Rank the Revival: A Numbery-Wumbery Breakdown (Part 1)

What’s the best rated year? Most divisive episode? Best showrunner? Guest contributor Joshua Yetman delves into the stats.

Across February and March, every episode of the revival was scored by the DWTV community, summarising 10 years of wonderful television with a comprehensive rank of episodes averages (summarised further by the diagram above, which really shows it’s literally been one hell of a rollercoaster!). But there’s a lot more to learn from the poll results than just a rank of all 117 episode averages, as my two articles on this matter seek to explore. Part 1 will act as, effectively, a “number dump” (though with added comments and explanations) for those of you who like cold, hard numbers. Part 2 of this article will perform deeper analysis in certain areas of the revival.

(1) Best to worst year of the revival

The average of a year is calculated as the average of all the episode averages in that year, though, for the sake of consistency, The End of Time Part 2 is lumped in with 2009, despite the fact it aired in 2010, as it is the second part of a story which began in 2009. The rankings of the yearly averages are as follows:

• 1: 2010 – 7.805
• 2: 2009 – 7.743
• 3: 2008 – 7.725
• 4: 2014 – 7.699
• 5: 2005 – 7.603
• 6: 2012 – 7.464
• 7: 2013 – 7.461
• 8: 2007 – 7.444
• 9: 2011 – 7.195
• 10: 2006 – 7.060

So, interpreting these years as series, Series 5 comes out as the best series, and Series 2 as the worst series, an expected and fair result. Still, it should be noted that the average of all these years is above 7/10, the benchmark for good, so every year was good on average!

(2) Most divisive episodes

To measure how divisiveness an episode is, we use a statistic called standard deviation, which measures the spread of data away from the average. Essentially, the higher the standard deviation of an episode, the more divisive it was amongst the DWTV community. The top 10 most divisive episodes of the revival are as follows, with the standard deviation given in brackets:

• 1: Love & Monsters (2.704)
• 2: The Doctor, The Widow, and the Wardrobe (2.555)
• 3: The Rings of Akhaten (2.371)
• 4: Fear Her (2.355)
• 5: In The Forest of the Night (2.347)
• 6: The Wedding of River Song (2.315)
• 7: The End of Time, Part 2 (2.302)
• 8: Let’s Kill Hitler (2.282)
• 9: Nightmare in Silver (2.267)
• 10: Evolution of the Daleks (2.263)

Love & Monsters naturally tops this list, though the reason why may not be obvious. Let me explain the reason. Although Love & Monsters has a reputation of near universal dislike (which wouldn’t imply it’s very divisive if everyone agrees its poor), in recent years the episode has developed a notable cult following and greater fan support, which essentially polarises the results completely, hence its position here. A similar episode in that respect, which also makes it into this list, is The Rings of Akhaten, which is also remarkable for how it polarises almost all fans into “love it” and “hate it” camps.

(3) Least divisive episodes

The top 10 least divisive episodes of the revival are as follows, with the standard deviation given in brackets:

• 1: Blink (1.298)
• 2: Silence in the Library (1.323)
• 3: The Empty Child (1.344)
• 4: Mummy on the Orient Express (1.354)
• 5: The Doctor Dances (1.366)
• 6: Forest of the Dead (1.368)
• 7: The Day of the Doctor (1.454)
• 8: The Eleventh Hour (1.501)
• 9: Utopia (1.509)
• 10: The Family of Blood (1.510)

The near universally acclaimed Blink naturally takes the crown of being the least divisive episode of the revival. It’s also interesting to note that 7 out of the 10 least divisive episodes were written by Moffat.

(4) Most and least divisive years

(Average of standard deviation is given in brackets)

I define the divisiveness of a year to be the average divisiveness of each episode in that year (so the average of the standard deviation of each episode), whilst I define the consistency of a year as the divisiveness of the episode averages (so the standard deviation of the average of each episode).

So, the year with the most divisive stories – on average – was 2011 (2.032), followed by 2009 (2.027). The year with the least divisive stories was 2005 (1.734), followed by 2010 (1.776).

(5) Most and least consistent years

(Standard deviation of averages is given in brackets)

The least consistent year was 2006 (1.365) followed by 2011 (1.125). Both of these seem appropriate, given the fact that 2006 saw witness to the two least popular episodes of the revival along with some incredibly popular ones as well, and a plot of 2011 (shown above) really emphasises how up and down Series 6 really was. As much as I personally love the series, I think I’ll now refer to it as the “yoyo” series as it seems to be all over the place quality wise, at least to the wider fandom!

On the other side of things, the most consistent year was 2012 (0.553) followed by 2009 (0.785). As these years only featured a handful of episodes, these results shouldn’t be surprising.

(6) Best and worst writers

The top 5 writers of the revival, based on the average score of their written/co-written episodes, is as follows (with the number of episodes they have written given in brackets for scale):

• 1: Robert Shearman – 8.827 (1)
• 2: Richard Curtis – 8.814 (1)
• 3: Jamie Mathieson – 8.741 (2)
• 4: Paul Cornell – 8.464 (3)
• 5: Matt Jones – 8.245 (2)

The two showrunners are placed as follows:

• 6: Steven Moffat – 8.199 (34)
• 14: Russell T Davies 7.397 (31)

Although Moffat ranks only 6th on average, it is to be noted that the entire top 5 stories, 10 of the top 20 stories, and 17 of the top 30 stories are all written by Moffat, a testament to the sheer wiring prowess of this incredible man. It should also be noted that Moffat’s average in the RTD era was 9.05, whilst his average in his own era was 8.02. Although it has fallen (which is natural due to the high quantity of episodes he writes now compared to the RTD era), an average of 8.02 is still excellent.

The bottom 5 writers are as follows:

• 20: Stephen Thompson – 6.515 (3)
• 21: Helen Raynor – 6.450 (4)
• 22: Stephen Greenhorn – 6.372 (2)
• 23: Matthew Graham – 5.732 (3)
• 24: Frank Cottrell Boyce – 5.324 (1)

The calamitous reception to In The Forest of the Night puts Frank Cottrell Boyce in stone dead last. Mark Gatiss escapes the bottom five (just), being placed sixth from last with an average of 6.534.

(7) Most and least consistent writers

The most consistent writer over different stories is currently Jamie Mathieson with a standard deviation of 0.09 (it’s so low as Mathieson has only written two stories, and both of them are similarly acclaimed).

The least consistent writer (over different stories) is Neil Gaiman with a standard deviation of 1.216. Given the strong positive reception to The Doctor’s Wife, and the dismal fan response to Nightmare in Silver, this result certainly makes sense in context!

Now, let’s consider the consistency of the writing of our two showrunners:

• Steven Moffat – standard deviation of 0.829
• Russell T Davies – standard deviation of 0.947

Both showrunners are placed low in the rankings for consistency, which is understandable given the sheer quantity of stories both have contributed to the show. They’re bound to have their good days and bad days, a cycle less prolific writers are more likely to avoid. Plus, it should be noted that, despite the fact these figures for both showrunners are relatively high, in absolute terms, they’re still fairly low and do not indicate massive deviations in quality from either individual, though Moffat is the more consistent of the two.

(8) The distribution of scores

The episode which received the highest proportion of 1/10 votes is – unsurprisingly given its reputation – Love & Monsters with a tremendously high 18.91%. The episode which received the lowest proportion of 1/10 votes is – again, unsurprisingly – the fan favourite Blink, with 0.45%. On average, 2.37% of the votes for each episode were for the 1/10 score option.

The episode which received the highest proportion of 10/10 votes is The Day of the Doctor with a colossal but understandable 73.16%. The episode which received the lowest proportion of 10/10 votes is – rather curiously – The Lazarus Experiment, with just 2.77%. Still, as this is the lowest proportion of 10/10 votes, it shows that every episode had at least one 10/10 vote, which signifies that our tastes and opinions as a fandom remain as beautifully varied and diverse as ever. Even generally unpopular episodes like Fear Her and Love & Monsters had a relatively significant amount of 10/10’s, the latter of which receiving a notably high proportion of 5.15%.

Furthermore, on average, 19.69% of the votes for each episode were for the 10/10 score. The most commonly chosen score was, on average, 8/10, with an average proportion of 19.94%. The least commonly chosen score was, on average, 2/10, with an average proportion of 1.38%. Finally, the 10/10 score was the most commonly voted score for a whopping 37 episodes (19 Moffat era episodes, and 18 RTD episodes) of the revival.

This all shows that the fandom (or, at least, the DWTV community) is plainly enthusiastic and highly positive about the revived era, a tribute to either how nice we are, or (more likely) how much sensational television we’ve had over the past 10 years! Plus, the Moffat era and RTD era seem to have performed almost equally well, a factor we will explore in greater detail later.

Now, let’s move on to the “5/10+ proportions”, which is a statistic of my own making. The 5/10+ proportion of an episode is, simply, the percentage of readers who gave said episode a 5/10 or higher in the DWTV polls. I believe it provides a slightly clearer indication as to how well received an episode is, by directly considering the distribution of votes.
The three highest 5/10+ proportions are:

• 1: Silence in the Library (98.72%)
• 2: Blink (98.57%)
• 3: Mummy on the Orient Express (98.53%)

70% of episodes attained a 5/10+ proportion of 90% or more. The average 5/10+ proportion was 90.43%.

The three lowest 5/10+ proportions are:

• 1: Fear Her (42.69%)
• 2: Love & Monsters (48.27%)
• 3: The Doctor, The Widow & The Wardrobe (64.73%)

So, Fear Her and Love & Monsters are the only episodes of the revival to receive more 4/10 or below votes than 5/10 or above votes. Also, note the considerable jump between positions 2 and 3, which again emphasises just how unpopular Fear Her and Love & Monsters are compared to the rest of the drudgery of the revival!

(9) Best Companion and Doctor

(Based on episode averages)

Now, obviously, it would be imprudent to judge which Doctor and companion is the best based on episode averages, but the following statistics will still highlight which companion and which Doctor had the best stories on average.

For companions:

• 1: Donna Noble – 7.795
• 2: Clara Oswald – 7.595 (note: this figure excludes episodes focusing on a Clara echo).
• 3: Amy Pond – 7.567
• 4: Martha Jones – 7.463
• 5: Rose Tyler – 7.424

So, bolstered by Series 4’s strong fan reception, Donna tops this list as the companion with the best stories on average. Rose, though a very popular companion in her own right, is let down by the relatively poor reception to Series 2.

For Doctors:

• 1: Twelfth Doctor – 7.699
• 2: Ninth Doctor – 7.588
• 3: Eleventh Doctor – 7.486
• 4: Tenth Doctor – 7.446

Peter Capaldi’s brilliant Twelfth Doctor tops this list, a testament to the perceived strength of Capaldi’s first series. Now, there may be allegations of “recent episode syndrome” playing a big part in Twelve’s placement here, and that’s something we will be looking at in part 2 of this article.

In any event, it’s interesting that the Tenth Doctor is placed last in this list. Although he is probably the most popular Doctor of the revival, it seems the community feel he had the weakest stories on average (though an average of 7.446 is far from being anyway bad). This is understandable, however, due to the fact he had three whole series in the role, so there was more chance for bad episodes to crop up than in the single series runs of Twelve (to date) and Nine, pulling down the average. This applies to Eleven as well.

(10) Best showrunner

The debate as to which showrunner is better – the Giant Welshman or the Mighty Moff – is seldom a pleasant or constructive one. I personally love and respect both, but some Whovians are rather fervent in their opinions on this matter, and I was initially concerned of the possibility of instigating such a debate with this statistic. But then, I saw what the statistics actually were, and I was more encouraged. Because they’re thankfully close. Very close indeed:

• RTD era average: 7.477 (standard deviation of 1.83)
• Moffat era average: 7.535 (standard deviation 1.90) (Revival average: 7.505)

Now, taken at face value, these figures indicate that the Moffat era is better than the RTD era, but we have to remember that these are samples from a wider population. Depending on what part of the population you sample, you will likely end up with a different sample average, and so samples do not necessarily reflect the population. Thus, the overall population may feel completely differently! There’s no way of knowing this for sure, but we can actually test this possibility using some pretty nifty statistics called two-sample hypothesis testing.

I won’t get into the nitty gritty of hypothesis testing, as it’s quite a bemusing concept (at least at first), but, essentially, hypothesis tests consider the how likely a given sample statistic would occur under an assumption about the population. If this probability is low, then our assumption may be wrong. Our assumption in this case is that the RTD era and Moffat era are equally loved by the wider population.

For stats nerds, I will briefly summarise the procedure of this test. The sample size was so large for this poll series (approximately 3000), we can assume that the test statistic is normally distributed despite the fact the population variance for both eras is unknown. The test statistic in this case, using a pooled sample standard deviation of 1.865, is 1.204. At the 5% level of significance, this is not significant!

Ok, in plain English for everyone else, there isn’t enough evidence, from this sample alone, to suggest that the wider population have a preference for either showrunner. That’s right, with the power of statistics, I can resolutely conclude that, based entirely on this poll, the RTD era and Moffat era are equally loved in the wider population! Isn’t that brilliant? Of course, in reality, there will almost certainly be a preference either way, but I can’t conclude anything about that from this poll.

So that concludes part 1 of this article. In part 2, we will perform some deeper analysis of the data in order to answer some burning questions. Is recent episode syndrome a thing? Which part of a double-parter is the best part? What is the best episode 1, 2, etc? All will be answered.

Now, as I couldn’t fit all the statistics I wanted into this article, if anyone wants further information, like how divisive a certain episode was, how many people voted 10/10 for a certain episode, etc, just ask in the comment section and I’ll endeavour to answer as many of your questions as I can!