We've recently released the initial version of LoLPortal's pro player database! Check it out here.
The dataset includes hundreds of matches from almost every major tournament from IPL 5 onwards (including the Champions Winter games that were played prior to IPL 5). Still have some work to backfill data from some of the minor tournaments.
I'm still working through adding the CS numbers for quite a few matches. This is slower going than the KDA's, as the scoreboard sometimes gets hidden for the last team fight of the match (especially in Champions Winter). Maybe the upcoming Riot API will allow me to accurately grab all this data and more from tournament realm matches automatically.
For the player table, you can click on the columns to sort by each statistic. I used per time period calculations rather than per game to give a fairer comparison between players. Right now, the player table requires a minimum of 10 matches to have been recorded with data for players to appear (otherwise, you get a constantly changing leaderboard for the various statistics if new teams/players come on the scene and do well in their first few matches).
The colorization of the player table is "dumb" at the moment in that it doesn't take into account the player's position (so the CS stat for supports for example is currently always red). I'll be changing this in a future update.
The KDA formula I'm currently using is (Kills + Assists) / (Deaths + 1). When implementing it, I had considered using (Kills + Assists) / Max(1, Deaths), as someone on my Reddit thread suggested, but this would mean that a player who has, for example, 6 kills, 0 deaths, and 7 assists would have the same KDA as one who has 6 kills, 1 death, and 7 assists, which didn't seem right to me. Suggestions on what formula to use and why are welcome!
The actual player profiles are a bit spartan at the moment, but I've got some improvements and more stats/visualizations planned soon.
Some interesting statistics
CLG's Doublelift lives up to his reputation as a farming king, boasting the second highest CS/30 minutes figure in the world, and by far the highest in North America.
For the solo Top lane position, the highest "kill contributions" (i.e. the percentage of a team's kills that a player was involved in, either with achieving the killing blow or with an assist) are dominated by Koreans, which speaks to the stereotype of Korean Top laners being the best in the world. SKT T1's Reapered currently holds the highest kill contribution, with a percentage of 73%, indicative of how hard he carries his team. Interestingly, TPA's Stanley, has one of the lowest kill contribution percentages at 57.3%, but also has one of the lowest death scores, which speaks to how well TPA executes their split push strategy.
Royal Club's bottom lane duo of MikakoTabe (support) and Uzi (AD Carry) top the assist and kill score categories, respectively. They also have some of the highest death scores, along with the rest of their team members, which points to their overall aggression as a team.
What interesting statistics or patterns do you see in the data? Let me know!
Other site improvements
While the player database has the been the major focus for LoLPortal recently, I've also made some smaller improvements across the site in the last few weeks:
Added a champion profile page for each champion where you can see which teams pick and ban them the most.
Redesigned the tournament pages to give more horizontal room for you to explore the tournament statistics such as picks and bans. I also added a player performance table for each tournament on which I have data.
Added a recent matches section on the home page. In the coming weeks, you'll be seeing some more significant design changes to the home page to provide you a better overview of what's going on in the League of Legends competitive scene.