Sports Graphics Production: Sports Data Management

Those who work with data driven systems know that reliably acquiring data is only part of the battle - data is often not in the right format and needs to be delivered for use structured in the right way.

As broadcasters compete to produce the most compelling graphics for sports events, creative considerations can sometimes overshadow the task of sourcing and organizing the information to be displayed. Finding that data, connecting it to rendering systems, and providing an operator with useful controls means drawing together a lot of threads, and is more often a service than a single static product.

Data Sources

Leagues, teams and other sports organizations make various data available in various ways. Information might exist as SQL databases, XML-formatted information, or other online or cloud-based resources. Meanwhile, different rendering systems from different vendors might require different interfacing using different, perhaps proprietary protocols - and all of this might change significantly depending on the sport and the desires of the director.

Jeremy Torres is Head Of Software at RCS, a role which has kept him deeply involved with that sort of integration for over a decade. “Sport to sport, league to league, every league and team has a different availability of data,” Torres points out. “A lot of the time that doesn’t get discussed and it gets pushed down in priority. That’s really what most of my job is - fitting those square pegs into round holes. Some data is available for some sports, other stuff we have to conjure up out of bits and pieces.”

The results may end up filling in the blanks for a scoreboard or lower third, but the demand for graphics is ultimately limited only by the desires of the director. “A lot of it is reactive data, which must automatically populate graphics,” Torres elaborates.  “We’ve done plenty of automated tickers for sports networks - when there’s a goal scored in soccer it breaks into the ticker and says which game and match it is, and the updated score. It can be manually operated in terms of data entry and then integrated, as well as fully automated.”

Feedng Creative Teams

Coordinating that effort with the creative team is key, as Torres goes on. RCS centers its approach around “project producers. They’re the ones who take charge on the creative aspect of it from a production level, not from a technical perspective. Those are the people I typically have the conversations with about where to put this data and when. Most of our people depend heavily on the project producer to be the one who knows the sport well, who knows the show well.”

Torres recalls a project on which those factors were thrown into sharp relief “a couple years ago, when we had our first cricket project. I didn’t know anything about it at the time… thankfully one of my colleagues knew what a baseball fan I was and guided me through it. It was an experience: I didn’t know how that sport works, how a production for that sport works, or what the graphics operators expect in terms of what data they would be expecting to put on air.”

These situations inevitably involve several departments at several organizations and people at various levels of creative and technical involvement. Original software development is usually involved, and lead time, as ever, is key. “Once upon a time,” Torres reflects, “there used to be months to work on these kinds of projects.” Estimating the work required intersects with a lot of different factors, but Torres explains that his work can begin “once I have the answer I need in terms of where the data’s coming from, and what it is expected to look like.” Answering that question before a sports season starts often involves simulated data, approximating the formatting of the real thing for any given sport. 

Torres’s initial work, he says, is often based on this simulated data - but once broadcasts begin, change is inevitable.  “Typically [projects are completed] on a per-season basis, but we’re always expecting things to happen, especially in the first season. You can only simulate data so much, and ninety per cent of the time the simulated data looks nothing like the real data. When it goes on air, you see what you have versus what you wish you had. The first couple of weeks of a project being on air is when I’m at my most heightened alert to get things turned around quickly. For the first few weeks of games there are always a lot of changes to make.”

The tendency for broadcasts to evolve creatively, perhaps in response to audience feedback or the interest of producers, are another motivator of changes that may make new demands on data handling. “Maybe the producers want to add more graphics or change how the graphics work,” Torres suggests. “But once it’s set for the season there’s often not a lot that needs to be done. A lot of the time we have to get used to how the story changes, when it comes to standings for playoffs and things like that. Once you get into playoffs the format of data changes.”

Focusing Minds On Data

With everyone involved anxious to keep a complex situation as simple as possible, Torres is keen to point out a few common sources of trouble. “Clients sometimes come to us, wanting a data-driven graphics package, and they don’t have access to any of the data. One hurdle is getting access, or working with a data provider.” The demands of live production, too, can concentrate minds toward the creative demands of getting things looking right, perhaps representing a distraction from the technical realities of populating those graphics with meaningful information.

“The biggest problem is that people sometimes don’t think about the complications of the data. They’re only thinking about the graphics, the look, the production but they’re not thinking about where this data is coming from. A lot of time they’re assuming we know how to get the data, but sometimes it’s not that easy, especially when it’s data that’s licensed. Some of the leagues are very open, some of them will only give it to people who’re licensed.”

Anyone with experience in database interactions will not be surprised to discover that this can create some inconsistencies which Torres and his team must clean up. “It’s very much taken for granted you’re getting data from a lot of places,” Torres says. “A lot of them use different keys and tricodes… it’s taken for granted but we do a lot of that cleanup job underneath the hood.” Even a single league might have more than one subtype of data. “Using NFL as an example, they have two big buckets. One is the game stats, the stuff that doesn’t really update in real time. That’s distributed every thirty seconds or minute. The other is clock data - that is a different beast. That’s a different data set.”

Tools Of The Trade

Torres develops the software required to bring everything together using familiar tools. “My main programming environment is Visual Studio. C# has the biggest, broadest set of data tools - not just SQL, though that’s what we first started with, an SQL server. The whole concept was how to get data into our databases, to collate them, to distribute them to our graphics systems as needed. That was fifteen years ago.”

It will be clear by now that there can be no single piece of software which could accommodate every possible scenario in sports graphics. Instead, companies such as RCS have created platforms and architectures - suites of software libraries which can be quickly assembled to connect common data sources and graphics rendering devices with appropriate processing between the two. RCS describes its Foundation architecture as a data service intended to collate and distribute information, made modular for flexibility.

“It started with the idea that, once we get the data into one place we control, we can start to build output modules that can go to anything,” Torres explains. “The system has since moved to the cloud, because it makes so much sense to have everything collected together then those who need the data can subscribe wherever they are. We have APIs and it’s a lot cleaner for everyone to go this way rather than multiple users trying to grab the data sources and do things themselves.”

With increasing experience, opportunities to streamline this sort of workflow begin to emerge. “We’re building the tools to build the product with the greatest efficiency possible. We’re using our experience from doing this so many times, with so many data services, with so many endpoints, that we’ve got very good at having a somewhat modular toolkit at the ready to build what we need to build.”

That approach, Torres estimates, has greatly moderated a once-intimidating level of complexity. “It gets us seventy, eighty per cent of the way there, rather than having to build everything every time. Every project is so bespoke that we’re not going to have an off the shelf product but thankfully, since we’ve done this for so long, we’ve covered all the sports and we have a sense, an awareness of what the requirements are on a sport by sport basis. So, it’s got a little bit routine as far as that’s concerned!”

Supported by

You might also like...

Sports Graphics Production: Data Sources For Live Sports Graphics

The first step in data driven sports graphics production is gathering the data itself. The nature of that data can vary dramatically from sport to sport. Here we discuss some of the data gathering technology and techniques required.

Growing Momentum For 5G In Remote Production

A combination of factors that includes new 3GPP 5G standards & optimizations that have reduced latencies & jitter, new network slicing capabilities and the availability of new LEO satellite services are bringing increasing momentum to the use of 5G for…

Monitoring & Compliance In Broadcast: Accessibility & The Impact Of AI

The proliferation of delivery devices and formats increases the challenges presented by accessibility compliance, but it is an area of rapid AI powered innovation.

Sports Graphics Production: Part 1 – Data Driven Visualization

Welcome to Part 1 of our new series on Sports Graphics Production - a collection of six articles presented in two parts, that examine two key areas of live sports broadcast; data driven visualizations and the rise of virtual studio environments.

Monitoring & Compliance In Broadcast: Monitoring QoS & QoE To Power Monetization

Measuring Quality of Experience (QoE) as perceived by viewers has become critical for monetization both from targeted advertising and direct content consumption.