The Road goes ever on and on; Down from the door where it began;
Now far ahead the Road has gone; And I must follow, if I can;
Pursuing it with eager feet; Until it joins some larger way;
Where many paths and errands met; And whither then? I cannot say.

[JRR Tolkien, Lord of the Rings]

Saturday 5 January 2013

TGO Challenge Statistical Analysis (Part 5–First Timer Retirees and One-Timers)

I am a bear of very little brain, and this latest set of analysis has taxed me greatly. After another afternoon of staring at numbers and concocting ‘countifs’ and ‘sumifs’ formulas, I was struggling to know my own name, never mind to work out what exactly it was that I was trying to achieve.

“You have to help me here!” I said to Mick finally, in desperation.

“What is the question you’re trying to answer” he replied?

I thought hard, staring at a column of numbers for a while. I tore my hair out a little bit more (a lot of hair-tearing has been going on). Then I answered “I don’t know”.

So, I put it away and came back to it this morning. The calculations suddenly became much simpler (helped by having written down a list of the questions that I was trying to answer) … until I got sucked down another rabbit hole and suddenly the questions became difficult again. Both Mick and I struggled over the data and the combination of my few brain cells with Mick’s many brain cells seems to have come up with some answers (whether they’re the right answers is a different matter).

First off, let’s look at some retirement statistics. I’ve already shown, in Part 2, the percentage of all participants who retire each year. The question that raised was how many of those retirees were first timers? Is it more likely that you will retire if you’re new to the event? Here, in absolute numbers, are the retirements for First Timers, plotted along with the retirements for non-First Timers (non-First Timers being those people who have started at least one Challenge before; there are a small handful of people who have started multiple Challenges but haven’t completed one*, but they’re few enough not to skew things unduly):

image

Absolute numbers aren’t that useful though, as they don’t put the numbers into context with how many participants were First Timers in that year. It’s probably more useful to look at the numbers in terms of percentages. This one shows Retiree First Timers as a percentage of all First Timers and Retiree Non-First Timers as a percentage of all Non-First Timers:

image

From that, it seems to me that it’s not a given that you’re more likely to retire if you’re a First Timer.

From looking at First Timers I went on to look at what I called ‘One Timers’ – those people who did one Challenge and never came back for another. It’s always possible that some of those people will come back one day, but for the purposes of this analysis the ‘Yes’ category are those who have already got multiple starts under their belts; the ‘No’ category are those who did their first Challenge between 1980 and 2007 but haven’t returned in the last five years; and the ‘Maybe’ category are those who did their first Challenge between 2008 and 2012 and haven’t yet been back, but they’re recent enough to consider that they may well do in the next few years.

image

I then looked at whether there’s any trend as to whether One Timers (taken as a percentage of all Challengers in any given year) are increasing, decreasing or staying the same, but on plotting that I realised that there must be a correlation between First Timer numbers and One Timer numbers (because you can only become a One Timer if you’re a First Timer, so, for example, in the year when there was only 1 First Timer there were no One Timers created). The chart below shows that in general it holds true that the one line follows the other, but the gap between the two isn’t constant. Where the gap is the smallest, a greater proportion of First Timers became One Timers than when the gap is bigger:

image

Remember that 1989 was the odd year when there was only 1 First Timer

That led to the question as to which year(s) put the most First Timers off? That is, in which year(s) did the highest proportion of First Timers opt not to take part in the event again?

image

Again, for the purposes of this analysis I’ve assumed that if a First Timer hasn’t returned in the last five years then they won’t come back, so have included all years up to 2007.

That begs the question as to what it was about 2003 that caused 62% of First Timers to not come back for a second Challenge?

------------------------------------------

Right, I think that’s enough statistical analysis and graphs for now. Surely, if anyone has made it through all of these posts, you must now be as sick of graphs as I am? For me, I think that enough time has been lost this week in complete immersion in the Monster Spread Sheet, so I’m now going to put it away for a while. I’m sure that there are many other statistics that can be drawn from it (together with a few lies and damned lies…), but they’ll have to sit patiently within the data for another while.

(*The biggest non-completer has been on the list for six Challenges but has completed none, with five retirements and one ‘Did Not Start’.)

 

Click to go to other Parts of this series of posts:

Part 1     Part 2     Part 3     Part 4     Part 6

13 comments:

  1. I'm just wondering how much of this is weather related, ie. how many first timers got soaked on the first day. we are due an exceptionally dry year (there's optimism) so it will be interesting to see if a high completion percentage goes with a dry warm crossing, enjoy.
    Cheers J.P.

    ReplyDelete
    Replies
    1. It's a pity that we don't have weather data to see what patterns emerge. What I know, or we have been told either in comments to these posts or otherwise: 1994 was a dry year with no extremes of weather over the Challenge fortnight, but deep snow was lying over the west coast mountains; 1998 was very hot; 2003 started with bad weather and had another bad spell later; 2007 was dry and hot; 2009 saw a wet first three days, saw four stunning days in the middle and then was quite damp for the rest of it; 2011 was also quite damp, with the tail-end of a hurricane blowing through towards the end; 2012 was wet for the first 9 days, with day 3 being particularly foul and windy, then hot and sunny for the rest.

      As 2012 saw the highest number of retirees in the history of the event, it will be interesting to look back in 2018 and see how many of 2012's First Timers don't return.

      Delete
  2. All utterly fascinating! Mind you, I'm still curious about how solo female Challengers fare.

    ReplyDelete
    Replies
    1. Louise.
      You're a very very naughty girl... You know Gayle won't be able to resist...

      I'm wondering if those Challengers who have died affect the statistics of those doing one Challenge and then not returning...

      Delete
    2. Louise - I've got my hands over my ears and I'm going 'La la la'. (Although perhaps it would have been more effective to put my hands over my eyes...)

      Alan - Eight One Timer Challengers are recorded as having died, of whom three died within five years of having done their only Challenge. With there being 1530 One Timers to date (including those in the 2007-2012 'may come back' category), I don't think that those three will be skewing things too much. In my opinion they wouldn't affect the statistics anyway - they're still One Timers, just not by choice...

      Delete
    3. It turns out that Alan was right - I couldn't resist. I've put some numbers together (interesting, too!). Another night I'll generate some graphs to illustrate them.

      Delete
    4. Excellent!! (Fingers drumming...)

      Delete
  3. Gayle

    Very interesting but at the risk of opening up another rabit hole,I wondered what,if any statistical correction you applied to weed out multiple challengers (female I guess) who may have changed their name due to marriage/divorce or whatever.

    Nick

    ReplyDelete
    Replies
    1. Name changes have been recorded and are accounted for in all of the numbers, so there should be no duplication. As the Entry Form asks how many previous Challenges you've done, I would expect that there's little scope for name changes being missed in the raw data.

      Delete
  4. Gayle. I think I can answer your question about the 2003 first timers not returning. We had dreadful weather for the first three days of that year: Heavy rain, strong winds and very cold. This was followed by another nasty spell a few days later.
    So even though there wasn't a particularly high drop out rate, the experience was probably such that the first timers wondered it was worth all the effort?

    ReplyDelete
    Replies
    1. So, in 9 years time, we'll be able to see in Gayle and Mick's annual graphs if the somewhat dodgy weather has had an effect on last years first timers...

      Delete
    2. Nine years, Louise? Five years is what I'm using as the not-coming-back cut off, so 2018 will be when that information is available. Not that I'm saying that I'm going to be getting the data each year and feeding it into the Monster Spread Sheet...

      Delete
  5. I know, at the time I had a reason for stating nine years. Now, of course, I've forgotten what it was. Nothing new there then!

    ReplyDelete