Sandra Vi going for record across America

what the data means wrote:

Sam's data shows a dual histogram, with data centered around 74 and 94 rpm, which I interpret as likely time spent fast walking (74 rpm) versus jogging (94 rpm).

If that being the case, it would be interesting to look at these histogram plots, and the amount of 'tightness' or 'scatter' around the walking cadence versus jogging cadence, as Mimi progressed along her journey. All the data taken was likely hers (assuming no relay handoffs). But maybe, as the injury progressed in the latter part of the journey, we might see: the amount of walking rpm increase relative to the jogging rpm, the walking cadence rpm distrubution maybe become more erratic, etc.

something to consider wrote:

A sample movescount cadence data set for Mimi from 9-13, found here,

http://www.movescount.com/moves/move176812117

, plots data points every 28 seconds. Quickly, looking at a random section of the data, one can see cadence values of 70, 71, 72, 73, 75, 77, 78, leading us to believe we do not see the periodic missing cadence values in the movescount data presentation that you are seeing in your data, taken evidently from Strava(?).

Both Movescount and Strava apply smoothing algorithms into the data when presenting them on graphs and are unreliable and why I analyze the raw underlying data instead. One example why their smoothed graphs can't be trusted is that it smears the data so much in the speed plots in Movescount that are slower speeds, you don't even see any variation in pace (which was discussed way back in the middle of this thread). I will try to get at least all the raw data and analyzed files uploaded tonight for everyone to have access to.

AquaDyne wrote:

No, but that's also probably an unlikely explanation. It's just a hint toward the fact that rounding is clearly involved somehow in the calculation.

Consider stride data rounded to the nearest tenth of a stride per second (e.g., 1.4, 1.5, 1.6). Now, maintain a running average of the last 3 readings with the caveat that you'll always have 2 of the same number and one of the number above or below it (e.g., 1.4/1.4/1.5 or 1.4/1.5/1.5 or 1.5/1.5/1.6). Average those 3 numbers, multiply by 60 (to minutes), divide by 2 (convert cadence to rpm) and round to nearest integer. The results are 40, 41, 43, 44, 46, 47, 49, 50, etc. These match the "holes" in the histogram.

There's still an explanation needed for why the 3 readings are never identical (1.5/1.5/1.5) or consecutive (1.4/1.5/1.6) but it seems reasonable to think that there's an earlier calculation step involving roundoff that leaves this as an artifact.

Bottom line, I'm not suspicious.

I think that is way too many assumptions and seems convoluted. Besides, it doesn't explain why I haven't found another Suunto user with the same Swiss cheese problem.

Guy who ran across Britain with a Suunto watch: https://www.strava.com/athletes/6923242#interval?interval=201726&interval_type=week&chart_type=miles&year_offset=0

lurkr wrote:

Mimi claims she started running to cope with her anorexia. Isn't running (a sport that encourages thinness/weight loss) the last activity you would choose if you suffer from anorexia?

.

She is mentally unstable and this obsessive running is linked to her personality disorder.

Excessive exercise is very common in mentally ill people with eating disorders.

It's all linked - wanting attention, history of anorexia, obsessive running, seeking validation from people that you are "special" and marvelous.

To paraphrase a '60s saying, "Running to counter eating disorders is like f***ing to prevent pregnancy."

scam_watcheroo wrote:

I think that is way too many assumptions and seems convoluted. Besides, it doesn't explain why I haven't found another Suunto user with the same Swiss cheese problem.

And your alternative assumption/hypothesis/theory is....?

I think your one assumption of guilt until proven innocent is way too many assumptions.

I recall reading a few pages back that Mimi had her watch on the highest possible resolution recording of cadence, one that is not the default nor normally selected by other users that you are selecting at random.

I'm seeing obvious evidence of internal rounding + averaging calculations that I can reproduce with only a few minutes of effort and zero knowledge of internals of the Suunto. I'm rather sure the quick and dirty calculations I came up with are not exactly correct. However, they demonstrate how rounding and averaging, together, can lead to the data you see without any nefarious cause. The onus of proof is on you to provide any alternative explanation, and evidence that supports your theory.

If you really want to prove this is suspicious, I suggest you spend your money on a Suunto watch and set it to the same settings Mimi used and record your own data.

Until then you are speculating and assuming guilt because you (and I) don't understand the internal math the watch uses.

lurkr wrote:

Guy who ran across Britain with a Suunto watch

I've downloaded all of his run across Britain data and analyzed it with my tool, he does not have the Swiss cheese cadence problem in his histogram and it looks normal. I'm in the process of uploaded all the GPX files I have for everyone along with their Excel analysis file right now. Should be done in a half hour and then everyone can click through the gallery of histograms on my Google Drive.

5-MILE SPLITS

Day # 42, October 22nd

All times are Garmin Time in the Pacific Timezone

Time: Miles ~Split Time +/- 3 minutes

03:00 AM: 00

04:16 AM: 05 ~76 min

05:25 AM: 10 ~69 min

06:34 AM: 15 ~69 min

07:45 AM: 20 ~71 min

08:55 AM: 25 ~70 min

10:04 AM: 30 ~69 min

11:20 AM: 34-1/2 Lunch Break Start

11:56 AM: Lunch Break End (36 minutes)

12:02 PM: 35 ~82 min

01:14 PM: 40 ~72 min

02:19 PM: 45 ~65 min

03:10 PM: ~49 MILESTONE - Crosses into Eastern timezone

03:27 PM: 50 ~68 min

My alternative theory is aliens. Mimi is not unique in using the 1 second data recording on her Suunto watches. The other Suunto watch users I've analyzed have varying levels of watch interval recording, including at 1 second intervals. In fact, the Kris King that ran across Britain has a lot of his data recorded at 1 second intervals and don't have the Swiss cheese cadence problem. You will be able to see for yourself when my upload to my Google Drive finishes. Inside each Excel file, I have a column called "Time To Previous Point (seconds)" that shows whether a particular run was recorded at 1 second intervals or at other intervals.

scam_watcheroo wrote:

Deer Scam. Please start another thread to discuss Mimi's data issues. The AquaDyne you are chatting with is a name thief. Mimi is no longer a part of the run across America. She removed all references to it on her website.

- Thanks.

Make a new thread wrote:

scam_watcheroo wrote:

Deer Scam. Please start another thread to discuss Mimi's data issues. The AquaDyne you are chatting with is a name thief. Mimi is no longer a part of the run across America. She removed all references to it on her website.

- Thanks.

+1

Yeah, I think this current thread should be focused upon Sandy, her updates, and when she will reach NYC.

I understand how some people are driven to prove Mimi didn't do what she said she did, but -- truly -- no one really cares anymore. She dropped out. It's over. Move on.

scam_watcheroo wrote:

For example, below is a raw data cadence histogram for one of Mimi's runs showing all the missing cadences. The second is from a random Suunto watch user I found on Strava. Notice how the random Suunto watch user isn't missing cadence values at regular intervals? I'm trying to find a logical explanation for all the holes in Mimi's cadence data (all her data is like this), but I cannot think of anything reasonable. I've also analyzed Sandra's cadence data and she is doesn't have "swiss cheese" cadence data either. Can anyone explain this to me?

https://ibb.co/kLx1o6https://ibb.co/evr0gR

Wow, that is a VERY odd looking cadence histogram.

A couple of things I'd look for:

1. Are we certain that GPX data downloaded from Strava is the same as the raw data that was uploaded? And isn't somehow processed by Strava's algorithms?

2. Maybe this kind of data only appears when the Suunto watch is set for low-fidelity / longer battery life recording? The other Suunto example you looked at may not have used the same settings.

3. Maybe this kind of data only appears for runs more than 6-ish hours in length? The Suunto watch (or Strava) might apply some kind of compression or averaging only once the size of the collected data exceeds some threshold. So again, the other example you looked at may not have been representative.

Total speculation: the weird histogram you found is similar to what I might expect if the original data used a different system of units, and each data point was an integer with no fractional amount, and then the data was later converter to a different system of units.

Make a new thread wrote:

scam_watcheroo wrote:

Deer Scam. Please start another thread to discuss Mimi's data issues.

Agreed. This thread should never have been hijacked in the first place. This complaint should have been made weeks ago.

Make a new thread wrote:

The AquaDyne you are chatting with is a name thief.

No he isn't.

Make a new thread wrote:

The AquaDyne you are chatting with is a name thief.

No he isn't.[/quote]

Prove it. What do you know about the vehicles Sandy is using? You can answer this in one second.

Alpha Romeo wrote:

To paraphrase a '60s saying, "Running to counter eating disorders is like f***ing to prevent pregnancy."

She simply substituted one obsession with another. From food to running.

It's very common. Loads like her that have an issue with food, attention and self worth.

She wanted this record for public approval, not for herself.

Yes, most (if not all) ultrarunners have addictive personalities (which can manifest in both very positive and very negative ways in life), but what makes you think you can diagnose Mimi and her motivations for doing a transcon? It seems AWFULLY presumptuous to say "she wanted this record for public approval, not for herself."

Personally, I have no idea what her motivation was for her transcon attempt. But she fell short of her goal; why can't you guys show some grace and stop the hatred?

Please post regarding Mimi's run across America data here:

— Thanxs

http://www.letsrun.com/forum/flat_read.php?thread=8496783