Join the new world

Hey eIreland,

This article contains the results of 7 days of data collection from the Government's Free Tank Scheme. The aim of the study was to demonstrate the possible benefits of data collection in this scheme to time management and supplier recruitment , with the hope of influencing future e-government policy.

* Want to just skip too the main point of the article? Scroll down to the Summary and go past the "Results" to you hit "Reflections from the Author".

Reasons for the study:
Popularity and success of the scheme has led to large number of requests per day. In a few noted cases there had been long (5-7 hours) delays between request and supply, hence investigating the patterns in peak demand should aid the MoD in better time management of their suppliers.

The ability to pinpoint time periods where supplier staff are needed would aid the MoD in specific recruitment.

Promote the ethos of accountability.



Summary:
There is only a small difference in the average number of requests during the week and the weekend, however there is significant variation when peak demands occur over the studied period. The variation shown in the results indicate that time management based on TEDEN (8 GMT+0 onwards) and ONE primetimes is only a partly efficient system. There is always a peak demand between 20:00:00 - 21:00:00 (GMT + 0) or 13:00:00 - 14:00:00 (eRep Time) during the week, which time management based on TERRA primetimes accommodates, however there are peaks throughout the day which total up to more requests than alliance primetimes alone; this represents an area which can be improved to reduce waiting time of supplies, but also increase availability of weapons to non-primetime players. Therefore, future Irish governments should be strongly advise to collect and store supply data to better predict peak demands and assign sufficient staff.



Source of Error:
7 days is not long enough to provide averages and reliable predictions of how peaks will behave throughout a week in the future. While this study was only meant as a preview, currently supply data is deleted after 3 or 4 days by the MoD hence the study only had a small pool of data to utilize.

My method of data collection was to divide a day into 1 hour intervals and count the number requests within that hour slot. The format of the raw data meant I had to count the number of requests per interval, and without the use of a excel formula; hence there may be mistakes in the counting up.


Results:

Data was first recorded on the 9th (Wednesday) and finished on the 15th (Tuesday), however this constitutes a week and I've arranged the graphs to show a normal week. Each day has two graphs; GMT + 0 and eRep Time. The comments section focus on three (draft) advised main time periods Morning 02:00:00 - 09:00:00 (GMT+0) / 19:00:00 - 02:00:00 (eRep Time), DayTime 09:00:00-20:00:00 (GMT + 0) / 02:00:00 - 13:00:00 (eRep Time) and PrimeTime 20:00:00 - 02:00:00 (GMT+0) / 13:00:00 - 19:00:00 (eRep Time).

1) Monday


Comments: Highest peak was during PrimeTime with 7 requests, and an average of 2.33 requests per hour. DayTime saw three peaks with 4 requests each and an average 2.36 requests per hour. Morning had peaks no higher than 1 request, and an average of 0.43 requests per hour.

2) Tuesday


Comments: Both PrimeTime and DayTime had the highest peak with 5 requests. PrimeTime had an average of 2.17 requests per hour. DayTime had an average of 2.64 requests per hour. Morning had peaks no higher than 2 requests, and an average of 1.0 request per hour.

3) Wednesday


Comments: Highest peak during PrimeTime with 8 requests, and an average of 3.17 requests per hour. DayTime has relatively constant requests, multiple peaks no higher than 4 requests, and an average of 2.64 requests per hour. Morning has peaks no higher than 2 requests and an average of 0.43 requests per hour.

4) Thursday



Comments: Highest peak was during PrimeTime with 8 requests, and an average of 2.83 requests per hour. DayTime's highest peak was 5 requests, and an average of 1.9 requests per hour. Morning's highest peak was 3 requests and an average of 1.0 request per hour.

5) Friday



Comments: Highest peak was during DayTime with 7 requests, and an average of 2.36 requests per hour. Primetime has two smaller peaks with 5 and 4 requests, and an average of 2.17 requests per hour. Morning has no data point higher than 1 request, and an average of 0.57 requests per hour.

6) Saturday



Comments: Highest peak was during PrimeTime with 5 requests, and an average of 1.67 requests per hour. DayTime had two distinctive peaks with 4 and3 requests, and an average of 2.0 requests per hour. Morning time had a distinctive peak with 3 requests, and an average of 0.57 requests per hour. Lowest number requests on this day.

7) Sunday



Comments: Highest peak was 6 requests in both Morning and PrimeTime. PrimeTime had an average of 2.17 requests per hour. Morning had an average of 1.43 requests per hour. DayTime had two peaks at 4 requests and an average of 2.36 requests per hour.

Further comments: Tuesday being an odd ball, the week days show little variation in their peak patterns. However, Saturday and Sunday both have large peak demands in all three main time periods.





I've shown you can identify peak demands pretty easily from just a week worth of data. At max it took me around 10mins each day to collect and sort, and 30mins to put the results together at the end of the week. This is not a backbreaking hassle for any government; it just adds a few more clicks to the day but can add a wealth of knowledge. Collection and analysis of supply data is a realistic target for any Irish government.

Of course, 1 week worth of data is too small a sample to provide reliable results. If this was adopted, incoming governments would be tasked with collecting and analysing the data (Like I have), and then pass on the results to the next government so they may draw up their own supplier schedule. 1 month (term) of data would be sufficient to base a supplier schedule on; Having collection longer than 1 month would mean two governments are involved which creates hassle and any data older than 1 month runs the risks that the results may not be relevant to current circumstances. If suppliers recorded when they logged in to fulfil requests, that'd help too.

But you're thinking: Is this worth the extra hassle? Are the benefits that great?

If the best outcome to a shorter delay between request and supply was Ooh isn't that nice then I wouldn't have bothered!

Yes, this is worth the extra hassle. I am not advising anything new; Accessing when a supplier needs to be online has been common practice of any decent military unit. Another common practice of military units is trying to get their troops online at alliance primetime; some units have held supplies until primetime, some only supply if you are online at primetime etc etc. Does it work? Hell No.

While there is benefit of all troops in military units fighting in primetime, it's impossible to achieve on a daily basis. Afterall, this is a game and players will play when the time suits them. If a player in a MU keeps missing supplies due to not logging in at primetime, they'll find another unit. And thats cool, but I'm talking about a government scheme open to everyone. Under our program, a player has to be on the IRC and filled out the form to get supplies; I have noted delays of up to 6 hours before now, I don't think MoD expects anyone to stay on that long for the free tanks so they'll get their supplies nearer to primetime, when most people/suppliers start coming on for long periods.

^ If fighting at primetime = goodie was the only grade mark, that would also be cool but most Irish players do not produce enough food to get their full fights for the day. The free weapons would boast the max amount of damage they can inflict in one day. If a player is requesting outside primetime, they are going to be fighting barehanded when they are logged in (thats if they don't wait, and if they wait; thats usable health and damage down the drain). And if they do log in again at primetime, a good coordinating MU will be supplying around then and the free weapons will be less valuable than if they had got them outside primetime. My point being our time management is centred on the old style Alliance primetime markers, and it excludes casual players when it doesn't need too.

Thats all this article is highlighting: Collect data and use it to place suppliers when they are needed. Currently, the deletion of supply data after 4-5 days is throwing away a resource but its also not accountable for many things.


There are many ways to use the data:
See if there has been an increase in the number of requests over a term. Is that government doing more to get people on the IRC?
Is it the same people requesting supplies, or has there been an increase in the number of new players requesting?
How much per tank has the government spent? Did the previous government do it cheaper?
Etc, etc.


This is no way a criticism of a successful program; rather I am suggesting ways to improve it. Suppliers recording when they log in would be great, but all it would really take is a Minister of Defence to separate the fulfilled data from the suppliers requests, and hand it over to a logistics team or do it themselves on another excel spreadsheet.

This is also a wonderful opportunity for lazy good-for-nothing Dail shmucks to start opening up actual useful topics on government performance, and put some life into their dead forum.

Is mise,
Seanan