The Terror that is Sion: Winning by Losing

During patch 8.20, a build that Riot dubbed “Strategically Minded Sion” rose in popularity (and infamy). Known more commonly as “inting Sion,” the build sacrifices KDA to draw as much map pressure as possible by mindlessly pushing towers, either winning the game himself or enabling one’s team to do so.

Intuitively, the strategy might seem ineffective, because dying gives the enemy gold and tends to relieve map pressure—however, like a new TSM roster, appearances can be deceiving. It really doesn’t matter how many times Sion dies, since with his Soul Furnace’s HP-stacking passive he’ll be inherently tanky as long as he knows what minions are, and the more he dies the less gold he’ll be worth (although as of preseason 9, the new bounty system has changed this). As a result, it becomes worthwhile for Sion to just trade his life for turrets. League is ultimately an objective-based game, and Sion’s abilities allow for a lot of turret damage, as well as waveclear that increases pushing power. Combined with Demolish, Titanic Hydra, Zz’Rot, and Sterak’s, Sion might as well have a wrecking ball instead of an axe. Then, there’s the issue of solo queue. Dealing with potent split pushers of any kind requires proper coordination amongst team members, and given with the Hue Jackson-like leadership of many teams, inting Sion can cause quite a snafu if not dealt with properly.

On October 3rd, a Reddit post put this build into the spotlight. To understand what came next, we pulled game data from September 15 to the end of November, counting every single time Sion died in a ranked solo queue game to compare “OG” Sion to Sion’s more current form. This data was plotted by day, with average deaths per game and games played shown. Note that the first vertical bar marks when the inting Sion post hit Reddit (and Mean Girls day). The second bar marks November 12th, when the ranked season ended. To give context to Sion’s death rate, we added a blue curve showing the overall death rate for all champions.

This plot compares Sion's death rate with the overall death rate over a period of 70 days. The 'Sion Games Played' trendline is shown using a rolling average with a 7-day window, to smooth out day-to-day variability and make the overall trend more clear.

As it turns out, Sion’s death rate used to be below average, but after the Inting Sion Reddit post, he experienced a massive spike in play- and death-rates. Perhaps everyone wanted to feed before, but didn’t have an excuse to do it until now. Sion’s death and play rates remain consistently above average weeks after the post, but afterwards Sion’s death rate increases substantially while overall play rates decrease. This makes sense, since we expect fewer people to play solo queue after the ranked season ends. Those who didn’t either stopped trying, or decided now was the appropriate time to try some troll builds like inting Sion or picking Ivern. This may have caused an overall increase in death rate amongst all champions during the period after November 12. If you had an AP Shyvana in your game, you likely weren’t the only one.

We also looked at ranked games from Patch 8.20 (October 10-23), and counted each time any champion was killed each game as well as each time they were executed, meaning killed by something other than a champion: like minions, towers, or jungle monsters.

This plot shows the death and execute rates for all champions over Patch 8.20 in NA solo queue ranked games.

As you can see, our undead, KDA-impaired friend averages 6.4 deaths a game, and has more executes than any other champion, including Singed. At first, we were surprised to see this: Inting Sion players may have a Teemo-may-care attitude towards giving up deaths, but that doesn’t exactly explain the high execute rate. Then we realized that these executes could be due to the common strategy of taking a jungle camp and dying at level 1. To confirm this hypothesis, we looked at the distribution of execute times for Sion games.

As it turns out, Sion players really are fans of the well-aged cheese. The classic “die to jungle camp” strat seems to be alive and well.

Tidbits on Deaths and Executes in Patch 8.20 Overall

We noticed some interesting things about other champions from the deaths vs. executes plot. In particular:

• Junglers have notably higher execute rates than other roles, which makes sense. If we had a nickel every time a Nidalee or Rek’Sai died trying to clear Krugs or Raptors, we’d probably have enough money to build a Death Star or something (not that we would, our enemies are taken care of). Shyvana is especially high up in the executes—maybe she tends to be too ambitious with early game Dragon solos.
• Out of all markspeople, you’d think Kog’Maw would have the highest death rate to tactically feed maximize passive use, but actually Kalista averages 7 deaths a game, We’re not saying she might be a bad champion, but she might be a bad champion.
• Yasuos average 7.2 deaths a game. Like being the best Zoe player, this is both a good and bad thing. It means that while Yasuo players do feed more than an average champion (let’s be real, we all knew this was true), you'll look good as long as you can avoid going full clown mode.

Finally, some “int”eresting anomalies from our data:

• In one game, an Ezreal and Brand on the same team combined for 44 deaths. Possibly with the power of friendship and the other team’s incompetence, they won that game.
• The highest number of deaths from one champion in a game was from a Nunu in a Silver 5 game, with 4 executes and 42 total deaths. He lost that game.
  • Incidentally, the Sion that game had 7 deaths.
• A Jinx in a Bronze 2 game had 30 deaths, yet won the game. We’re not exactly sure how, but we’re guessing the 19-death Tristana on her team wasn’t the cause.
• Some of you may think these outliers only happen in lower ELOs. However, in one D5 game a Karthus died 20 times before winning his game. Seeing as it is Karthus, 20 deaths is somewhat excusable, but possibly excessive (Karthus averaged about 7 deaths per game overall on patch 8.20).