#dojomadness

Berlin, 03/03/2016

User growth continues to speed up yet again for LOLSUMO. We grew our Daily Active Users to 30,000 just two months into 2016. This tripled the numbers from our December update.

We expect our DAU to continue along their present upward trend with our new Superlatives feature as well as through polishing and incrementally improving our current features.

Superlatives provide League of Legends players with their best and worst performances across distinct categories. As an example, “Signature Champion” shows the top five champions that you win most games with. “See No Evil” makes you aware of the five champions that you placed the least wards with.

This allows our users to see interesting and fun facts about their own play over a longer period of time. They are able to spot key information that helps to determine their strongest champions. Users are also enabled to determine the champions they play often but need to improve on. Superlatives furthermore serve to highlight hidden discrepancies in your play style. A support player might realize that he places less wards on Thresh than with Nami or Morgana and correct this mistake.

We would like to thank our community for the continued support in comments and app ratings:

LOLSUMO is putting the finishing touches on a new sophisticated system to calculate your personal best build for your current match.

Becoming better at League of Legends depends on a lot of factors like discipline, time investment, acquiring new skills and refining existing ones. The same is true for LOLSUMO.

LOLSUMO’s item builds are powered by big data. Every build is calculated specifically towards your current matchup. We are factoring in data like the ranks of everyone in the match, all champions involved, the lanes these champions will be on, and a lot of additional information. This system is right at the heart of LOLSUMO, the engine that powers it all. Engines can be tweaked and tuned, and us being DOJO MADNESS after all, we like to go a step further.

In order to come up with potential item builds, you can gather a lot of data focusing on how often an item ends up winning.

We have been able to put a number on how good an item build actually is. That is understandably a lot more complex, but also a lot more sophisticated than the method mentioned before. We are able to rate item synergies as well as factor in your champion, your lane and your opponent. As an example, we know that for Champion A, “Item 4”  is strong when used with “Item 67” against Champion F. Being able to put a number on this small piece of the puzzle allowed us to rate complete builds.

Once we were able to do that, we unleashed the (data-) kraken and collected a metric ton of information to calculate the best item builds used for a specific champion against a specific opponent on a specific lane.

This does not mean that there may not be better builds around that no one has found yet. The beautiful thing here is that as soon as someone tries them, we will know them, too - and by extension, any of our users who finds himself or herself in the same matchup. We are constantly processing big data and updating our guides accordingly to give you the best one possible.

Why is this better than our previous system? 

One example: 

Suppose there is a true killer build, but the player using it loses his match due to a teammate disconnecting, and he never tries it again. When purely looking into win/loss statistics, you will miss out on this build. Not so with our system. 

And the cherry on top of our cake: We have mathematical proof that our rating system works. We test it blindfolded against reality, and when it says a build is good, reality thinks so, too.

That is all for now, dear readers. 

Youngbuck - Coach for G2 Esports: “You finished week 1 with 2-0, but did you think at any point that you might lose one of these matches?”

Kikis, Top Laner for G2 Esports: “Who do you think are your biggest rivals at TOP lane?”

We wonder, sometimes. Then again, we are a company dedicated to supporting eSports and helping gamers master their play. Just a normal day in the office, then!

MOBA-Games like Dota2 are highly complex constructs of rules and choices. The decision of which skills to spec and when shapes the way the game plays out and may propel you ahead to win, or leave you behind to lose. Our data science department made such a path visible based on the Frankfurt Major.

The image below (click here for a larger view) depicts a tree of skill choices being made with the Winter Wyvern and the amount of times they were chosen (Numbers from 0.01 to 1.00 are percentages from 1% to 100% respectively. The second number depicts the amount of times the choice was made). The most successful build path is marked in red.

Merry Christmas and a happy new year from the DOJO MADNESS team!

Berlin, 12/14/2015

Continuous and accelerating growth has led us to new heights with our LOLSUMO App. Only 5 months from our global launch we just surpassed 10,000 Daily Active Users and doubled our November averages. The mobile app is holding steady at 75% Android and 25% iOS users, while user numbers continue to grow on our recently released Web App as well.

We want to thank all our incredible users for their continued support and encouragement, which we see reflected in an average of 4.87 stars over the last 1,000 reviews and ratings in the mobile app stores.

Here is a snapshot of what the community had to say about LOLSUMO: