Should the Goldenblock start holding Bingo Sundays? Are Incumbents in Vogue? These Questions and More

Westinor

Deputy Minister
Citizen


Some Quick Investigations Using EEPED
By Westinor



I recently released the first version of the Europeian Executive and Presidential Elections Dataset (abbreviated to EEPED, which is on-point because I am so eepy writing this), which you can check out here! To demonstrate the dataset's utility in performing casual research on interesting Europeian electoral questions, I thought I'd try and answer a couple quick questions utilizing data stored in EEPED. Though it is not currently in a completed state, EEPED still contains a good amount of data on elections and candidates from the Executive Re-Merge until the present day. There are a ton of interesting questions that it can answer, though, so I'll take a stab at a few here!! These are extremely rudimentary analyses, but hopefully they'll prove interesting enough to some readers :>

Analysis of Age on Vote Percentage

An age-old question: do older, more experienced candidates win more elections? Europeia is filled with its fair share of old, wizened political sages who occasionally step into the electoral arena to show the youngsters how it used to be done. In the most recent presidential election, two of the region’s oldest members, HEM and Sopo, drew eyes with their late bid for the presidency. An EBC poll in reaction to their entrance saw what had been a previously fairly well-positioned run for incumbent Vice President Grea Kriopia transform into a neck-to-neck-to-neck race between all three tickets. Previous polling shows that experience is a perpetually salient issue among voters, outshining even an emphasis on young blood and talent at times. Despite regular debates on upward mobility and providing chances for new faces in the executive, it is clear that our older citizens remain political heavyweights by way of their hefty records in the region.

Drawing from data in the EEPED, we can first check substantively how much being very old matters in terms of winning elections. The dataset stores information on the cumulative ages of a ticket, from the day each candidate joined Europeia and applied for citizenship to the day of each election. A quick analysis shows that the oldest tickets in any given election do not actually win more often than other tickets – in fact, they only win just over a quarter of the time!



We can further divide up the results to see what places the oldest ticket in each election ended up in.


A quick note: replicating this will yield a "first place" count of 11, but there are only 8 "wins" for the oldest tickets. This is because of three elections: September 2021 and its runoff, and November 2021. The former saw Lloen and Lime go to a runoff and tie, which I coded as a first place for both of them but not a victory. The latter saw a runoff between Darc and Sky, which Darc won. Darc was the oldest ticket and received the most votes, but because there was a runoff he was not coded as "winning" the general election.

This is perhaps a better representation of how well old tickets tend to do – they don’t often win, but they perform pretty solidly, coming second in half of all elections. However, drawing any conclusions from this analysis would be misleading – many elections see only two tickets compete, with one of them perhaps only somewhat older than the other. Older tickets may simply fall prey to a variety of other headwinds that push their experience in the region out of consideration in voters’ minds. For the most part, tickets that are particularly older than their competitors perform no better than an average ticket, but it appears the oldies of the region are dragged down in this particular range of elections by a single outlier – HEM, who has not won a single of the three elections he has run in since the executive re-merge, while absolutely exploding a ticket’s age in the process.

There are better ways to visualize how larger and smaller differences in a ticket’s overall age might affect a ticket’s chances of winning a larger share of the vote. Leveraging our data on the absolute age of a ticket, we can visualize how well older tickets do depending on their actual age in terms of vote percentage.



Surprisingly enough, there’s a negative relationship between the two. It doesn’t seem particularly significant, though, and it’s pretty clear there’s an outlier that’s dragging the results down – that’s the aforementioned HEM/Sopo ticket from before, which is actually (by a large margin) the oldest ticket that has ever run in Europeian history. By removing that particular datapoint, we arrive at a bit different but not necessarily clearer picture:



The relationship isn’t so stark anymore, and in fact, it would probably be pretty bold to claim that there was one at all. Running regressions on ticket age and vote percentage, electoral placement and electoral victory yields statistically insignificant results. Disappointingly enough, it seems like ticket age perhaps really does not matter whatsoever.

However, this doesn’t necessarily rule out experience as a relevant factor in a ticket’s electoral chances or as a useful predictor in predicting the winner of an election. Ticket age is probably not the best representation of experience – the relationship between time spent in a region and how much one learns is not necessarily linear, so perhaps a more nuanced or wholly different approach to representing experience may fare better. EEPED also contains a variable indicating the number of terms served as the Head of Government per ticket, which may be a better indicator of relevant experience in the executive. How do tickets that have served more terms as head of the executive fare?



Pretty well! It’s not an incredibly tight relationship, but this points us perhaps in a different direction. We should expect that having served in the job one is running for previously might help with convincing voters that you’re fit to hold it once more. Further regressing whether or not a ticket won an election as well as the percentage of votes won on both ticket age and previous terms served as the head of the executive yields a statistically significant relationship between both ticket age and previous terms as head of government (results can be seen in the table below). Ticket age is confirmed to have practically no effect on winning an election at all, but previous terms served as head of the executive seems to have a fairly positive relationship with electoral success – the results are not necessarily anything to write home about, but they confirm what is perhaps generally accepted knowledge in Europeian politics with an important qualifier.

Old tickets are not guaranteed to win races – but experience almost definitely matters. This incredibly surface-level analysis lends credence to the idea that a history in the executive strengthens candidates’ chances of winning, but being at the top isn’t the only way to earn your stripes. With more data on candidates’ experience in the executive, particularly via Cabinet terms and other roles, we might be able to create a more interesting and fine-grained analysis of how exactly experience in the executive shapes one’s fortunes in the political arena. For now, though, even limited data on a constrained group of variables can help us ascertain the answer to some perennial questions.

Does Executive Satisfaction Polling Predict Vote Percentages?

Standing at the front for consideration in many Presidential Elections are the loyal vanguard of the previous administration: the Cabinet Ministers. These integral parts of any Executive government are often considered strong successors for the role of head of government due to their proven successes in the past term. How proven their success is, however, varies. Luckily, Europeian media is incredibly vibrant and often collects citizen satisfaction on Cabinet Minister performance, a strong metric for how well a Cabinet Minister is considered to have fared in their term.

I collected this data and stored it in EEPED. Since many presidential and vice presidential candidates are returning cabinet ministers, quite a few tickets have mid-term and end-of-term satisfaction polls associated with them. However, this is not the case for all tickets, or even all returning ministers, as I was unable to find satisfaction polls for some terms. For the tickets that did have a satisfaction rating for the candidates that were returning cabinet members, though, I sought to perform a very basic examination: did satisfaction ratings help predict a ticket's performance?

To dive deeper into this, I found the average of each ticket's satisfaction ratings (if there was a satisfaction rating for one candidate on a ticket but not their partner, I took the average of the existing data and replaced the null data). This was the best way I could think of to represent a ticket's satisfaction ratings, though it might be potentially problematic to assume that the existing average is a good placeholder for data that wasn't collected. The imputation didn't affect too many tickets, though, so the results should remain fairly unchanged.

Here is a basic visual representation of the data:

It's not clear that performance in satisfaction polling correlates with electoral success, nor that the performance in the previous term's satisfaction polling is a good indicator of how much of the vote a ticket might receive. There is a lot of missing data in this section, so the ambiguity of our result is not surprising. With more data, and perhaps a stronger commitment to holding satisfaction polls for each term, we might be able to more strongly ascertain the existence (or lack thereof) of a relationship here in the future!

Incumbent Success in Elections

I would like to quickly go over one final interesting conclusion from the data I gathered this time around - how well do incumbents perform in elections? Europeia has a tattered history of struggling incumbents, with recent examples including Rand's unsuccessful run in the wake of an interregional crisis in the summer of 2023, and JayDee's defeat at the hands of ROE in January 2024. At the same time, however, we have just witnessed a year of history-making continuous presidencies, with JayDee's three terms and upc's record-breaking four having stretched almost the entirety of 2024. How successful are incumbents at winning elections? Are they burdened by the weight of their previous failures, or do they stand high atop the accomplishments they have worked so hard to achieve?



The quick answer is a resounding shrug. Incumbent presidents are currently tied in terms of electoral victories, and incumbent vice presidents hold a two-election margin. Something to note is that upc and JayDee's terms counted for 6 total wins for incumbents (and one technical loss) - perhaps incumbent presidents of Europeia past ought to hang their heads in shame.




That's all I have for now!! I apologize for the poor writing, it is very late for me and I am exhausted, but I hope the data is at least mildly interesting enough to make up for it :) Thank you for reading, please feel free to check out my dataset if you have the time and have a gorgeous weekend!
 
JayDee's terms counted for 6 total wins for incumbents (and one technical loss)
I refuse to acknowledge this as a real loss :P

This is really fun analysis though, thank you West!
 
Last edited:
I'm very impressed with the amount of data collected, but disappointed there were no tight correlations uncovered
 
For the most part, tickets that are particularly older than their competitors perform no better than an average ticket, but it appears the oldies of the region are dragged down in this particular range of elections by a single outlier – HEM, who has not won a single of the three elections he has run in since the executive re-merge, while absolutely exploding a ticket’s age in the process
This is the single most violent thing anyone has ever said about me!!!

To be fair, I don't even remember running three campaigns since 2020!!!

Overall, this is fascinating stuff. Thank you for writing this all up.
 
Does the data look very different if you exclude HEM’s recent runs?
 
This is the single most violent thing anyone has ever said about me!!!
if only you had run with me that one time...

this is marvelous work west, i have really enjoyed reading it!
 
Does the data look very different if you exclude HEM’s recent runs?
I believe the fourth figure is just that! Not a ton different is the answer though

I'm very impressed with the amount of data collected, but disappointed there were no tight correlations uncovered
It was disappointing to me too :( I’m optimistic for a few reasons, though! Firstly that the breadth of data with regards to something like age here is actually relatively small - though there are tons of observations, they’re restricted to a particular time period of Europeian history where I figure there’s more of a sort of bias against older candidates. The older end of the more politically-active candidates observed (HEM, Sopo, Kaz, McEntire etc.) were, at this point, either past their electoral “prime” - I’m laying in bed without the spreadsheet on hand at the moment but the former two I believe served something like 11-12 terms as head of government between them before the executive reunification. They both really went on to play supporting roles in a couple tickets during this
time period, usually as the VP. Kaz and McEntire both were also repeat-candidates who ran multiple tickets; Kaz successfully won one term but lost in re-election while McEntire did not win any at all. These candidates (and a couple others, I believe perhaps CSP and Calvin) had disproportionately large ages due to this being the most recent data available - a lot of them come from the same age of Europeia in terms of political activity, and the years probably stop mattering after a bit (if you’re ever considering age or age as a proxy for experience an extra five years of HEM being HEM isn’t going to change a voter’s mind, but it will change what the data says) while having significant disadvantages in terms of “electability”, usually running in last-minute tickets or playing supporting roles to perfectly viable candidates and drawing the short straw. What would’ve been a more reasonable way to go about this instead of jumping straight in without any assumptions a priori would’ve been to sort out the candidates by age range and check their electability - I think you’d probably find this age-associated relationship with the far end of older candidates and the very new end of candidates being on the losing end. That’s probably something someone could do in a couple minutes and with the benefit of being able to stew on the data instead of typing it all up in a frenzy I can look back and see there was a much better way to go about this! Additionally, just stretching out the dataset probably gives us better insight into a stronger relationship for something that’s more of an age-old issue, literally. The portion on satisfaction data was a bit poorly done, with more comprehensive data on that I would take a different look at it in the future, but it’s mostly ragtag because you’re not going to have polling data for a lot of candidates who didn’t serve in the most previous cabinet - but you could supplement their most previous cabinet term and probably fill in most null datapoints but that would come with a potential cost in terms of what the data is actually representing. Finally with regards to the incumbency results… I have no clue, more data would probably be good there too!! Notably (I should have made a graphic for this) though, Incumbents minus JayDee and upc in particular fare incredibly poorly, which is something to chew on. Elections in which incumbents ran were relatively few but surprisingly chaotic and spicy - while collecting data these were the ones that stood out to me, and I can name quite a few off the top of my head (yours, Kaz’s, Darc/upc’s, Pichto/Kura’s). Honestly if I’m counting right no incumbent has won aside from Jaydee, upc, and Gleg? That probably would’ve been a cooler headline!!!

note to self: collect cooler data next time
 
note to self: collect cooler data next time
Oh no, it's very cool. I think it speaks to the electoral open-mindedness that incumbents and experienced citizens don't have a huge advantage.
 
Honestly if I’m counting right no incumbent has won aside from Jaydee, upc, and Gleg? That probably would’ve been a cooler headline!!!
You are correct. I even did an article covering this anomaly.

 
Honestly if I’m counting right no incumbent has won aside from Jaydee, upc, and Gleg? That probably would’ve been a cooler headline!!!
You are correct. I even did an article covering this anomaly.

I can’t believe I’ve forgotten this!
 
I’d be interested to see visualization of the age of tickets the oldest ticket is running against
Took a bit to properly think it up, voila


That took longer than expected ;-; I did not know how to do what I was trying to do

(note: I am realizing now I forgot to drop non-contested runs from the original analysis so the actual number of oldest ticket wins is five...)
 
Last edited:
I’d be interested to see visualization of the age of tickets the oldest ticket is running against
Took a bit to properly think it up, voila


That took longer than expected ;-; I did not know how to do what I was trying to do

(note: I am realizing now I forgot to drop non-contested runs from the original analysis so the actual number of oldest ticket wins is five...)
I’m not entirely sure if average age of ticket excluding the oldest is the right way to look at it because that data can be greatly skewed by a much younger ticket.
 
I’d be interested to see visualization of the age of tickets the oldest ticket is running against
Took a bit to properly think it up, voila


That took longer than expected ;-; I did not know how to do what I was trying to do

(note: I am realizing now I forgot to drop non-contested runs from the original analysis so the actual number of oldest ticket wins is five...)
I’m not entirely sure if average age of ticket excluding the oldest is the right way to look at it because that data can be greatly skewed by a much younger ticket.
I considered this, but a. there are only I believe three elections in which this is the case and b. I couldn't really think of a better way to do it! The data gets really condensed if you set a bar for each ticket (there are I believe over eighty observations and a majority of them would be in the top row!) but also the "average age" excludes the oldest ticket as well, so it should be a fairly holistic and probably the best estimator for most cases (tell me if you can spot the elections dragged down by a very new ticket!! two of them are right next to each other)
 
How are we determining age? Is it when someone joins Euro or when they join NationStates.

Also I’m pretty sure I beat Elio in January :P
 
How are we determining age? Is it when someone joins Euro or when they join NationStates.

Also I’m pretty sure I beat Elio in January :P
Euro!! and i’m not sure why that is happening, i thought i had fixed that but admittedly was so tired after the hour of editing code that i must have missed it ;-; i think it is probably cos you were coded as losing the initial election, i will doublecheck that thank you for the callout!!
 
Back
Top