Archive:Leaguepedia Articles/Patch Analysis of 3.14
|Statistical Analysis of 3.14 in Competitive Play|
With the big preseason patch, 3.14, Riot publicized a lot of their gameplay goals through Reddit posts, forum posts, and a video. Now that it has been several weeks since the patch, and there has been a sizable amount of competitive games played on the new Summoners Rift, I wanted to take a look back and see how this big shift has really changed the game and if they matched up with Riot’s initial goals. The four areas of changes I am going to be focusing on are supports, junglers, game pacing and snowballing. We have analyzed 156 competitive games: 85 on patch 3.13 (S3WC and early OGN Winter) and 71 on patch 3.14 or 3.15 (BotA, NA LCS Promo, EU LCS Promo, and later OGN Winter), all of the data here (unless stated otherwise) was from this collection of 150+ games, which I think is a suitable sample size.
The first thing we are going to look at is the role changes, leading with the supports. Riot had ambitious goals for the supports; to increase the gold flow, provide more itemization options, and reduce their burden to ward. Lets take a look at if these goals were fully realized. We collected several stats (seen on the right) which analyzed the gold and items of supports before and after this patch.
The first thing to look at is the GPM difference. Overall, supports have significantly more gold post-patch with an increase of 15% in GPM, which is quite significant, and I think it is due almost singlehandedly to the new support items which give much higher gold over a game than the previous gold earning items. Add in the fact that these supports now have these gold generating items in their core build, building them every game, while some supports before the patch would never even get a Philosopher's Stone or Kage's Lucky Pick at all - I believe they were definitely successful in their goal to give supports more gold flow.
The next stat that is significant to this discussion is Item Value and “% of gold spent on items.” I think these were the biggest changes in the patch for supports, as now since you can’t place more than 3 wards at a time, supports no longer have to buy more wards throughout the game and thus spend a majority of their gold on real items such as Talisman of Ascension or Mikael's Crucible. The end game item value (the total cost of their items at the end of the game) over doubled, and while some of that is accountable for the slightly longer games, it is more significantly attributed to the fact that they no longer have to buy as many consumables, and thus their % of gold spent on items went up drastically. However, while they definitely are buying more items now, I am skeptical about the goal of diversifying options as a large majority of supports are only buying the same 2 items. Despite the addition of AP ratios on many of the new supports, less than 10% of the supports in the sample bought any significant AP items. Overall, Riot was successful in increasing a support’s gold flow, and allowing supports to spend their money on big-ticket items, but they still need to balance out the items so there is more diversity.
The next big change to look at was the changes to junglers. Riot’s goals for the jungle changes were for junglers to be able to reach their “end game fantasy,” for there to be more customization in the jungle, and to bring more potential for farming junglers. To look at this we collected stats similar to supports, on the left.
Like supports, junglers also saw an increase in gold flow. Their GPM went up decently though not as much of an increase as supports - but still an increase. This is interesting though because while their GPM did increase, they didn’t farm much more at all, only gaining a 1% increase in Minion Kills Per Minute. I think the GPM increase was from the unique passives added to jungler items which give them additional gold on monster kills and the increased gold bounty on monsters as the game progresses, not because jungler’s styles have changed. Also, similar to supports, junglers now get to spend significantly more of their cash on actual items, due to the removal of oracles and the vision changes. Overall, Riot was successful in letting junglers reach their “end game fantasy,” but the addition of the Wight is still a puzzle to me.
Even though the changes helped spawn diversity and creativity in the jungle, this has not transferred into competitive play. We took a look at the jungle items that people purchase, and it continues to be very lopsided towards Spirit of the Ancient Golem, yielding 80% of all purchases. And while there are slight variations in routes, the playstyle is still primarily gank focused tanky-oriented junglers who sacrifice farm for the solos.
The next thing to look at is game pacing. Riot didn’t have very many goals specifically for game pacing (they more tackled snowballing which is its own section below). Instead, Riot’s goal for game pacing is just that it would be more “dynamic and exciting.” We looked at some benchmark moments in the game, but the main focus of this section is looking at the average game time, which went up by almost 5 minutes or 14.2%. Games are drawn out more than before the patch, although there is more variance, which was part of Riot’s goals. The other noteworthy thing we found was how first bloods typically came later in the game, while dragon and tower kills seemed around the same time. In fact, I was expecting a later dragon timer due to the fact that they are worth less early, but for some reason people get dragon slightly sooner now. While “if Riot really achieved their goal of more dynamic and exciting games” is too vague to quantify, one product of this recent patch is the significantly longer and more drawn out games.
The last thing to look at is snowballing. Riot has tried to decrease the snowball effect caused by early objectives and inhibitors. To look at this, we analyzed the win rates after taking first objective, seen on the right. We saw an overall decrease in “snowballing,” but two things stuck out: the 3.6% increase in first blood win rate, and the 11.9% decrease in first dragon win rate. The dragon decrease, I believe, is a misconception in the current meta where people overvalue early dragons despite nerfed rewards at lower levels. However, the first blood increase is strange because in this preseason patch they actually nerfed early kills by having them give less gold before 4 minutes. I think this might be related to the game pacing changes and how they have affected the laning phase. First blood is on average 25% slower in post-3.14, and the distribution is more right-skewed, meaning there are significantly more extremely late first bloods. For example, in pre-3.14, only 1% of first bloods were after 10 minutes; yet in post-3.14, 15.6% of first bloods were later than 10 minutes. As Team Dignitas’s analyst, Vlanitak, theorized, “[First Blood] happens later now in a big amount of games and you can do a drag or tower after.” One last thing to consider though when analyzing snowballing in competitive games, is that less mistakes are made than in normal games; so while the inhibitor changes did not dent the impressive first inhibitor win rate in competition, it probably had a bigger effect in public games.
In the end, I believe this patch was overall successful conceptually. Most of the goals that Riot put in place went through correctly. However, the biggest plague to this patch is the balance which disturbs some of these goals that were in place. Many items such as Talisman of Ascension, Spirit of the Ancient Golem, Spirit Visage & Sunfire Cape, and Athene's Unholy Grail are bought every single game, and this counters one of Riot's big goals of this patch to increase dynamic play. And while there will never be an end to snowballing, nor will supports ever truly reach that "end game fantasy," this patch is definitely a step in the right direction and it's effects ripple through competitive play.