Sunday 20 September 2015

Championship Report: 2015 Times National Sudoku Championship

There was a time and place, dearest reader, when posts on this blog (and I really still do hate that word) were given titles of Radiohead tracks.  Whilst some things never change - such as hopefully the continuation of these reports - somethings do and so we'll all have to deal with a boringly functional title.

The build up to this year's championships was a little different from previous years, with organisation behind the scenes being handled by some Times admin staff; the usual suspects David Levy and Tony Corfe were to take over on the day.  There had been one major change however: in previous years the competition format meant that the grand finalists were the best 8 from solvers who had solved two preliminary sessions.  This year reverted to a much earlier format (going back to perhaps 2007 or 2008?) whereby solvers were entered into exactly one of two of the preliminary sessions and the best four from each session would make it through to the solver.

The letters through the post confirming entry were rather delayed, but this little bombshell was something that Mark Goodliffe for instance had initially missed.  I was not the only one pondering this little complication, as it soon became apparent that the top 3 qualifiers from last years competition - Myself, Mark and David Collison - along with two-time champion George Danker had all been entered into the afternoon session.  This seemed to be running the distinct danger of overloading one of the two preliminaries, which I think George took up with the organisers who then promptly asked both George and Mark whether they wouldn't like to take part in the morning session itself.

Anyhow, sticking to the afternoon session was entirely fine by me; I definitely like to have more rather than less sleep on a Saturday morning.  My trip down from Milton Keynes to London was very uneventful, the main highlight being that instead of trekking over to Tower Hill and walking over to Thomas More Square I could instead get the Northern line straight to London Bridge, emerge from its warren of tunnels and emerge at the base of that most delightful Qatari tribute to Barad Dur.  Having been suitably dazzled by that for a minute, I quickly located the shiny new News Corporation building, and decided to check in with the reception.  As it happened I was a little early, but not early enough to get any of the seats, so I decided to head to M&S and get some sandwiches and a bag of mints.

After returning we had to wait another 15 minutes before actually being let up to the top of Times Towers.  Some of the morning session had decided to stretch their legs, including 2013 champion Stephen Gerrard, as well as Mark and 2015's UK Sudoku Champion Heather Golding who I'd not realised was going to be there.  In retrospect it seems that the transfer of George and Mark to the morning session may have inadvertently lightened up the load for the afternoon!  This impression was further reinforced when I got to the top floor, and bumped into general UKPA hero Mike Colloby who had also been there in the morning.  After a good chat with him, and with George + his fiancee Sophie, I was able to digest a little information about how the morning session had gone.  There had been one particularly nasty puzzle, but the top 4 had ended up being Mark, George, Stephen and Mike.  Poor old Heather had had a bit of a nightmare, but those four certainly represented a strong half of the draw.

At this stage the following piece of trivia might help to explain my mind.  Having won in 2007, I made a stupid mistake in 2008.  I then went on to win in 2009 - and in 2010 then went on to make a stupid mistake.  In 2011 I made the grand final, eventually coming 3rd after freezing up in the final, but in 2012 made a stupid mistake (which agonisingly meant I could not go head to head with my great friend Rishi Puri).  In 2013 I stupidly missed the qualifiers so did not take part in the event.  In 2014 I made the grand finals, eventually finishing 2nd to Mark - perhaps you can spot what kind of forecast might logically be extrapolated for 2015?

With the addition of where I judged my form to be based on my performances on sudokucup.com and fed sudoku (check out my results for Saturday 12th), combined the volatility that is inherent to the I think it is safe to say I wasn't expecting too much from myself this year.  That's perhaps not say that I didn't consider myself the favourite - I do most years - but when any one from eight can win over a set of puzzles where any solver might miss something or make a silly mistake I don't think any one individual can have an a priori probability of winning greater than say 40-50%.  With Mark in tremendous form, I think I had my own chances at 25-30%, a little lower than in previous years, where I had not really come close to winning anyway.

So on to the preliminary.  I'm not sure there's a huge amount to say, other than the quirky booklet format remained, which meant that the first thing you see is puzzle 8 (as opposed to puzzles 5, 6 or 7).  Having not apparently recovered from my shaky online solving from the morning, I made a shaky start to number 8 before deciding I'd leave the puzzle half way through and instead have a look at the others.  This was a good idea - puzzles 5, 6 and 7 were much more to my liking and i whizzed through them all with out any incident.  After staring at puzzle 8 for a bit, I went into panic mode and made a "bifurcation" (rest assured that I still love that particular word, dearest reader).  My first attempt met with a fairly quick contradiction; the other route slowly led to the resolution of the puzzle - and after a couple of minutes of very carefully checking the filled grid for mistakes, I handed in after 20 minutes.

Having gone back through both sets of preliminary puzzles, I hadn't missed anything overly difficult in puzzle 8 - just a simple case of a naked triple in box 2.  This seemed in contrast with the stinker (puzzle 3) from the morning session which I think only resolves using x-wings (although please correct me if I'm wrong), and certainly put Mark's finishing time of 22 minutes in a very good light.

At the end of the hour I was very relieved to hear I'd made it through to the grand final.  It also turned out that I was 10 minutes quicker than the next finishers - of whom there were 5 or 6.  Joining me in the final would be Andrew Dyson, David and Ned Walker.

As always, the preliminaries are over at 2pm with an hour's break before the grand final starts at 3.  As always I wasn't particularly hungry, but I forced down a bag of crisps and a chicken and bacon sandwich anyway in one of the meeting rooms.  This was actually reasonably relaxed, with plenty of good banter from Mark and Heather (who will be joining me and David McNeill in representing the UK at the World Championships next month) amongst others in the room.  Nevertheless the excitement does slowly build as the clock nears 3, with the inevitable return to competition.

I think it's safe to say that as we were taking our seats all I was really focused on was being pleased that I had a seat in the front row, which removes any temptation to be distracted by any of the other competitors.  It had turned out that in one of the preliminaries, David Balderson had handed in answers to the four puzzles spread over two booklets in a time that would have placed him in the top 4.  I am not sure whether this was the morning or the afternoon session, but having initially been left out of the final David B was then invited to be the 9th man in the final.  This was certainly the fairest resolution to this unfortunate state of affairs, and an option that I sometimes wish was open to us for the 2014 WPC.  But like I say, I wasn't paying too much attention to this.  On to the final.

I started by making a conscious effort to solve the puzzles in order this time.  And things started well!  Puzzles 9 and 10 were both done before 7 minutes were up.  I know this because there was a great innovation for this year's competition: screens displaying countdown timers!  I was solving with good rhythm and spotting the naked subsets almost immediately.  I thought the same was true for puzzle number 11, until I was about 2/3 of the way through puzzle, where a contradiction suddenly popped up from nowhere.  When this happens it is possibly a devastating blow to your solving, because not only is your rhythm thrown completely off kilter, but it also introduces the nagging doubt in the back of your mind that it is perfectly possibly for you to make a silly error somewhere along the line and be totally oblivious to making it - until, if you're lucky - it blows up in your face and you catch it.  If you're unlucky it blows up much later when you think you've correctly solved it, and it turns out you haven't.  This time I did manage to catch it, and after the inevitably temptation of spending a few seconds looking for a quick fix (there is almost never a quick fix) I had to go with the option of rubbing everything out.

Solving a puzzle for the second time can be a very odd experience.  Some of the time, because the memory of key solving steps are fresh in your mind, it can end up being reasonably quick and any time lost is minimised.  However what can equally happen with your confidence shaken is that you completely miss all the steps that you had previously made, worrying about the faint imprints still visible in the papers and wondering whether you might actually be able to use them as a shortcut back into the puzzle (Don't do that!!!). Happily my experience this time round lay somewhere between the two.  I was quite careful up until the point I reached the area of the grid that had yielded the previous contradiction, but on entering in different numbers with no contradiction the adrenaline started kicking in and the rest of the puzzle went into the grid in something of a rush.

Which just left puzzle number 12.  I'm not sure exactly what the clock was showing at this point, but I'm guessing this was roughly 15 minutes in.  What my intuition was telling me was that Mark might well be up with me so the gloves had to be off for this last puzzle and this had to go down as quickly as possible.  Maybe this was letting emotion get a little the better of me, because I got about half way in before my progress ground to a halt.  Panic mode definitely set in and I set about looking for good places to bifurcate.  When you've put yourself under pressure this can sometimes be harder than it should be: all you need to do is go looking for pairs, or else numbers which can only go in one of two places in a row/column/box.  My first attempt looked like it was about to come up trumps for me, until with about 10 empty cells a contradiction loomed large.

I have previously described my guessing technique as "shambolic" in an explanation of why I never used to guess.  What might previously have happened is that I would forget where I had started bifurcating and then simply repeat the same branch - however maybe it's the case that I am now older and wiser and my technique has sured up a little.  What I had done here was for the bifurcation ensure that any entered numbers were entered in lightly and then circled - meaning they can very easily be rubbed out.  In this case this was about 1/3 of the grid so this definitely helped me out.

What can also happen in these circumstances where one branch yields lots of sequential information before getting to the contradiction, is that the other branch gives you sod all and then you have to make another guess.  For a fleeting second I thought this might be the case here - however I soon spotted the next deduction and I was on my way.  I was fairly confident that no-one else had yet handed in as I was filling in the finale few empty cells, so all I had to worry about was the structural integrity of the lead in my pencil which had suddenly seemed to have failed right at the last, with the final number to be placed only going in at the third attempt.

There was no time for doing anything other than a quick empty cells check before handing in, shortly after 21 minutes.

As is traditional, the remaining ~40 minutes sat in silence waiting for everyone else to finish entails a special sort of purgatory.  I attempted to do some killer sudoku from the book in the Times goody bag, but by this stage I think my mental energies were spent and I just couldn't concentrate.  Instead I was mentally preparing myself for the seeming inevitability of having made a mistake.  Indeed, when the hour was up, nothing seemed to happen for about 5 minutes before David Levy announced they needed a further 5 minutes for the final results.  With the 9 of us sat in limbo, I announced that this was indeed a special sort of purgatory, and circulated my bag of mints whilst musing over the afternoon's football scores.

Eventually Tony and David came back, announcing the final results.  It turned out that 3 of the finalists had only solved 2 puzzles correctly, and that 2 of the finalists had solved 3 puzzles correctly.  When it came those who had all 4, with my name not having been mentioned coupled together with the fact I had handed in first, I permitted myself a celebratory punch of the air to confirm my 3rd victory!

And I suppose that pretty much concludes this write-up.  The new Times puzzle editor had made the effort to come to the prize giving, which was very good of him, although I'm ashamed to say I don't recall his name.  There was no one-on-one session with a photographer this time round - alas the picture in the paper has my golden envelope rather neatly blocking the 2014 WSC/WPC logo on my red polo shirt.  And because I'm neither hugely vindictive, nor looking for an "angle", I won't say anything about the journalist who tried to interview me.  With this write-up I firmly believe I have set the record straight!

P.S. If anyone has copies of the grand final puzzles, please let me know!

1 comment:

  1. Sorry Tom, only just read this blog - good write-up, all very fair to everyone. Congrats as usual. I've got the Grand Final puzzles. Mark

    ReplyDelete

Contact Form

Name

Email *

Message *