Meet The Dota 2 Statsman For The Manila Masters: Noxville

May 11, 2017

We got to talk to Ben Steenhuisen about some of his prep for the Masters

One of the most important things that sets Dota 2 apart from traditional sports – and even other esports – is the sheer volume of data we’re able to extract from the game. Whether it’s simple KDA, or something as complex as comparing win ratios between heroes between players (and between positions and between compositions and so on!), Dota 2 has more accessible variables than any other game out there and we the fans love it. It’s what makes Dota 2 such a complex intellectual contest, on top of the mechanical skill required to play it at a high level.

And at the center of all the numbers and algorithms are the statsmen, unsung heroes who have applied their expertise in statistics and big data analysis to the game so that we can confirm our favorite players with more than just personal bias.

Enter Ben “Noxville” Steenhuisen, one of the best of such statsmen in Dota 2 since 2014. The work he does cannot simply be reduced to the phrase “crunching the numbers”. He is also heavily involved in event production, is a software developer for datDota, and maintains his own Dota 2 data website.

Previously, Ben would prepare “stats bibles” with friend and fellow South-African Anthony “scantzor” Hodgson. These bibles were sort of like player and team stats Sparknotes for the broadcasting talent to quickly reference between matches, but the stats game has evolved since then.

“We stopped preparing stats bibles in 2015. Instead, my primary focus has been on creating tools (initially on my own stats site, but more recently with the datDota relaunch) to allow analysts to easily get useful information about players/teams. In the future we're also going to add nifty functionality to compare/contrast a set of teams - which will hopefully be a useful tool for talent in their preparations for big LAN events. Obviously, I still act as a resource for talents to use, whenever they need some specific data to support their talking points or in preparing themselves for events: they often contact me for some data.”

(From Ben's website: Noxville's Dota 2 Statistics)

During pre-production and the actual LAN event itself, Ben also works behind the scenes to bring up the statistics audiences see on-screen from time-to-time during the broadcast.

“I normally prepare maybe 20-30 general stats per team before the event begins. In the games where there's super interesting stuff happening, I just look up and post unique/live stats. In the slower and more standard games I use some of the prepared stats to keep the conversation going.

Barring some significant changes (7.00 being the most significant recent change; but in the past the movement of the Roshan Pit has also always been significant), historical statistics always form the benchmark by which we compare current performances. Early on in patches, another type of narrative is pushing those boundaries – like when players are trying to set a record in a specific regard. The most interesting games from the stats perspective are the ones where we're seeing a lot of crazy and unconventional stuff happening and you're using the historic stats to frame just how incredible the game is.”

On top of adding factual comparisons between teams to add depth and quality to the broadcast, player stats are also the closest we can get to getting an unbiased power ranking for professional teams. We also asked Ben how far public opinion usually is from what the data says are the “best” teams. In particular, we wanted to know which team he thinks has the best shot at winning the Manila Masters (from a statistics perspective, of course).

“I think that each scene is moderately biased towards themselves (a China Weibo poll pre-Kiev Major put Wings as 5th most likely to win the event), but often the public opinion is 80-85% accurate about a team. What they don't see though are the scrim games happening behind the scenes - and this distorts our perspective of teams.

I expect to see OG in the finals, probably against IG or EG. Faceless, NP and Newbee are probably the weaker teams; the Filipino qualifier winner is a completely unpredictable team. I'd have said Secret might also surprise - but they (like NP) made a recent roster change - so it's hard to tell now how they will perform in a few weeks’ time.”

Ben is currently busy working with the production team for the Manila Masters in preparing all the various tools and graphics that will come into play during the broadcast of the event. During the event itself, he will be right there behind the scenes helping bring up stat comparisons for casters to use, and MVP numbers for audiences to get hyped for.