Quick Cryptic stats for 2024

Many of us like to keep statistics on our solving times – it’s not compulsory of course, and certainly the QC is not competitive, but it offers some fun and is in keeping with the origins of the TfTT site, which was originally to compare solving times.

In past years people have commented in-line on the post of the day, which has meant they can both get in the way of those not interested, and get lost among other posts for those that are.  So this post is to allow people to share their experiences of 2024.

Let me start the ball rolling.  It was a year with a much higher than usual turnover of setters:  we had no less than seven new “names” (even if that does not necessarily mean seven new people behind them!), as we welcomed Asp (5 puzzles), Bjorn (just the 1), Bubbles (also 1), Cheeko (5), Jimmy (13), Lupa (5) and Pipsqueak (8).

On the flip side the year went by without any further puzzles from previous setters Bob & Margaret, Hawthorn, Marty, Monty, Parkin and Rodney.  Is it Au revoir or Adieu?

With this relatively high turnover in the setter common room, I was not surprised to find my times and completion rates more variable than in previous years.  While my average time decreased very slightly (from 11:50 to 11:47), the standard deviation of my times has increased quite noticeably, from 3.75 minutes to 3.93 minutes – that may not look much to non-mathematicians but in a population of some 300 entries, it is significant.

A case of new setters finding their range perhaps?  We certainly had some tough puzzles as they did so, and my SSS (“slowest successful solve”) was 23 minutes, a full 4 minutes longer than my previous SSS.  And my outright DNFs (including Finished-with-errors) went up from around 10% of puzzles to around 12%.

None of these figures include the best three setters of them all – John, Sawbill and Phil, who once again provided puzzles of the utmost quality and amazingly consistent standard (my standard deviation for their Sunday Specials is just 2.1 minutes).  Many thanks to them, and indeed to all our setters, for the year of entertainment they have given us.

Cedric

14 comments on “Quick Cryptic stats for 2024”

  1. 2024 saw a change to my solving approach. Feeling worn down by a perceived run of “are these getting harder” puzzles in late 2023, I decided I’d quit at 20mins. Occasionally, if I was within an answer or two, I’d continue on. This approach left me feeling fresher but didn’t need to last too far into March as I was finding the puzzles well-pitched and rarely taking longer than 30mins.

    In June I began putting my solves up on Youtube and this changed my mindset. No longer was I just doing my own thing against the clock, I found myself becoming mindful (worrying?) about whether watching a video of some bloke umming and ahhing and alphabet trawl is as boring as I’d expect!

    Although I’ve since found the Pause Recording button, rewatching my own solves has been interesting to see where time is being wasted as well as how often I can be saying the right things for an answer even if it doesn’t click immediately. Reading out clues, talking and explaining has certainly added some time to my solves in the latter half of the year but I feel a better rounded person for it.

    The summary of my 2024 stats …
    – 222/314 successful solves (71%)
    – Median average time was 18min16 with mean of 20:00.
    – SCC escapes 150 (with a further 31 DNFs in under 20mins).
    – Longest time spent on a puzzle 1:00:30. Thirty-one (31) puzzles took over 30mins plus 19 quits.
    – Fastest time 4:45 but it was a DNF, so official PB currently 5:36

    Timewise these are big improvements on my first two years of solving. My success rate has stalled at 71% this year. I put most of that squarely on tougher QCs because I know I’m better solver than I was a year ago but recognise some of it is on me as I tend to be racing through and less willing to persevere as I once did.

  2. ‘Marty’ and ‘Rodney’ were both pseudonyms of the late Richard Rogan so it’s unlikely we shall see any more, although a few puzzles left unpublished still turn up occasionally as by ‘Felix’.

    ‘Hawthorn’ was the former Puzzles Editor David Parfitt who relinquished the role in May 2022 and has evidently moved on to other things.

    The single puzzle by Parkin was a special with a Nina marking the departure of David as Puzzles Editor and the arrival Mick Hodgkin. Since it contained the hidden TATA DAVE I suspect that ‘Parkin’ was either RR or Mick who subsequently adopted the name ‘Peridot’.

    1. David Parfitt has taken over from John Grimshaw as the Times Concise compiler. He also does the daily word watch in the Times2 section of the paper. I will have to do a bit of analysis of my own times for 2024 and leave another comment here.

  3. Looking at the Times app it looks like I have been solving online since about the start of 2024. That would have been sporadic at first as my previous daily routine was to print out the 15×15 and solve it over couple of pints at lunchtime with a workmate. Stumbled across this site through trying to find explanations for clues we had failed to parse (I word I was hitherto unaware of). I then started routinely checking the comments on the 15×15 and the SNITCH to see if it tallied with our subjective experience of the relative difficulty from day to day. I probably only attempted the QC on a sporadic basis perhaps on the homebound commute if bored.

    I then started to ‘lurk’ frequently before making my first post I would say somewhere halfway through the year and now for the QC at least becoming an almost daily poster.

    Since packing old job in at the start of Summer the old routine has gone out of the window and my focus is very much on doing the QC often as a first act of the working day. The QUITCH stats tell me my average is 11:01 but not sure over what period that is. I only started looking at the clock towards late summer (probably coinciding with coming on here) and I got consistently faster as I improved. My sense is however that my times have tailed off a bit these last few weeks as not always found wavelength with some of the apparently new setters.

    I also note that over the last month I have only submitted one incorrect solution whereas this used to be several a month due to fat fingers solving on a tiny iphone SE. This used to typically cost me some of my better times as if it had been a particularly quick solve I would rush to submit it without checking. I now solve on a laptop which is both a. quicker and b. less prone to fat fingers.

    Anyway cracking corner of the internet this and pleased to be part of such an erudite community.

    Happy New Year all.

    Horners 1974

    1. Thanks, Horners.

      In answer to your question on the QUITCH, the average is calculated at the end of each month over the previous six months. A few things to note:

      • the “month” ends on the last Friday of the calendar month and the new month starts the next Monday (You can see the exact dates for each month on the “Trends” page.)
      • only data collected by the (Quick) SNITCH is used, so if it didn’t get recorded on the day, it won’t be included in the average
      • it’s not quite a strict average, but is the average of a statistical distribution (the gamma distribution), which is better for rejecting outliers. [You may find it’s a bit less than a strict average.]

      Probably more information than you wanted 😉

  4. Well done for setting up this link Cedric, and I’m happy to report my stats for the year 2024.
    I have in recent years recorded my weekly average on a Friday which has made my calculation a good deal easier. My stats confirm that my daily average based on my weekly total was 10.33; or just 33 seconds over my target time. Although I cannot be sure because I haven’t checked previous years, my feeling is that this yearly average may be my worst for quite a while.
    My best weekly average recorded in June was 7.07, with two further weekly averages of 7.43 and 7.55. My worst recorded time for a completed weekly average was 17.43 followed by 14.49. That is not to suggest that 17.43 was my worst time, but my worst weekly average. I think I recorded a 28 minute completion somewhere within the stats.
    I suppose one thing it does prove is that my ten minute target is realistic and generally attainable. My New Year’s resolution is to beat it in 2025!

  5. Interesting to see other solvers’ way of thinking about things! I started solving the QC in early December 2023 and since that time my weekly average has gone from about 40 minutes to about 20, though last fall it dipped for a few weeks into the mid-teens. My longest time was 1:17:18 and my shortest 5:40. I only recorded my completion times at first, so not sure what percentage I solved in my first six months. I think I became a bit less persistent and more impatient as my skill increased, but not sure. Since June 2024 my completion rate has been about 91%.

    I’m having a complicated life at the moment so not doing a lot of puzzling and not solving very fast, but the QC is still a joy when I get to it. Thanks to all who blog and comment for making this a real bright spot in my existence.

  6. Quite simple: started doing QC about two years ago, got quicker by average of ten mins. Since June i have lurched back three mins. Or are they a bit harder?

  7. Happy New Year to all. This is probably a good place (thanks for setting this up, Cedric) to explain why I don’t record a solving time (just in case anyone was wondering?!?). Crosswords must be all about enjoyment (no pressure, after all?), and the way I enjoy it is, rather than lock myself in a cell with a stopwatch, to do it over breakfast. The one, I believe, precludes the other. I do sometimes imagine you all with your stopwatches and wonder what your scenarios are!

    1. Couldn’t agree more. However, I do record whether I solved the crossword or not. This year I solved 233/263 (I only started recording them at the end of February). That’s 89%. Maybe I can make it over 90% this year. For the ones I failed to complete, I had either 1 or 2 clues outstanding, except for number 2726 (12/07/2024) when I gave up with 12 still left.

      1. Thank you – that’s a good idea! I might try that – January 2nd is a good time to start, isn’t it.

    2. I think your philosophy is excellent! I started tracking my times for a couple of reasons: when I started my times were so long that if I didn’t get faster I would need to give it up as too frustrating and time-consuming; the second reason was that I was struck by the, to me, novel idea, reflected in this blog’s title, of caring how long it took and thought I would try it out. It’s been an interesting psychological self-study, as I teeter between the insouciance of a Martinů and the teeth-grinding agonies of a GaryA. Caring and not caring at the same time is quite the challenge, and in my current rather weird life situation, a useful exercise!

Comment

Your email address will not be published. Required fields are marked *