Ancient Coin line (support item) changes
tl;dr The new Ancient Coin will make you as rich as you were in patch 5.4; but why were you so rich in patch 5.4?
In patch 5.10 the Ancient Coin line of support items got some sweet buffs. I have been wondering what affect the changes had on support item picks, and also if the stated goal of giving more money to supports with yellow items was achieved.
Happily my accumulated dataset makes digging possible answers to such questions easy. To make the charts in this article I looked at 1,651,985 recent NA ranked solo-queue matches played by players of all skill levels on patches 5.3-5.11.
I will treat the separate questions of 'which item?' and 'more gold?' separately.
First, charts about rate at which each support line is chosen:
Pick rate, here, is a measure of how many green support items are bought per-game. Remember, each team typically has a support, so the actual rate of purchase is likely about half the displayed value. Not precisely half, of course, since some games have no support items purchased, and some have more than two.
I expected this chart to be relatively flat. The Relic Shield based items (green) have not changed significantly since their introduction in preseason 5. Two notable features of this chart, to me, are the dip in purchase rate for patches 5.4 and 5.5, and the slight dip after patch 5.9.
The first dip I cannot really explain. I have skimmed the patch notes from those patches and those around them, and did not seen any obvious suspects. If you know what happened around those patches to cause that first dip, I would be interested to know (tweet me).
The decline after patch 5.9 makes more sense to me. Players are foregoing Relic Shield in order to try out the new Ancient Coin, as we will see below.
This chart is similar to the one above, but is for the Spellthief's Edge (blue) line of support items. There is a sizable drop off after patch 5.9 here as well. I do enjoy seeing explicable, quantified player behavior.
Here is the plot for the yellow support items:
Looking at this plot, again I see two things that stand out. In patch 5.5 players definitely abandoned Relic Shield for Ancient Coin. Looking at the patch notes for 5.5, I do see that Bard was introduced. Could that swing these charts this much? I am not sure. Also, after patch 5.9 we see the bump we were hoping for. Players are reacting to this change by buying more Ancient Coins.
The downward trend from patch 5.6 to patch 5.9 also catches my eye. The span of time there is about two months. By looking at trends like this one, and modulating that information through professional league play around the world, one can see how Riot might think now is a good time to buff the Ancient Coin line.
Regardless, it is also clear from these plots that the Ancient Coin line is still much less popular than the other two choices for support items. Notice the scales of the y-axes in these plots. Players are still quite a bit less likely to buy an Ancient Coin in any given game.
The buff to Ancient Coin was largely about making it provide more passive gold. You get essentially one more gold for each minion that dies near you, and you do not have to be as close (1400 units instead of 1100) to get the passive gold. Does that translate into more gold for supports who buy this item? We can measure.
The following plots show the average gold/game for support champions for games played on a given patch. There are three charts (one for each support item line), and each bar gives, for that patch, the average gold earned by a champion who bought a support item from that line.
First lets look at gold/game for Relic Shield purchasing supports:
Fairly flat. This is to be expected, since there have not been big changes to this item since its release.
Here is the chart for Spellthief's Edge purchasers:
Also flat, but when I look at this chart and the previous chart side by side I get a little bit excited. First, and its a bit hard to see, but blue-item buying supports earn a bit more gold/game than green-item buying supports, on average. But more exciting than that is the contour of these two plots, and how similar they are. What was it about patch 5.4 that gave a small bump in gold/game to both blue- and green-item buying supports? What is the tiny dip on patch 5.8? I do not have good answers to these questions. It might be interesting to chart gold/game for different roles (or even for everyone) across patches to see if it was just supports that were affected or if everyone follows this same contour.
Anyway, interesting ideas from what I thought were going to be a couple of boring charts.
Here is the chart for the Ancient Coin line (what we ostensibly came here for):
Again, those anomalies on patches 5.4 and 5.8, but lo, look after patch 5.9. A few hundred extra gold per game for Ancient Coin users on patches 5.10 and 5.11. The change makes them essentially equally as rich as Spellthief's Edge buyers. Mission accomplished?
I get the same feeling from making these posts as I do from putting together sets of Lego. The main difference is that I get to share these posts with you, while I have to fly my Lego X-Wing around by myself. Ha!
Back in preseason 5, when these support items were brand new, I took a look at how players were choosing which item to go with. It is interesting to see that those patterns that were established super early have persisted with very little change throughout season 5. Even the gold/game numbers on specific champions are close. The recent changes to the Ancient Coin line represent a fairly conservative move, in my opinion; but I am constantly impressed with the great job Riot does balancing the game.
Ok, time to go back to the index and find yourself some more insights (and hopefully a couple of laughs). If you want to be the first on your block to read future articles, get up on the rss feed and follow me on twitter.
The centers of the faces of a cube are dual with the vertices of an octohedron. But league and math fit together in an even more satisfying way. And so we LeagueMath.com.
No mage support gold income was harmed in the making of this article.