MIMS: Three rivers, Three races

Readers may be curious to see split times for the full rivers (East, Harlem, & Hudson) – not just the “segments” shown in the previous two posts. So here they are.

You’ll notice a new swimmer in the mix here: Sarah Thomas from Colorado. Originally seeded 8th, Sarah managed to slip into 5th place overall with a roaring swim down the Hudson. Well done, Sarah!

The story of the splits: East & Harlem Rivers

In the “GPS snapshots” I’ve shown in the last couple posts, you can see how far apart each swimmer is (6 of them, anyway) in terms of distance. Four hours into the race, for example, Erica Rose was 455m ahead of Ollie Wilkinson, who was in turn 135m ahead of John Van Wisse.

Another way to model the race is to look at when each swimmer passes a given landmark. This shows how far apart each swimmer is on a different dimension – time. Using the GPS tracks provided by NYC Swim, we can actually calculate “split times” for each swimmer between any landmark we choose. And, using those split times, we can calculate each swimmer’s speed (including current) for each segment.

For the purposes of this study, I chose 11 landmarks – three in the East River (Pier 11, Queensboro Bridge, and the Randall Island footbridge), two in the Harlem (Macombs Dam Bridge and Spuyten Duyvil), and six in the Hudson (GW Bridge, Riverbank Park, 79th St, 34th St, Pier 40, and the finish at South Cove). Hopefully they’re all fairly obvious reference points. Here they are on a map:

So, here’s what happened in the East River:

As you may recall, Erica Rose was about 85m ahead of Ollie Wilkinson and me at Pier 11. However, the GPS tracks were taken at 1-minute intervals – so I was only able to calculate splits to about half a minute of precision. That’s why Erica, Ollie, and I have the same split for the Start-to-Pier 11 segment.

In any case, everything here pretty much corresponds to what we’d expect. From the start to Pier 11 we were swimming through mostly slack water – thus slower overall speed (probably 1 mph net current). The first part of the East River was fast (up to 6 mph average speed). In the latter part of the East River (along Roosevelt Island), the current slowed by about a knot.

Erica pretty much dominated the East River.

What about the Harlem?

Some thoughts:

  • According to these splits, the field was pretty evenly matched in the Harlem. Erica gained a minute on John (and 2 minutes on Ollie) between the Footbridge and Yankee Stadium, but then gave back 30 seconds to John (and 90 seconds to Ollie) in the upper Harlem.
  • This isn’t necessarily because Erica was swimming less well. It’s possible that the trailing swimmers gained some relative advantage by swimming in the Harlem during a faster part of the tide cycle. Erica entered the Harlem River at 11:32am –  5 minutes ahead of 2nd place Ollie and 7.5 minutes ahead of 6th place Miguel S. Those may have been crucial minutes. Based on the slow speeds of all the swimmers in the lower Harlem (just under 30 minutes per mile), we were probably swimming into a head current at first.
  • The current really picked up in the upper Harlem. Again, the trailing swimmers were relatively advantaged by this, as they spent more time in the faster current. This may explain Erica’s surprisingly slow split in the upper Harlem.
  • I had a terrible 1st half of the Harlem; but, I already knew that.

Here’s the order of the top 6 swimmers as they passed under the Spuyten Duyvil railroad bridge and into the mighty Hudson:

  1. Erica R. – 2:08pm
  2. Oliver W. – 2:13:30
  3. John V.W. – 2:14
  4. Miguel S. – 2:15
  5. Miguel A. – 2:16
  6. Evan M. – 2:17:30

Are you ready?

Swimming around Manhattan, tides are (almost) everything

Just over a week ’til the Manhattan Island Marathon Swim! Hard to believe it’s already upon us.

Sometimes people ask me if I have a “goal time” for the swim. That’s an interesting question. As anyone who’s spent much time in open water knows, the relationship between time and distance is somewhat complicated; even moreso for marathon swims.

MIMS is a different beast, though. I’d go so far as to say that MIMS times are pretty much meaningless — as an indication of speed. The typical winning time of 7 hours, 30 minutes works out to just under 59 seconds per 100m. So: world record 1500m pace, 28.5 times in a row. In MIMS, the tides are king – perhaps moreso than any other major marathon swim.

How important are the tides? Think of it this way. My ultra-marathon pace is about 2.3 knots. A world-class marathon swimmer? About 2.6-2.7 knots. The slowest swimmer in the MIMS field? Maybe 1.6 knots. Why am I describing swim speeds in terms of knots? Because that’s how river currents are measured. Now, guess what sort of current we will encounter when we round the southern tip of Manhattan and enter the East River? 3.1 knots. The river will be moving faster than we will be swimming.

It won’t always be 3.1 knots – that’s just the max flood tide at the Brooklyn Bridge, at 9:52am. But you get my point: MIMS involves navigating three separate rivers, on three separate tide cycles. At the start (from South Cove until we hit the East River), we’ll actually be swimming into a slight head current. The Harlem will max at about 1.4-1.9 knots. In the Hudson, an ebb tide of about 2 knots (max) will carry us home to the finish.

How important are the tides? You could take the best marathon swimmers in the world and, if you dropped them in one of the rivers at a particularly disadvantageous point in the tide cycle, quite possibly they would be unable to finish the swim.

And that’s not all! The currents don’t affect everyone equally – it depends on where you are in the river (i.e., how fast or slowly you’ve swum up to that point). For example, if you’re a fast swimmer, you may arrive in the Harlem River before the tide change – so you’ll swim into a head current at first. Slower swimmers arriving after the tide change will get a push from the outset. This has the effect of compressing the field (the trailing swimmers end up closer to the leading swimmers than they would in a slack current).

Same thing in the Hudson: the ebb tide doesn’t begin until 2:30pm – possibly after the leading swimmers arrive. And the tide doesn’t max out until just before 6pm – well after the leading swimmers have finished. In other words, the trailing swimmers get a better push coming down the Hudson, making their finishing times seem “faster” relative to the winner.

And that’s not all! Sometimes the tide change arrives early – or not at all. Sometimes (especially in the Hudson), a strong south wind will blow against the ebb tide and whip up a nasty chop – in effect, neutralizing the current assist.

The point is: Who the heck knows what MIMS times mean? The only thing that matters is the order of finish. All this is a round-about way of saying: I don’t have a goal time. The rivers will do what they will do; my job is to keep swimming until I finish.

There’s a saying among English Channel swimmers (attributed to Alison Streeter, I think), that “even a paper cup will eventually wind up on French soil.” I don’t know if that’s literally true – but to the extent it is, it’s probably even more true of MIMS. You don’t have to be a fast swimmer to finish MIMS. The tides will carry you. The important thing is to keep swimming until you get there.

So, if you ever hear anyone talk about their MIMS “time” as if it indicates anything about swimming speed – now you know to ask: “So how does that compare to a paper cup?”

The fastest swims around Manhattan (Part 2)

Part 1.

Shelley Taylor-Smith recorded the fastest swim around Manhattan (5:45:25) in 1995, but it was a special “record attempt” swim scheduled on an unusually fast tide. What are the fastest swims in the regular Manhattan Island Marathon Swim race, which is typically held on a slower tide? Over the 29 years of the modern MIMS, the 10 fastest swims are as follows:

  1. Tobie Smith, 1999, 6:32:41
  2. Tammy van Wisse, 1999, 6:51:31
  3. Rob Copeland, 1999, 6:52:49
  4. Susie Maroney, 1990, 7:00:27
  5. Matthew Nance, 1990, 7:04:53
  6. Jim Barber, 1991, 7:06:34
  7. Kris Rutford, 1991, 7:06:44
  8. Matthew Wood, 1990, 7:07:32
  9. Susie Maroney, 1994, 7:08:10
  10. Igor de Souza, 1991, 7:08:20

Interestingly, 9 of the 10 fastest times happened in just 3 years – 1990, 1991, and 1999. The 3 fastest times were all in one year – 1999. Perhaps these years were “stacked” with outstanding swimmers. Another possibility is that these years saw especially favorable conditions (faster currents, smoother water, warmer water, etc.).

One simple method of estimating the effect of conditions is to find the median time in each annual race – and compare each individual to the median of that year. The fastest swims relative to the median would therefore be judged as the “most outstanding” swims. Here are the top 10 swims, compared to the same-year median time:

  1. Shelley Taylor-Smith, 1989, 7:32:34 (84 minutes faster than the median)
  2. Shelley Taylor-Smith, 1998, 7:18:07 (-77 minutes)
  3. Jay Benner, 1998, 7:19:40 (-76 minutes)
  4. Chris Derks, 1998, 7:24:02 (-71 minutes)
  5. Jim Barber, 1989, 7:45:15 (-71 minutes)
  6. Shelley Taylor-Smith, 1988, 7:27:44 (-71 minutes)
  7. Susie Maroney, 1990, 7:00:27 (-70 minutes)
  8. Penny Palfrey, 2007, 8:36:01 (-69 minutes)
  9. David Strasburg, 1989, 7:51:10 (-65 minutes)
  10. Matthew Nance, 1990, 7:04:53 (-65 minutes)

And… there’s Shelley Taylor-Smith again! In fact, 3 of her 5 swims (all 1st place overall finishes) are among the top 10 “most outstanding” swims. In other words, she was demolishing the field.

No matter how you look at it – solo “record attempt” swims or regular MIMS races – Shelley Taylor-Smith is the greatest Manhattan circumnavigator of all-time. It’s really not even close.

Swimming vs. Tide Surfing around Manhattan (Part 1)

It’s well known that Shelley Taylor-Smith holds the record for the fastest swim around Manhattan: 5 hours, 45 minutes, 25 seconds.

What’s not quite as well known is that she achieved this feat on a special “fast tide” – a convergence of maritime conditions in the East, Harlem, and Hudson Rivers that occurs only once or twice a year, if at all.

With the founding of the modern Manhattan Island Marathon Swim race in 1982, and more sophisticated understanding of tide cycles, a string of specially planned solo “record attempt” swims were undertaken in the ’80s and ’90s, all on fast tides. After Diana Nyad‘s 1975 swim in 7 hours, 57 minutes, the record was lowered six times by four different people over the next 20 years:

  • 7:14 – Drury Gallagher in 1982
  • 6:48 – Paul Asmuth in 1983
  • 6:41 – Drury Gallagher in 1983
  • 6:12 – Shelley Taylor-Smith in 1985
  • 5:54 – Kris Rutford in 1992
  • 5:45:25 – Shelley Taylor-Smith in 1995

Her record has stood ever since, despite an assault last year by world-class marathon swimmers Petar Stoychev and Mark Warkentin.

To give a sense of how much the tides matter in these swims, Taylor-Smith’s 5:45:25 works out to a pace of 45.2 seconds/100m (12.1 minutes per mile – almost 5 mph). If you assume a 20 minutes/mile average swimming pace (world-class for marathon distance), that means she derived 40% of her overall speed from the river currents!

All this is by way of saying: There’s a reason nobody ever comes close to 5:45 in the annual race. MIMS is never held on the special “record attempt” tides. Why not? Though I haven’t seen this stated explicitly anywhere, I assume it’s because while the “record attempt” tides may push a fast solo swimmer around Manhattan very quickly, it may not be suitable for getting a group of swimmers (of varying speeds) around Manhattan in a reasonable amount of time. The “MIMS tide,” as I understand it, is actually selected to punish the fastest swimmers with head currents (they arrive at Hell Gate and Spuyten Duyvil ahead of the tide change ). As a result, the MIMS field tends to compress much more than would a current-neutral swim of similar distance.

So basically, the MIMS race and the special “record attempts” are two entirely separate categories of swims. We all know that 5:45 is the fastest overall swim, but that was on a “record attempt” tide. What are the fastest swims on the regular MIMS tide?

Stay tuned for Part 2….

NOTE: Much useful information in this post was gleaned from Capt. Tim Johnson’s wonderful History of Open-Water Marathon Swimming.

Don’t underestimate Tampa

Some people do the Tampa Bay Marathon Swim as a “warm-up” for one of the triple crown swims. And it makes sense: Tampa is early in the season, 8 weeks before MIMS and more than 3 months before high season for channel crossings.

But thinking of Tampa as a “warm-up” might tempt a person to take it less seriously – and that would be a big mistake. TBMS is one of only four annual organized ultra-marathon (25K or longer) swim races in the U.S. (along with MIMS, Ederle, and Swim Across the Sound), and it may be the toughest. While water temperature is not usually a factor, pretty much everything else is. Glancing through the archives, tide changes and rough seas seem to be the two big ones.

Swimmers typically start with the flood tide, which pushes them up Tampa Bay — for a while. If you don’t swim far enough over the next few hours, though, the tide reverses direction and starts to push you back towards St. Petersburg – making it effectively impossible to finish.

Tampa Bay is also quite large, so conditions can mimic those in the open ocean. Here’s what the Bay looked like in four recent years (click to enlarge photos):

2010 – 4/7 finished
2009 – 7/11 finished
2008 – 6/8 finished
2005 – 7/19 finished

Since Ron Collins’ pioneering swim in 1998 (9 hours, 52 minutes), there have been 149 solo entrants in the annual race. Of those 149 swims, 70 were DNF’s – they didn’t finish the full 24 miles. That’s a success rate of 53%. By comparison, over the same time period, 90% of MIMS entrants have successfully rounded Manhattan (239 of 266, not including those who withdrew before the event).

There are other factors at work, of course. The MIMS selection process likely “weeds out” swimmers least likely to finish, based on swim speed or previous cold water marathon experience. I don’t believe Collins has yet rejected anyone from attempting TBMS – which is a good thing, in my opinion.

But Tampa Bay has humbled some great swimmers. In one recent edition, a swimmer who is perhaps the best non-professional marathon swimmer in the U.S. retired due to seasickness. Last year, a well known swimming guru (who had twice finished MIMS) planned to swim TBMS, Catalina, and the English Channel, all in the same year. After Tampa (which to his credit, he finished), he decided: maybe I’m not a marathon swimmer, after all. In 2007, tragically, one swimmer passed away from a heart attack.

The point is, this swim is a beast. I’m preparing for it as such.