Sunday, 1 October 2023

Bonus Conversion Rates on UC

This post comes with a government health warning. In fact it comes with several:-

 - I am not a mathematician and am all too capable of making basic errors.

 - Even my miniscule knowledge about statistics tells me that 10 is a very small number from which to draw any kind of meaningful conclusions.

 - Even when statistics provide us with questions, it’s rash to jump to conclusions over the answers.

Right then, let us begin. After the current series of UC had been running for a few weeks, let’s say five, it started to strike me that the bonus conversion rates seem noticeably higher than in previous years. So I waited until after last week’s 10th match. By bonus conversion rate I mean the percentage of bonus questions each team gives correct answers to. So, for the sake of example, if one team were to only answer one starter correctly in the whole show, then if they answered none of the 3 bonuses that followed correctly their rate would be zero. If they answered one correctly it would be 33.3 (recurring). If they answered two correctly it would be 66.6 (recurring) and if they answered all three correctly it would be 100.

I feel that a bonus conversion rate is a useful measure to look at rather than a team’s overall score. I haven’t done analysis on these so I don’t know, but I wouldn’t be surprised if the teams’ scores are noticeably higher than 2023’s. I yield to no one in my admiration of Jeremy Paxman but it’s understandable that his speed of questioning in his final series didn’t match the speed at which Amol Rajan has been asking questions in he current series. The more questions that are asked in the match , the higher the teams are likely to score. Whereas the BCR is not as strongly affected by the number of questions asked in the show. This is why it is perfectly possible for runner-up teams to have higher bonus conversion rates than the teams that beat them.

I’ve compiled a table showing the bonus conversion rates for the teams in the first ten matches of the current series, and a second table showing the rates for the teams in the first ten matches of the 2023 series. Here they are.

Bonus Conversion Rates

Current Series

Winning team

BCR (Bonus Conversion Rate)

Runners Up

BCR

Manchester

54

Trinity, Cambridge

67

Aberdeen

55

Birmingham

49

Birkbeck

80

Oxford Brookes

88

Christ Church, Cambridge

43

Southampton

67

Emmanuel, Cambridge

59

Jesus, Oxford

60

UEA

54

Strathclyde

52

Hertford, Cambridge

69

Open University

73

Imperial

71

Balliol, Oxford

66

Sheffield

67

Loughborough

72

Warwick

61

Wolfson, Cambridge

62

Average winners

61

Average runners up

66

 

2023 series

Winning team

BCR

Runners Up

BCR

Durham

57

Bristol

57

Newcastle

64

Open

43

University, Oxford

48

LSE

67

Cardiff

49

Coventry

66

Royal Holloway

63

Cranfield

36

Queens Belfast

29

Glasgow

50

S. Andrews

49

Gonville and Caius

50

UCL

70

Sheffield

52

Christ’s

48

Oriel

42

Jesus

52

St, Catherine’s

50

Average - winners

53

Average – Runners up

50

 

So what does this show us? Good question. As I said we should be careful about drawing conclusions. So let’s start with what we CAN say.

The first ten matches of the current series have seen higher average bonus conversion rates by both winners and runners up than we saw in 2023.

What MIGHT this point to?

- Maybe the bonuses are easier on the whole than they were in 2023.

- Maybe the teams are stronger than they were in 2023.

- Maybe the average BCR (bonus conversion rate) was lower in 2023 than previous years and this season has seen things revert to normal.

- Maybe it’s just a blip that looks far more significant than it is because 10 is such a small number for any statistical analysis.

It will be interesting to do the comparison for the whole series. In fact it would be helpful to do an analysis of, let’s say, the whole seasons over the last ten years. Interesting, but not something I have the time to do at the moment.

Opinion

Yes, I thought I’d better highlight that this is just my opinion and not based on anything it says in the tables above. The thing which made me start to look at the whole question over whether the bonuses have been easier this year was the BCRs, not any gut feeling that the questions were easier. It’s difficult to quantify the relative difficulty of questions as it is. After all, to use a widely quoted adage, they’re all easy if you know the answers. But no, I haven’t been sitting in front of my telly thinking that the show has ‘dumbed down’ at all. There have been the odd bonus sets where I’ve thought that they were rather easy, but that’s happened in previous years as well.

However, I seem to recall that in one of the interviews prior to the broadcast of the first match, Pinki Chambers, BBC Commissioning Editor for Entertainment and Comedy, was quoted as saying  “the questions are harder.” Well, I know it’s the kind of boast you would make, but the evidence of the first ten shows suggests that this is going to be a hard claim for to prove.

No comments: