On 10th August 2020, the Board of Control for Cricket in India (BCCI) received approval from the Indian government to hold the 13th edition of the Indian Premier League (IPL) cricket tournament in the United Arab Emirates. Due to the rising number of coronavirus cases, it was deemed too risky to hold the lucrative sporting event in India.
Why did BCCI consider UAE as the suitable host for IPL in 2020? Weren't there suitable locations within India which could have been viable?
Since this is a problem statement involving 'location', let us see what map-based Location Analytics can indicate to us. Initially, I'd request you to spare a minute to think of potential venues within India or beyond which could've been a suitable alternative to the UAE. Compare your thought-process with the output mentioned towards the end of this article.
At the outset, I think both of us would concur that shifting the IPL location to a country outside Asia - be it South Africa (like in 2009) or an Australia or England wouldn't have been viable because of the high expenditure involved, quarantine restrictions, logistical challenges, and so on. Since there wouldn't have been any in-stadium viewership (and no ticketing revenues as a result), scheduling this cricketing extravaganza in a distant land during the coronavirus crisis was always a less practical option.
Within Asia, I couldn't think of any reason why India should have forfeited the opportunity to be the host. After all, we have so many cricket stadiums around the country. Surely, we could find a couple of viable venues in less virus-affected areas and hold the matches there.
Well, as soon as I saw the data, I knew I was under the wrong impression - we have just 23 active international cricketing venues in India - operational, with floodlights, International Cricket Council (ICC)-approved i.e. with adequate sporting & residential amenities, and so on.
Nonetheless, 23 is still a large number - in fact, the largest in the world for any country - surely, we could have found suitable venues within India from these options, I thought to myself.
Note: The dynamic map-based application for this study can be accessed from- https://arcg.is/0jD519. You can navigate the application as you read this article as there are many facets which would appeal to you more meaningfully once you explore the data for yourself - the images included below may not be fully self-explanatory. You may be asked to create a free account on the application landing page.
The next question was What parameters will I choose for selecting venues within India? I felt it should be equivalent or better than what UAE is offering to us.
The UAE offers us 3 cricket stadiums which are in close proximity to each other. I assume that this is the primary criteria that BCCI considered - it would be easier to schedule games and logistics for the eight teams (up to 4 teams play everyday) from these three closely located venues. Modern training facilities and the availability of suitable (bubble) accommodation at each of these locations are an added advantage.
Modern Location Analytics platforms (GIS) are quite powerful. Rather than plotting the straight line distance between locations, here we use Drive Time as a measure to calculate coverage in a particular timeframe from a given location (Drive time tool can be configured with the existing Road network layer, Average driving speed of vehicle, Time of the day, live Traffic information, and so on).
In the output above, you are observing the coverage from Sharjah stadium if one is driving in a car for 2 hours, in any direction. You will infer that all the three stadiums fall within a 2-hour travel window (no traffic). Besides, road connectivity between these locations is excellent (in-situ inference).
Certainly, these venues in the UAE appear to be an attractive proposition - less time involved in travel, less touch points and hence, lesser risk of getting infected (direct source to destination travel on ground - no flights), restrictive but luxury accommodation for the cricketers - all these combine to create a strong 'bio-secure bubble' that the organizers are aspiring to create for the contingent.
Next, going purely by the 'time taken to travel' parameter, do we have anything viable in India?
Upon loading the stadium data onto a map as depicted above, it dawned upon me that most of the locations in India (International cricketing venues i.e.) are geographically dispersed. Upon doing a similar Drive-time analysis as I had done for the UAE earlier, the output generated is depicted below-
The finding was not very encouraging - No three stadiums fall within the 2 hour drive-time window.
At best, there are three pairings of 2 locations which could be considered- (the obvious one being Mumbai's Wankhede and Brabourne; Delhi & Greater Noida and Lucknow & Kanpur being the other two). The closest three location pairing would be Mumbai's Wankhede & Brabourne + the MCA Stadium in Pune.
(Apparently, the DY Patil Stadium in Navi Mumbai has been converted into a football stadium and therefore doesn't feature in the list of active international cricketing venues. If it was still an international cricket venue, then Mumbai (2) + Navi Mumbai (1) would have been the ideal 3-location pairing, purely from the proximity perspective).
Let us ignore the virus cases for the time being and continue to deal with this problem statement purely from a location standpoint. If India is not suitable, can other countries in Asia be suitable? Within Asia, of course Pakistan is not an option - so Bangladesh & Sri Lanka are the only two major cricketing nations left for our consideration. Because both are smaller countries by size, the likelihood of finding 3 locations close to each other could possibly increase.
Bangladesh has only 4 international cricket stadiums (with floodlights). Unfortunately, no three locations fall within the two-hour window, here as well. Only Dhaka and Fatullah (both being part of the Dhaka administrative division) can be paired together. Chittagong and Sylhet are actually between 4-6 hours Drive time away from Dhaka. Also, by virtue of being one of the world's most densely populated countries affected by the virus, I can understand why BCCI would have been reluctant to select Bangladesh even if the location considerations were to meet the proximity requirement.
We are left with Sri Lanka. Let us see what the island nation to the south of India has to offer:
Close! Although not exactly within 2 hours, the trio of Colombo, Pallekelle and Dambulla (with Colombo and Dambulla at the opposing ends) seems to be an appealing alternative to UAE (Sri Lanka, like Bangladesh, has only 4 operational international cricketing venues with floodlights at the moment).
With location analytics platforms such as these, one can avail the benefit of clubbing & analyzing non-spatial data as well provided that we are able to 'link' the positional data to the non-spatial attribute. So I proceeded to gather the total coronavirus cases (cumulative) as of today - 20th September 2020 i.e. - at all the cricketing venues and linked it to the positional data of the stadium location. Since I couldn't obtain city-wise virus case data for all the locations, I decided to use the Administrative region-wise Virus cases dataset and divide it by the population of that region to derive a (normalized) percentage for comparison. For India, I believed I could use the administrative region - a State, for UAE - an Emirate, for Sri Lanka - a Province and for Bangladesh - a Division.
Unfortunately, I wasn't able to access the virus data, administrative region-wise for UAE and Sri Lanka. As a result, I obtained the total cases for the country and multiplied it with the percentage of population that lives in that particular administrative region. For example, if UAE has 200 coronavirus cases and Sharjah Emirate within has 20% of total UAE population, then I have assumed that Sharjah has 40 cases of virus (20% multiplied by 200). While this method is flawed, for this study purposes it appeared to be the next best alternative to me. Fortunately, the total number of cases for these countries were low and therefore, the chances of high variance / misrepresentation are reduced considerably.
Once I integrated the virus information with the stadium positional information and plotted it on the map, the output took me by surprise (refer figure above).
While the total cases in UAE is relatively less (85,000 cases), when compared to its population, the percentage of cases to population is actually high - between 0.74% and 0.99% i.e. nearly 1 affected in 100 - it falls in the second-worst class range among the Asian venues used in our study. The whole of east India and Bangladesh fares better than UAE, which was bewildering for me although one can contest these stats due to the low testing rate in India, under-reporting of data etc.
Unfortunately, the venues most suitable in terms of location in India - the Mumbai-Pune belt fares worse than UAE in terms of the total virus cases parameter.
The trio of Dehradun, Dharamshala and Mohali looks an appealing alternative to me. However, I also understand that the Himalayan foothills would be a difficult terrain to navigate, especially at night once the matches conclude, with the winter season just around the corner. The 3 stadiums in Uttar Pradesh are even better placed in terms of proximity, but not the best location pairing to host the IPL due to the high population density in the state.
Hence, I concluded that, from a location proximity and from a virus cases perspective, there doesn't seem to be any attractive venues within either India or Bangladesh.
Which leaves us with the island nation Sri Lanka. With total (cumulative cases) in the entire country at 3287 as of today, active cases at just 186, case to population % is the lowest for all the Asian locations at 0.02%. Additionally, a pristine environment, availability of suitable accommodation by virtue of it being a popular tourism destination, pleasant weather during this time of the year (possibly rainy, however), its proximity from Indian mainland, and a INR / LKR rate of 1 / 2.52 (expenses play a vital role) - the trio of Dambulla, Pallekelle and Colombo seems equally viable as the UAE venues, if not better.
Interesting Fact: Sri Lanka actually even offered to host the IPL way back in April. So what could have gone against it?
a) Charges of Corruption against the Sri Lankan Cricket Board and b) Security Threat - seems to me the two plausible reasons why the BCCI preferred to handover the organizing duties to the UAE.
That being said, there is still scope for plenty of refinement in the methodology used to shortlist the right location. We can add more parameters like active cases, projected rise in cases, location and quality of hotels available in the vicinity (ICC mandates a minimum of 4 star accommodation for international matches), location-wise expenses incurred, predicted weather information etc. to arrive at a better decision.
For this article purposes (purely academic & informative in nature), we'll leave it here, though.
One important learning which I personally derived was that there is a benefit in having sporting venues located in close proximity to each other. After all, it would help in organizing large competitions / global tournaments without all the long-distance travelling and accompanying hassles. No wonder, Olympic villages are built to facilitate the namesake tournament. While each state in India may build one or two cricket stadiums to satisfy its residents, logistically, it is a nightmare for athletes, telecast professionals and the entire ecosystem in terms of cost, time and comfort.
Update: 23rd Dec 2020 - Refer this article which highlights the benefits of having a 'compact' sporting tournament.
My idea was to take a problem statement which was appealing in nature and work towards a solution using location analytics. I hope you found reading this to be enjoyable and informative. Do check out the map based platform and study the output for yourself - https://arcg.is/0jD519. It is relatively simple to navigate and use, although some of the advanced filters and views may need some learning.
Have fun!
ABOUT US
Intelloc Mapping Services | Mapmyops.com is based in Kolkata, India and engages in providing Mapping solutions that can be integrated with Operations Planning, Design and Audit workflows. These include but are not limited to - Drone Services, Subsurface Mapping Services, Location Analytics & App Development, Supply Chain Services & Remote Sensing Services. The services can be rendered pan-India, some even globally, and will aid an organization to meet its stated objectives especially pertaining to Operational Excellence, Cost Reduction, Sustainability and Growth.
Broadly, our area of expertise can be split into two categories - Geographic Mapping and Operations Mapping. The Infographic below highlights our capabilities.
Our 'Mapping for Operations'-themed workflow demonstrations can be accessed from the firm's Website / YouTube Channel and an overview can be obtained from this flyer. Happy to address queries and respond to documented requirements. Custom Demonstration, Training & Trials are facilitated only on a paid-basis. Looking forward to being of service.
Regards,
Much Thanks to ESPN Cricinfo, John Hopkins Virus Dashboard, Wikipedia & Esri's Living Atlas for the datasets. Study has been done on Esri's ArcGIS Online Location Analytics Platform.