Matter of Stats

View Original

The Chase UK: How Good is Darragh Ennis?

Darragh Ennis became the sixth UK Chaser when he appeared in an episode that aired on the 19th of November 2020 where he won chasing a target of 16 against a team of two finalists, offering five pushbacks and answering the last question with 7 seconds still on the clock.

That started a remarkable run of 14 consecutive wins, which was the best start of any UK Chaser to date.

(All data used in this analysis comes from the One Question Shootout website - your source for all Chase UK data. It includes data for 1,930 episodes aired up to 11th November 2022)

(Note that all of the charts and tables in this blog can be clicked on to access larger versions)

Surprisingly, his run was ended by a solo player who set a target of 18 in pursuit of a £75,000 prize pool. In that episode, Darragh offered seven pushbacks, of which four were taken, and fell three questions short of the target when time ran out.

In total, he answered 26 questions in the two minutes available, which is about average for a typical UK Chaser, though a little faster than his usual pace.

Since that episode, Darragh has won 23 of the next 30 episodes in which he’s appeared, giving him an overall 37 and 8 record, which is the best record of any UK Chaser across his or her first 45 aired episodes, eclipsing Anne’s and Paul’s 35 and 10 records.

NOT ALL EPISODES ARE CREATED EQUAL

Darragh’s win-loss record so far has been impressive, but looking solely at that metric fails to account for the quality of the contestants that he faced, and the consequent targets that he has been set and number of finalists he has encounted in Final Chase.

Looking firstly at how he has performed against different target sizes, we see that he is the only UK Chaser with a perfect record against Final Targets under 15, with the win percentages of the other Chasers ranging from Shaun’s roughly 93% to Mark’s and Jenny’s 99%.

We also see that he has the second-lowest win rate against Final Targets in the relatively common 15 to 18 range, ahead only of Shaun, and between 5- and 10-percentage points behind the other UK Chasers.

His record against relatively large Final Targets of 19 or more is impressive, although he has faced them in only 13% of his episodes, compared to 22% for Shaun, 28% for Anne, 24% for Jenny, 34% for Mark, and 29% for Paul. Not too much should be read into this statistic, however, as it is based on only 6 episodes, and so is highly subject to sample variation.

However, the fact that such a small percentage of contestant groups have managed to set Darragh a Final Target of 19 or more means it’s probably fair to say that the average quality of contestant he’s faced has been lower than the average faced by other UK Chasers.

It’s true, of course, that the Final Target is heavily influenced by the number of contestants who get back to the desk, which is itself inflluenced by the Chaser’s ability to catch them head-to-head, but I’d contend that the Final Target set is far more influenced by contestant quality than it is by Chaser ability.

One other interesting feature of Darragh’s data is the relative consistency of his speed of answering, irrespective of the size of the Final Target, which is in stark contrast to the other UK Chasers. It will be fascinating to see, assuming this continues, the extent to which it might lower his win percentage for larger targets.

As just noted, there’s a direct link between the number of players who get back to the desk and the size of the Final Target that is set.

Part of the reason that Darragh has so infrequently faced large Final Targets is the fact that he has so rarely faced four contestants in the Final Chase. It’s been the case in only 7% of his episodes, as compared to about 11% for Shaun and Anne, 12% for Jenny, and 15% for Mark and Paul.

In those episodes - and, to be clear, there have only been three - Darragh has offered considerably more pushbacks, and answered considerably less accurately and more slowly than other UK Chasers when facing four Finalists.

More generally, we again see here relatively little variability in Darragh’s answering speed as we move from final teams of one to final teams of four, which is a phenomenon that we do not see for any other UK Chaser.

It’s particularly noticeable for Paul, who moves from answering 12.7 questions per minute when facing only a single Finalist, to 13.8 questions per minute when facing a full desk. That’s almost a 9% increase.

To give these numbers some practical context, Paul’s 13.8 answers per minute to Darragh’s 12.4 means that Paul will, on average, answer three more questions than Darragh over a two-minute period. Given that Paul’s accuracy is, overall, quite similar to Darragh’s means that Paul is more likely to run down large Final Targets should those numbers continue.



CONTROLLING FOR DIFFERENCES

To finish, we’ll build a simple statistical model that will allow us to estimate the probability of a particular UK Chaser being successful when facing a given Final Target and number of Finalists.

The model we’ll build is called a binary logit, and the fitted version of it is summarised at right.

If you don’t have a statistical modelling background, just note the following things:

  • The numbers next to each Chaser name give an indication of how likely they are, in this model relative to Shaun, to beat a team after we adjust for the size of that team and the Final Target they’ve set

  • Shaun, because he is what’s called the model’s “reference category”, is assigned a value of zero. Larger values mean that a Chaser is more likely to win, and smaller values that a Chaser is less likely to win

  • On that basis, we can order the UK Chasers from least- to most-likely to win as follows (the numbers in brackets are the Chaser’s actual win rate, unadjusted for player quality):

    • Anne (80%)

    • Mark (76%)

    • Paul (78%)

    • Jenny (79%)

    • Darragh (82%)

    • Shaun (69%)

It’s important to point out that this is just a best guess at an ordering, based on the data that we have so far If you look at the column headed 95% CI in the earlier table, you can see that the true values of the numbers we’re using to rank the Chasers could be anywhere within quite a large range.

If you want to use this model to perform some of your own calculations, here’s one example to show you how it works.

A team of 3 Finalists has set Jenny a target of 19. What is our best estimate of her chances of prevailing?

Estimate = 1/(1+exp(-(7.7 + 0.85 - 0.40 x 19 - 0.15 x 3))) = 62%

The results of performing a range of similar calculations appears in the chart below.

The key message from this chart is that, generally speaking, the estimated differences in ability across the UK Chasers is quite small, especially for typically sized Final Targets and team sizes. For example, for a Final Target of 16 set by a two-player team, the estimated Chaser victory probabilities range from just 86% to 89% without Shaun, and 74% to 89% with him.

SUMMARY AND CONCLUSION

Darragh has undoubtedly had an impressive start to his Chaser career, and is a formidable adversary.

His 37 and 8 win-loss record is the best of any of the Chasers, although the analysis suggests that this record might, at least in part, have been bolstered by his facing a slightly lower than average contestant quality, as evidenced by the smaller number of them making it back to the desk, and the smaller Final Targets they have set, relative to teams faced by the other UK Chasers.

If we adjust for contestant quality, Darragh’s record looks somewhat more like that of the other UK Chasers, though the relatively small number of episodes he’s appeared in makes any estimate of his ability subject to more sampling variation than estimates for other UK Chasers.

What’s particularly interesting about his style so far - and what sets him apart from the other UK Chasers - is the consistency of the pace with which he answers questions in the Final Chase, regardless of the size of the Final Target. This pace is slower than the other Chasers’, and it will be interesting to see what, if any, effect this has on his future ability to run down larger Final Targets.