 25-yard, 50-meter, 33-yard, or [any # between 10 and 50]-yard-or-meter pool.
The Hopefully To-Be-Improved
- Watch is far too susceptible to accidental power-ons, which quickly drains the battery.
- When using in a short-course pool, single-length splits are too subject to measurement/rounding error, and thus not useful. Should be able to set splits to 50’s or 100’s.
- Stroke counts are close-but-not-quite.
- Stroke-style recognition is good but occasionally quirky.
Stroke rate calculations include time spent turning at the wall(fixed in most recent firmware update).
- You have to press a button every time you start or stop swimming (annoying for interval training).
If you’re a data junkie, read on…
I put the Swimsense to the test with the following set:
- 200, taking exactly 12 strokes per 25-yard length
- 200 – 13 strokes/length
- 200 – 14 strokes/length
- 200 – 15 strokes/length
When I got home, I connected the watch (via included dock and cable) to my PC and uploaded the workout data to my online Swimsense account. FINIS has a slick “web app” that lets you visualize and analyze your workout, and then drill down to each swim within the workout.
Here’s the first page you encounter after logging into the Swimsense app – a “big picture” perspective on your workout:
The top section shows times for each swim and splits for each length (25 yards, in this case). You can drag this section horizontally with your mouse to view the other swims.
Below that are five graphs with information on your stroke count, pace, and efficiency. These are summary data, showing averages for each of the 4 swims. A “SWOLF” score is the sum of your stroke count and time on each lap (fewer strokes and faster times = more efficient).
Zooming in on the top section, we can see my times for each of the 200-yard swims, and notice that the Swimsense got the lap count perfect. (Accounting for the button press at the start & stop of each swim, my actual times were about 1 second faster per 200.)
One minor complaint: I don’t need or want to see my 25-yard splits. The watch isn’t precise enough (seconds rather than tenths or hundredths) to show meaningful 25-yard splits. As a result, when I look at the data for individual swims, it appears that my “pace” varies a lot more than it does in reality, because the watch extrapolates the 25-yard split to a 100-yard pace. Look at the chart of my first 200. It’s very unlikely that my “pace” varied between 1:08 and 1:24 over the course of this swim.
FINIS, if you’re reading — it would be great to have the option of setting split-length to 50 yards or 100 yards. At 25 yards, it’s too susceptible to measurement error and rounding error. Here’s another example of the problem with single-length splits (from a different workout):
There’s just no way my pace was jumping around that much. See how when one split is slower than average, the next one tends to be faster than average? If this graph showed 50-yard splits instead of 25, it would be a much smoother line (and more useful).
So, the Swimsense is pretty good at counting laps. What about counting strokes? For that, we have to drill down to the specific swims. Here’s the first 200:
Remember, on this first swim I took exactly 12 strokes per length. The first thing we notice is that Swimsense counts stroke cycles rather than strokes. OK, then: 12 strokes = 6 full stroke cycles. The watch, however, counted 7/8/8/8/8/7/8/7.
So, there’s an error in there somewhere. Perhaps the Swimsense is “counting” my arm motion during the flip turn as a “stroke”? It’s tough to say. Whatever the case, the count should be consistent – I did the exact same thing on each length. Push off wall, streamline for a bit on my left side, pull down with my left arm, recover with my left arm (then right, then left, etc.). Then, near the opposite wall, one final stroke with my right arm (my 12th) before pulling myself down into a flip-turn.
The other three 200’s tell the same story: Accurate lap count, but stroke count is off by a bit. (Click to enlarge.)
|#2 (13 SPL)||#3 (14 SPL)||#4 (15 SPL)|
I actually don’t mind if the stroke count is off my 1-2 SPL (and to be fair, the Oregon Scientific watch was off by much more). The more important data, to me, are lap count and pace. The problem is, if the watch errs on stroke count, all the interesting “downline” calculations get distorted – SWOLF, DPS, and stroke rate. I’d love to see my stroke rate trend over the course of a swim – but if I can’t trust the numerator of [strokes / time], I can’t trust the quotient, either.
The Swimsense is designed to recognize not only how many strokes you take, but also what stroke you are swimming – fly, back, breast, or free. This is a killer feature if it works. The Oregon Scientific watch, for example, requires two extra steps: (1) setting the watch to fly/back/breast/free mode; and (2) calibrating the watch with a lap in whatever stroke you intended to swim.
So, does it work? For the most part, yes. I should note – I didn’t really “test” the Swimsense on actual I.M. sets or multi-stroke swims. I use the watch only for long, steady-state freestyle swims. The watch did, however, almost always correctly recognize that I was swimming freestyle on these swims.
There were a couple interesting malfunctions. For example, see the following set of 300’s LCM, which I did all freestyle:
The colors for each 50m length represent different stroke styles “perceived” by the Swimsense. In this case, blue = freestyle and green = backstroke. So, the Swimsense thought I was doing backstroke on the 4th length – though in reality I was doing all freestyle. Watch what happens on the next 300:
Now, the Swimsense thinks I did backstroke on lengths 2, 3, 4, and 6 – and freestyle only on lengths 1 and 5. My pace didn’t change significantly between each 50 – why would the watch think I did backstroke? Next 300:
Again, the Swimsense thinks I did most of the 300 backstroke, but actually it was 100% freestyle. The next 300 is the most interesting:
Now the Swimsense thinks I did all backstroke. Almost as if it was “learning” (albeit incorrectly).
I have no idea what happened with this particular set of 300’s. To be fair, over the 50,000+ yards I’ve swum with the Swimsense, instances like this have been rare. On the other hand, compared to most people, I have fairly canonical stroke technique (i.e., I’m a forgiving test).
In the grand scheme, these are nitpicks. The Swimsense is a highly successful piece of technology, and worth the $200 I spent on it. Even better, because of the watch’s web integration, the firmware is update-able. So in a way, the Swimsense is “getting better all the time” – without you having to buy a new watch. In theory, every issue I’ve mentioned could be fixed in a future update.
The web integration is probably the Swimsense’s most valuable feature. When workout data are trapped on the watch (as with the Oregon Scientific), they’re just not very useful. Being able to export the data to a PC opens up so many possibilities for analysis – which the Swimsense web app takes full advantage of.