Joe-Zeloof/Joey_Zeloof_OpenML_SpeedDating

Overview:

After looking through different datasets i found one that includes data collected from a speeddating experiment that took place from 2002-2004, in which data was collected about the dates itselves, the outcome of the dates and background information of the participants. My hope was that through the analysis of the data i would be able to improve my chances of sucsess in the dating world.

Data handling summery:

I went over the metadata of the dataset and didnt find any major problems with the dataset, there were no empty fields and all data looked rational and uncorrupted.there were also no outliars. the only issue i found with the data was that there were effectivly 6 different features which had different scales for values then each other. after diiging deep into the problem i realized that there is no real way to unify the scale without distorting the data (even a minmax normalization wouldnt work). So i left it as it was and made sure during my analysis not to compare features with different scales.

Research questions and analysis

1. What is the difference between how beople rank their own attractivness to how they are percieved

• chart 1:

• several main insight sarise from the plot:

  • assuming that true attractivness is how you are rated by others then, people rank their own attractivness around 7% more then rality.
  • this begs the question of wheather there is a difference in how men and women rate themselves in relation to how they are rated.
  • the average value for attrcativness by others is 6.21 and not 5 as would be expected from an average of a scale from 1-10.

2. is there a difference in how men and women rate themselves in relation to how they are rated, if so whats the difference.

• chart 2:

• Insights:

  • surprisigly or maybe not on average, men on average rank themselve more then they are then do women.
  • men rank themselves around 10% higher and women rank themselves 4% higher
  • i find this facinating! after some thought i guess it makes sense because for my own personal expirence women tend to be harsher on themslves then men when it comes to beuaty.

3. out off all the interests in the experiment, which one has the highest correlation with wheather or not the dated wants to date

• chart 3:

• insights:

  • also facinating but very quickly i noticed that in this plot aswell there need to be a seperation between men and women as my gut tells me that the correlation between if a man is intrested in tvsports and his chances of getting a second date is as higher then the same correlation for women.

4. difference in corraletion of dater interests and chances of getting a second date between men and women

• chart 4:

  

• insights:

  • very few of the intersts in women are positvly correlated whereas in men there are many more.
  • cluubing ranks highest in both men and women in correltion which can be explained by the fact that people who are interested in clubbing would likely be more outgoing and therefore would likely have a higher chance of getting a second date
  • there are several interests that are positivly correlated in men but negativly so in women such as: concerts and dining. im not quite sure why that is tbh but intersting nevertheless.

5. how the age of a dater affects his attractivnes according to the person he is having a date with

• chart 5:

• insights:

  • women aged 18-20 are percived the most attractive and then it platoes till the 30s
  • men remain relativly stable until they depreciate after 26
  • important things to remember are thet all the subjects are in university and that may skew the data as they are probably more inclined to date people within university (of a younger age) which probably positivly impacts the attractivness of younger people.

6. Do men expect more dates than women?

• chart 6:

  

• insights:

  • not surprisingly, men expect more like but in reality recive less (and obviously the opposite is true for women). this finding goes hand in hand with how men and women value themselves. from both queries it seems that women are harsher on hemselves when in reality they have a bigger selection of men to choose from as is shown in the second plot above.

Research summery:

Across the analysis, several consistent themes emerged.

1. First, people tend to rate their own attractiveness higher than how they are rated by others—on average by about 7%, with men overestimating themselves more (≈10%) than women do (≈4%).
2. Interestingly, the average attractiveness given by others was 6.21, higher than the midpoint of the 1–10 scale, suggesting participants were generally generous in their evaluations.
3. Interests showed only weak correlations with the likelihood of receiving a second date, but important gender differences appeared.
4. Men showed many more positive correlations across interests, while women showed almost none, meaning that hobbies play a larger role in men's dating outcomes.
5. Clubbing was the strongest positive correlate for both genders, possibly reflecting confidence and extroversion.
6. Some interests even flipped direction between genders—for example, concerts and dining were positive signals for men but negative for women, a surprising and difficult-to-explain pattern.
7. Age also played a role: women aged 18–20 were rated the most attractive, after which their perceived attractiveness remained fairly stable until the early 30s.
8. Men’s attractiveness stayed stable until around age 26, then dropped more sharply. These patterns may be influenced by the fact that all participants were university students, making younger partners more accessible and socially desirable.
9. Finally, expectations versus reality showed a clear divide: men expected more dates but received fewer, while women expected fewer and received more. This aligns with the earlier findings—men tend to overestimate their desirability, while women undervalue themselves despite ultimately being selected more often.

Taken together, the analysis paints a consistent picture: people—especially men—overestimate their attractiveness and dating prospects, interests play a modest but gender-dependent role in dating success, and age preferences reflect both biological and social factors. These insights highlight the complex but predictable dynamics of early-stage attraction.

Link to explenation video

https://youtu.be/XzHOtQp-EX8