What's in a Name Gender: UK Edition
/Some time ago, I looked at the performance of Australian contestants based on the gender implied by their first names, but it turns our that I never got around to doing the same for UK contestants.
Today, we close that gap.
Firstly, a quick reminder that we are again using Derek Howard’s Gender by Name database to attach implied genders to names, and we are using the player-by-player data on OneQuestionShootout for our raw data.
Derek’s database attaches a probability of being male to a large number of first names, and we deem a contestant name to have been appropriately name-gendered if that probability is under 25% or over 75%. For the UK data that leaves only about 6% of the 9,260 contestants with an unknown name gender.
NAME GENDER PERFORMANCE BY SEAT NUMBER
We know from previous analysis that the producers of The Chase maybe place contestants of different abilities in different seats on the quiz, so let’s start by looking at some of the raw performance statistics by name gender and seat number.
Here we see that:
Name Gendered females in any given seat record, on average, slightly smaller Cash Builders than males. Note that this - and any other - name gender based analysis cannot be conclusive about the relative abilities of contestants with male-gendered names versus those with female-gendered names because we do not if the questions they face are of equal difficulty.
Contestants in Seat 1 tend to record the highest Cash Builders
Contestants with male-gendered names tend to choose higher offers than contestants with female-gendered names
Contestants with male-gendered names in Seats 1 and 2 get through to the Final Chase at higher rates than do contestants with female-gendered names. They also are a part of teams that win in Final chase more often than contestants with female-gendered names.
Contestants with male-gendered names in Seats 1, 2, and 3, on average win more money than contestants with female-gendered names in those same Seats.
INCLUDING THE OFFER TAKEN
We can drill down on the previous analysis by looking at which of the three offers each contestant has taken.
Here we can see that:
Contestants with male-gendered names from every seat tend to take the higher offer far more often than contestants with female-gendered names from the same seat number.
Generally speaking, contestants with male-gendered names tend to accept larger offers - whether they be low, middle, or high - than do contestants with female-gendered names from the same Seat number and who took the same ranked offer.
Contestants with male-gendered names from every seat tend to make the Final Chase more often than contestants with female-gendered names from the same seat number and who took the same ranked offer.
TEAM MAKEUP
It’s one thing to look in isolation at individual performance based on name-gender and Seat number, but we can also look at team performance based on the apparent gender mix that it has,
Here we can see that (ignoring mixes that have appeared in 30 or fewer episodes):
The Chase producers very much prefer a team that comprises two name-gendered females and two name-gendered males. That configuration accounts for over 80% of all episodes that can be confidently fully name-gendered.
Completely imbalanced teams comprising four name-gendered males or name-gendered females are almost unheard of in the UK version of the game.
It would be nice to see some experimentation by the producers of The Chase to see more name-gender imbalanced teams with 0 or 4 name-gendered females
Teams with just one female-gendered name have fared best in terms of win rates, setting a target, and in terms of the amounts won by participating contestants.
We can drilldown on this analysis based on exactly where the name-gendered males and females sat:
The first thing to note is that, amongst those compositions that have occurred at least 30 times, the four compositions with the highest win rate all have someone with a male-gendered name in Seat 1. Three of those four also have someone with a male-gendered name in Seat 2.
Amongst those same compositions
the Male/Male/Female/Female composition has tended to see the highest number through to Final Chase
the Female/Male/Female/Male composition has produced the highest average final prize fund
the Male/Male/Female/Male composition has produced the highest average target by quite some distance
the Male/Female/Male/Male composition has produced the highest average team winnings
ARE THE DIFFERENCES STATISTICALLY SIGNIFICANT?
We can conduct a variety of statistical tests to determine if the gender of composition differences we see are statistically meaningful and here we’ll produce what are called 95% confidence intervals to see if they do or do not include the value zero, which would mean that we can’t be highly confident that a real difference exists and what we’ve seen cannot be attributed to random variation.
(For the technically curious, I created these confidence intervals via bootstrapping)
INDIVIDUAL PLAYER STATISTICS
CASH BUILDER AMOUNT (MALE-GENDERED LESS FEMALE-GENDERED NAMES)
Seat 1: £112 to £464
Seat 2: £392 to £709
Seat 3: £211 to £507
Seat 4: £165 to £486
All Seats: £374 to £537
Conclusion: Contestants with male-gendered names tend to produce higher Cash Builder totals than contestants with female-gendered names. The difference is a bit less than one half of one extra correct answer
AMOUNT CONTRIBUTED TO PRIZE FUND (MALE-GENDERED LESS FEMALE-GENDERED NAMES)
Seat 1: £422 to £1,466
Seat 2: £905 to £2,073
Seat 3: £924 to £2,128
Seat 4: -£85 to £2,693
All Seats: £946 to £1,796
Conclusion: Contestants with male-gendered names tend to contribute more to the final prize fund than do contestants with female-gendered names. The exception is contestants from Seat 4 where we are less certain that the difference we see might not be attributed to chance. The overall difference across all seats averages out to about £1,400
PERSONAL FINAL WINNINGS (MALE-GENDERED LESS FEMALE-GENDERED NAMES)
Seat 1: £45 to £790
Seat 2: -£146 to £495
Seat 3: -£333 to £166
Seat 4: -£425 to £471
All Seats: -£28 to £325
Conclusion: With the exception of contestants in Seat 1, who do seem to win more if they have male-gendered names rather than female-gendered names, name gender cannot confidently be asserted as being associated with higher or lower winnings
TEAM STATISTICS
TEAMS WITH THREE MALE-GENDERED NAMES LESS TEAMS WITH TWO MALE-GENDERED NAMES
Difference in Number Who Make Final Chase: -0.14 to +0.13
Difference in Probability of Winning: -1% to +12%
Difference in Average Amount Won by Team: -£1,336 to £1,979
Conclusion: Teams with three male-gendered names do not statistically outperform teams that are balanced (ie have two male- and two female-gendered names) to such a level that we can confidently assert that the differences we see are not due to random variation.
BEST NAMES
For those of you who’ve made it this far, let’s finish with a fun one.
Specifically, let’s find out which relative common first name has been associated with the highest average winnings. For this analysis the only condition we’ll impose is that the name must have appeared at least 30 times across all episodes.
We see that Dan is very much your man with an average winning of £3,100, which compares to the all-contestant average of just £1,286.
Helen and Linda are also quite interesting in that they’ve won one-third of the episodes they’ve been in. That’s almost twice the all-contestant average of 17%.
