Thursday, 31 October 2013

Bonus Question
How many Only Connect stats are too many Only Connect stats?

Following on from my analysis of Only Connect's missing vowels round, and with my team's (the Board Gamers - I'm one of the ones sitting next to Jamie Karran) next appearance on the show just around the corner, I thought I'd put together a post detailing some fun scoring statistics from the show. This is unashamedly geeky, however, so if you're not prepared for some hot, raw, game show stats action, look away now.

Round-by-round scoring

In my previous post I looked at how important the missing vowels round was in terms of the total points scored along with how often it allowed a team to come from behind and win (spoilers: not very). Here I'll be going into a bit more detail about round-by-round scoring, specifically how individual teams have scored. The graphs below show the points scored by each team across the four rounds of the show for the first seven series (I've not included any statistics from series eight as it's ongoing). With 99 shows in total you're looking at data for 198 teams in each graph. You'll probably want to right-click and open in a new tab to get a bigger version; there are a lot of delicious numbers here.

Distribution of round-by-round scores by each team

Those who are interested in the details can doubtless puzzle over the graphs themselves, so I shan't clog up the post with too much of my own commentary. Broadly speaking, however, all the rounds except for the wall have similarly shaped distributions, with missing vowels a bit more spread out (reflecting the greater granularity of scoring in that round). On the walls, meanwhile, the most common score is the maximum 10, achieved by one in three teams, while over half manage at least seven. My team's score of four a few weeks ago, meanwhile, puts is in the bottom quarter. The wall graph also illustrates a fun quirk of that round's scoring, where the 'six points' bar upsets a nice steadily increasing pattern. This is because it's actually quite awkward to score six points on the wall; to do so you have to either spot all four groups but only identify two connections, or vice-versa (you can't get 'three and three' because it's impossible to find exactly three groups).

It is of course fun to single out a few 'records' that these data reveal:

Round 1 (connections) highest score: 12. Achieved by the Alesman (vs. the Pool Sharks) in a series 4 first round match.

Round 2 (sequences) highest score: 10. Achieved three times: Gamblers (vs. Bowlers, series 3 second round), Wintonians (vs. Quitters, series 6 first round), and Francophiles (vs. Cartophiles, series 7 semi-final).

Round 3 (walls) lowest score: 1. Achieved by the Second Violinists (vs. the Footballers) in the first round of series 6.

Round 4 (missing vowels) lowest score: -3. Achieved by the Geocachers (vs. the Brit Poppers) in the first round of series 4.

Round 4 (missing vowels) highest score: 22. Achieved by the Epicureans (vs. the Courtiers) in an incredibly long missing vowels in the first round of series 4.

It's only fair to add that the Geocachers' -3 should be put into context; they had some ground to make up and so needed to be a bit trigger-happy. Curiously their opponents in that game, the Brit Poppers, faced similar misfortune in their next episode scoring -1 in missing vowels against the Radio Addicts. This gives the Brit Poppers the fairly impressive distinction of appearing on the only two shows in Only Connect history to see a negative score recorded in the final round.

It came as some surprise to me, when putting these together, that the Board Gamers' round two score of 11 in our opening match this series is a new record for sequences. I have subsequently reminded our captain that, had he listened to the two people shouting "point!" at him (and not inexplicably said "warhead") on a throwover, this 'achievement' would be even harder to surpass.

'Winning' rounds

Another idea that interested me was whether certain rounds are more or less likely to be won by the team that go on to win the match. It turns out there is some variation across rounds in this regard, albeit not much. At the top end, it's round 2 (sequences) that best indicates which team will win; being carried by the eventual winners 70% of the time. Missing vowels comes in second, falling to the match winners in 64% of episodes, while the connections round is much more of a toss-up at 59%. The team that wins the match only wins the wall 52% of the time, but this is mainly thanks to a large number of draws; 28% of wall rounds end in a tie.


Finally, I thought I'd look at how many rounds are won by the team that win the episode. Obviously there are a lot of drawn rounds, and so I've produced a comprehensive table here, but if we look solely at the rounds won by the winning team we get the figure on the right. It does, admittedly, barely warrant a graph at all, as most teams either win two or three rounds (38 and 42 times, respectively) but I like the colours. What's impressive is that 11 teams have won despite only carrying one of the rounds (although all of these were accompanied by tied rounds), while eight teams have achieved a 'perfect game' of sorts, winning all four rounds.

There are, of course, still plenty of fun questions to look at, and I'm planning to do some slightly more analytical work on these data at some point soon(ish). In the meantime, if you have any burning Only Connect related questions (and let's be honest, who doesn't?), feel free to ask (either in the comments or on Twitter @statacake) and I'll see what I can do!

Sunday, 27 October 2013

With over 250 buildings at least 150m tall Hong Kong has more skyscrapers than any other city

The team name: Crisis on infinite oeufs

The attendees
1) The statistian
2) The doctor

The ones that got away
1) What was President Gerald Ford's middle name?
2) The Marqueasas Islands lie in which ocean?
3) Who was the retired gunslinger played by Clint Eastwood in the 1992 movie Unforgiven?
4) Malbec is a variety of which fruit?
5) Hong Kong lies at the mouth of which river?
6) Which airline began operations in 1927, becoming the largest international air carrier in the United States before ceasing operations in 1991?
7) Lake Tana, the source of the Blue Nile, is the largest lake in which African country?

The answers


Poll results: 14 votes with 1 scoring 6/7, 2 scoring 4/7, 1 scoring 3/7, 4 scoring 2/7 and 1/7. 2 voters joined us on zero :(

The average voter scored 2/7!


The excuses


The alternative questions
1) Ford was, unsurprisingly, not the only President known by a different name earlier in life. Which President's surname at birth - shortly after the death of his biological father in 1946 - was Blythe?
2) The Marquesas Islands are a group of Polynesian Islands. The name Polynesia derives from the Greek for which two words?
3) Despite his extensive CV, Clint Eastwood has only won Academy Awards for two films. One was Unforgiven, which earned him the Best Director and Best Picture Oscars, a feat he repeated with which 2004 film?
4) Which American author penned the 1939 novel Grapes of Wrath, earning him the Pulitzer Prize (as well as contributing towards the award of his 1962 Nobel Prize)?
5) First released in Japan in 2006, which game partnered Pokémon Pearl in the then-latest instalment of the popular video game series?
6) Pan Am were notable for the use of what word for a type of fast sailing ship in the names of their aircraft?
7) Since the secession of South Sudan in 2011, which country has overtaken the former Sudan as the largest African nation (by area)?

The answers

Sunday, 20 October 2013

The blue, white and orange tricolour of New York City reflects its Dutch history (and its former name of New Amsterdam)

The team name: The Raton Lovers (reception: suspiciously good for a French-for-raccoon based pun)

The attendees
1) The statistician
2) The doctor

The ones that got away
1) Which actor offered Demi Moore one million dollars to sleep with her in the 1993 film Indecent Proposal?
2) Kathmandu is the capital of which country?
3) In the name of the American broadcast network, what do the letters CBS stand for?
4) Name the three major airports serving New York City
5) From which country did Tunisia gain independence in 1956?

The answers


Poll results: 16 votes with 7 scoring 4/5, 4 scoring 3/5, 4 scoring 2/5, and 1 scoring 1/5. No-one joined us on zero :(

The average voter scored 3-and-a-bit/5!

The excuses


The alternative questions
1) Robert Redford founded which international film festival in 1978, which has been held annually in Utah ever since?
2) Nepal shares its borders only with India and China, a property shared by which other country?
3) Who hosts CBS's Late Show, a late night chat show that has been running since 1993?
4) Before being renamed in honour of the late president in 1963, what name - shared by a Scottish indie rock band - was previously used to refer to JFK airport?
5) Along with Tunisia and Turkey, six other UN member states have national flags featuring a crescent and (single) star symbol. Name four of them.

The answers

Thursday, 10 October 2013

One of Connecticut's nicknames is "The Land of Steady Habits"

The team name: There's no 'I' in Équipe (reception: non-existent)

The attendees
1) The statistician
2) The doctor

The ones that got away
1) In which month is Earth Day celebrated?
2) What is the capital of the US State of Connecticut?
3) In January 1953 over 70% of US television sets were tuned in to watch Lucille Ball give birth in which TV show?
4) Managua is the capital of which Central American country?
5) Hoss, Little Joe, and Ben were characters in which US TV series?
6) What is the name of Romeo's cousin and friend in Shakespeare's Romeo and Juliet?
7) How many years of marriage are celebrated with a lace wedding anniversary?

The answers


Poll results: 11 votes with 1 scoring 7/7 (!), 1 scoring 6/7, 2 scoring 5/7, 4/7, 3/7 and 2/7, and 1 joining us on zero.

The average voter scored a bit under 4/7!



The alternative questions
1) The first Earth Day was celebrated in 1970, but what time system - used in many computer operating systems - is defined as the number of seconds since January 1 of that year?
2) Hampton Court Palace was originally built for Cardinal Thomas Wolsey, a favourite of which king?
3) With 71.7% of all US television sets tuned in, more American households watched Lucy give birth than the inauguration coverage of which US president, the following morning?
4) With the exception of countries whose name (or at least an entire word/part thereof) is the same as their capital, such as Monaco (Monaco) or El Salvador (San Salvador), Nicaragua is unique in sharing the last 4 letters of its name with its capital Managua. Which Asian country is the only country to share the last 3 letters of its name with its capital city?
5) Bonanza is the third longest running live-action series in US broadcast history behind fellow western Gunsmoke and which rather more recent police procedural?
6) Who directed Romeo + Juliet, the 1996 film that cast the classic tragedy in the modern-day (and fictional) setting of 'Verona Beach'?
7) If you gave sugar as a sixth wedding anniversary gift, and should you live long enough, what might be an appropriate gift at your 216th wedding anniversary?

The answers

Monday, 7 October 2013

Bonus Question
Does missing vowels matter?

Delicious vowels
A charge often levelled by (some) Only Connect viewers is that the final round - in which quickfire word puzzles are solved on the buzzer - carries too much weight. All too often, it is said, a team works hard to establish a lead over the first three-quarters of the contest only to have their surely deserved victory snatched away by a team who are slightly quicker on the draw. My team's win (I make up one third of the Board Gamers along with Hywel Carver and Jamie Karran) over the Globetrotters last week seemed to stir the hornet's nest again, as our 19-16 deficit after the connecting walls turned into a 26-21 victory. (Although it probably didn't help that our somewhat unorthodox wall strategy made us look like dribbling idiots.)

Setting aside the rather more philosophical question of why one round of an established quiz show should carry less weight for a moment, I thought it would be interesting to see if I could find out just how powerful (or not) missing vowels really is. To answer, I did what any statistician would do, and looked for the data. Through a combination of question editor David Bodycombe's own records (thanks David!), episode summaries at Mastermind champion Dave Clark's blog (thanks Dave!), and my own trawling through old episodes on YouTube (thanks Dav- I mean thanks me!), I put together what I'm pretty sure is the only comprehensive dataset of regular series Only Connect scoring in the universe. With it I'm able to look at all sorts of fun Only Connect-related questions (some of which will follow in later posts) including whether or not the missing vowels round is particularly important.

First things first, what about all these come-from-behind victories I've heard so much about? It turns out that they're not that common after all. In 99 episodes over the first seven series there have been just 14 occasions where a team losing after the walls have gone on to win - and two of those were on a tie-break. What's more, these turnarounds are seldom particularly dramatic. Of the 14, five were carried out by teams who were losing by just one point after the walls, while four were by a team trailing by two. The biggest deficit that's ever been overturned in missing vowels is four points; you can score more than that in one question in the earlier rounds.

Continuing our journey into the unremarkable let's have a look at how many points are scored on average in each of the four rounds. This should give us a very rough impression of how each round contributes to a team's performance. Check out the fancy table below.

Average (combined) points scored per
Only Connect episode, series 1-7.
It may surprise some that it's the connecting wall that comes out on top, with a whopping 13.5 points scored per episode between the two teams. Missing vowels is second with 11.3 points per show, while sequences and connections are a bit further back with 8 and 7.4 points respectively. It's important to note that these figures aren't directly comparable as each round has slightly different scoring mechanisms - a team can't score more than ten in the wall round, for instance - but the numbers still provide some food for thought. There are also some curious trends across the series (the data for which can be viewed here) - teams seem to be getting gradually better at the first round (or perhaps the questions are getting easier) - but those are left to the interested reader. Overall though, it's fairly remarkable how well-balanced the rounds are, particularly in more recent series.

It seems, then, that while missing vowels is an important part of the show it's nowhere near as make-or-break as the Internet (or at least the vocal parts of it) would have you believe. On average you'll see a turnaround once in seven shows, or roughly twice a series - the rest of the time you may as well not bother with the final round at all; the outcome would be the same. Moreover, when it does happen, it'll more than likely be a lead of only one or two points that's overturned, which for a relatively large part of a show seems pretty reasonable to me.

So why has this myth persisted? The most likely reason is that while the first three rounds take a relatively long time, missing vowels seldom lasts more than two minutes. Points are therefore scored at a much quicker rate than earlier in the show giving the impression that the round as a whole is cheap. The walls, meanwhile, have their points 'disguised'; they're only added for both teams at the very end of the round.

One final observation is that there seems little reason why a team leading before missing vowels might be expected to go on to lose. After all, a team who are better over the first three rounds are, one might expect, likely to be better in the last one as well. Pleasingly, this hypothesis is borne out by the data: there's a (statistically) significant association between a team's performance in missing vowels and their lead before it.

In conclusion:

Ones that got away: Only Connect Special 1

Warning: the following contains spoilers for Series 8, episode 2 of Only Connect, first broadcast on BBC4 last Monday (30th September). I'd be surprised if anyone reading this both cares about the result, and yet hasn't seen the show, but if that's you look away now!

More like bored lamers, amirite?!
It's a Ones that got away Only Connect special! Rather than the usual post mortem of one of our pub quiz adventures I'll instead be going through our performance on last week's episode of Only Connect. So if you're interested in how the Board Gamers (Hywel Carver, Jamie Karran and me) would have fared with the other team's questions, or what exactly was going through our heads on that connecting wall, then read on. Naturally this will take a slightly different format to a normal OTGA post, and so rather than only look at the questions we got wrong, I'll instead be giving every question a small comment. First though, a tiny bit of insight into what it's like behind the scenes.

Before the show

An early morning (well, for a Saturday) saw us on a train out of London Paddington at 8.42am. The journey largely consisted of going through a book of connection quizzes and trying to explain to a Turkish woman how to use her mobile phone. At the studios we were shown to our dressing room and Hywel got told off for only bringing checked shirts. The first episode (between the Lasletts and the Pilots) was being recorded while we went through the pre-show motions. At some point Jamie decided to do roly-polies in the corridor, shortly after which the Lasletts - fresh from victory - appeared. A brief 'you're all happy and relaxed while I'm a bit stressed' conversation followed before we were whisked off to get our microphones. It was at this point we finally met the Globetrotters.

Predictably, Only Connect tends to attract some pretty seasoned quizzers, and so we were half-expecting to recognize our opponents from other game shows. They turned out to be entirely fresh faces, but it did emerge just before filming that they were part of a team from the Quiz League of London. This immediately had us slightly on edge; every Only Connect series-winning team has featured at least one QLL member, so our initial optimism from not playing a bunch of  'old quiz blokes' was short-lived. Still, we didn't have much time to let that sink in as all too soon we were heading into studio.

Did I mention that the production team are lovely? Well they are. Victoria Coren is too, obviously, but there wasn't time to think about that: the intro music had played, our awkward introduction was over, and the game was afoot.

Round 1: What's the connection?

The questions
1) Music: No Memory by Scarlet Fantastic; I'd Rather Go Blind by Ruby Turner; Love Don't Live Here Anymore by Rose Royce; Get the Party Started by Pink
2) Büsingen, Switzerland; Monaco, France; Berwick, Scotland; Swansea, England
3) Jet-propelled unicycle; Dehydrated boulders; Invisible paint; Iron bird seed
4) 1968 Black Power salute athletes; Curley in 'Of Mice and Men'; Freddy Krueger; A baseball fielder
5) The World: 1804; Facebook: 2012; Africa: 2009; China: 1982
6) Pictures: Lady Bird Johnson; Professor Green; Queen Latifah; Prince Naseem Hamed

The answers


The excuses
1) No Only Connect team wants to get the music question, and we were fortunate for it not to be our downfall here as our admittedly vague answer of 'colours' was accepted. The first three clues were well out of our musical comfort zone, and while I did at least recognize Love Don't Live Here Anymore, I would never have been able to tell you the artist. Fortunately we knew Pink thanks to Jamie's brief obsession with playing that particular track on EZ2Dancer machines (and there are people out there who think his extreme nerdiness is a front) and figured there weren't many alternatives for connections that could include that particular name.
2) A lovely question, but one I'm glad we didn't pick ourselves (I'm a pretty big football fan but only knew Swansea of the four). The mention of Berwick will send many a quizzing brain to "most northerly town in England" and even when it was thrown over we couldn't get beyond the idea of extreme locations. All credit to Globetrotter Suda for spotting this one, albeit just too late. Meanwhile, a large number of people now think (albeit inaccurately) that a guy called Hywel didn't know where Swansea is. Every cloud...
3) Iron bird seed was of course a big giveaway, but we got to it too late for the penny to drop. My only reliable memories of Road Runner, meanwhile, are limited to when I collected Looney Tunes Tazos during a 1996 Walkers promotion. Obviously.
4) As something of an Olympics nerd, and having studied (as I think everyone of a certain age) Of Mice and Men for GCSE English Literature, I had this off the second clue. I was a touch surprised the Globetrotters took the third (which, ironically, wouldn't have helped me), but better safe than sorry, of course.
5) The only question in the whole show I think I'm really disappointed about missing. We got preoccupied with Facebook's 2012 floatation on the stock market and hypothesized that the world's first stock exchange might have opened in 1804. This is, of course, utter nonsense, but it goes to show how quickly an idea can take hold under studio pressure. Having wasted time on this we were then running short to come up with any other inspiration and so missed something that was very much in our wheelhouse. Fair play to the Globetrotters for hoovering up the bonus.
6) Another question I'm glad we didn't have to face. Our record of recognizing people in pub quiz picture rounds is beyond awful, and I'm thankful that the nation was protected from seeing this for themselves.

So after a pretty poor opening round we were trailing 4-1. This was a disappointment but not too unexpected; we knew rounds two and four were our strong suits, so we were hoping this would prove only a temporary blip.

Round 2: What comes fourth?

The questions
1) 4: Parade's End; 5: Forsyte Saga; 6: The Chronicles of Barsetshire
2) Fred Dinenage children's TV show, IPC Media celebrity magazine, Red and green paints mixed
3) Pictures: Graph of y = x4; Graph of y = x3; Graph of y = x2
4) Flight; Shaft; Barrel
5) Adenine; Cytosine; Guanine
6) Pictures: Courtroom scene; Kevin Pietersen; A vinyl record

The answers


The excuses
1) After a bad start a literature question was not what we needed, but just as we were going to take the third and final clue Hywel spotted the possibility of it being books in a series. This would make the last in the sequence seven, which made a lot of sense given the existence of such famous examples. Hywel suggested Chronicles of Narnia but we played it as safe as possible with Harry Potter.
2) We were nowhere on this one, having forgotten who Fred Dinenage was and having no clue on the magazine. Not sure if we'd have got this with all three clues, but no way of knowing now.
3) If there's one thing a statistician likes to see on a quiz show, it's graphs (well really, we like to see them anywhere, but quiz shows especially). We were almost tricked by this one, though, thinking the first picture was y = x2 and the sequence was heading up to x5. Hywel's observation that it was really too flat to be x2, and mine that y = x is a much more likely answer than y = x5 meant we picked up another valuable three points.
4) Rather frustratingly, both Hywel and I were telling Jamie the correct answer when it was thrown over, only for him to whip out his warhead from nowhere. Fortunately (for him) it didn't matter in the long run, but 'point' became something of a running joke within the team for the rest of the day, because we're just that funny.
5) Following on from our graphs question, one about DNA bases was also well received. I'll admit I was nowhere (I have long since expunged GCSE biology from my memory) but after Hywel and Jamie spent a very long time persuading themselves thymine had to be the correct answer we bagged a tasty five points to give us the lead.
6) I know very little about cricket, but I do know that Kevin Pietersen is often referred to as KP. With the resulting sequence landing on MP I would have been tempted to go for this on three (although suspect I would have chickened out in reality). The Globetrotters took the third clue and our hopes of picking up a bonus were swiftly dashed.

Some rather delicious questions for us meant we'd turned the game around swapping a three point deficit for a three point lead. A respectable performance on the wall and we could hope to go into missing vowels at least level...

Round 3: The connecting walls

Less than 24 hours later >:(
The Globetrotters' wall can be played on the Only Connect website (wall 348). Groups (and connections) under the button:



Fun Only Connect fact: teams don't watch each other's wall performance. Instead you lurk in the green room eating sweets while the (lovely) production team tell you to stop worrying. Consequently the first we saw of the Globetrotters' wall was when it was uploaded to the Only Connect website just before the show last week. Playing at home Jamie and I managed to score 10, helped no doubt by the lack of pressure and the presence of board game manufacturers (although notably none of these happen to produce any particularly good board games).

As for our wall, that's number 349 on the website, groups and connections below:



So yes, this didn't go brilliantly, but on retrospect was always likely to defeat us. A lot of Twitter conversation (understandably) took issue with our 'strategy', which to the casual observer appeared to be centered on pressing at random and hoping for the best. In reality most of what we did was reasonably thought out, even if it did look a little manic.

With little inspiration hitting for possible groups of five (the standard four plus a red herring) I, as the 'designated presser', spent much of the time trying three from one group with every other thing on the board (for example trying canopic, jam, bell and everything else in the hope of finding that fourth jar). Having done quite a lot of wall practice in preparation, we were all used to playing in this way and so could keep thinking about other ideas while essentially getting a bunch of free attempts at hitting a group. (Fun fact, even if we were guessing entirely at random, then at an average rate of one group of four every two seconds you'd still have a 15% chance of hitting at least one purely by luck. If there's one way to guarantee you won't solve the wall it's by not trying.)

As it was, we were always going to be in trouble. None of us had heard of a kilner jar (despite subsequently learning that everyone I know seems to own one) and I'm still utterly bemused by how much everyone else seems to know about helicopter manufacturers. We probably should have got the bands, and might have done on another day, but there's no accounting for what being on the spot can do to you. Walls are much more about fundamental general knowledge than you might think, and ours was sorely lacking here. Still, four points for the groups meant we were within touching distance going into missing vowels, which was better than nothing.

Round 4: Missing Vowels

From the comfort of our own homes we knew that missing vowels was potentially a 'get out of jail free' card for us. Unfortunately, we also knew that when you watch on TV there's a small delay between when you see the clues and when the contestants do (necessary so as to provide time for the viewers to be shown the correct answer after it's given in the studio). Consequently we appreciated that any sort of missing vowels confidence could prove misplaced. There's not much else to say about how the round unfolded, but I remember realizing we'd probably done it after Suda's unfortunate mis-buzz on the 'furnished flat' clue. My performance here was poor (I've had a fear of buzzers since being rubbish on University Challenge and only picked up a couple of answers here) but a strong showing from Jamie, and especially Hywel, got us over the line 26-21.

So that was that. As we had hoped, it was solid scoring on sequences and missing vowels that saw us through, while I won't forget about that wall in a hurry. Victory meant that we had avoided the worst-case scenario of turning up only to lose twice, so there was relief all round on that front. It also meant we had the rest of the day off, so we did what any self-respecting quiz team would do: a wild night in a Cardiff Wetherspoon getting drunk and playing a quiz machine.