Quick Gameplay Thoughts: November 30

Hi folks,

Usual Disclaimers

These posts will often contain talk about future work we're doing, or planning to do, that isn't yet guaranteed to ship. The nature of the work could change or, depending on what we discover, projects mentioned may get delayed or even stopped. If you'd like to see a Tweet whenever a new one of these posts goes up: https://twitter.com/RiotMeddler

---

Marksmen

Been seeing more discussion about the state of marksmen below. Some notes below covering what we’re seeing, what we’re thinking, and an early take on what changes we’re considering.

Data and Observations

• We’ve been seeing some players on forum/social media argue strongly that Marksmen are weak. That argument comes in a few different versions, with the most common being ‘all marksmen are weak’, ‘Marksmen who aren’t on this short list are weak’ and ‘all crit marksmen are bad’.
• Based off that we did some surveying of players from North America, Korea and Brazil this week to better understand how representative those feelings are. Results from that:

All three regions feel jungle is the strongest position, followed fairly closely by mid Korean players see top, bot and support as all around the same power level North American players see top as 3rd most powerful, followed by support, followed by bot Brazilian players see bot and support as equally powerful in 3rd place, with top clearly the weakest position in their eyes

• We then also looked at which roles players have been choosing as their primary selection when queueing as another data point (what people are choosing to play versus how they perceive power). Overall we’re seeing some decline in players picking bot as preferred position over the few weeks since Worlds ended. Again there’s quite a bit of regional variation, with a larger drop in NA for example but little change in Korea. Also worth noting that desire to play support goes down as much or more in most regions over the same period, and that the decline is seen primarily in Ranked, with normal draft affected but to a lesser degree. Also important to note that declines in both support and bot preference starts before preseason, so this isn't caused by preseason changes, at least not primarily.
• Turning to in game data, win rate isn’t very helpful here, given most games still have a marksman versus another marksman. Pick rate tells us that, on average, marksmen are still almost always being picked bot lane, with the most common non marksman being Karthus, who composes about 1% of champ plays in bot lane. That does vary noticeably by context though, with Karthus’s pick rate in bot at 2% in high MMR (nearly 5% in NA), versus no measurable play at all bot lane in Brazil.
• Worth noting that Marksmen, including some crit marksmen, remained dominant in professional play as of a few weeks ago. We believe that will still be the case when professional play resumes too. Context is pretty different however, both because play between organized teams is very different from solo queue and because the pro meta is noticeably different overall (e.g. more tanks, more macro play, more effectiveness from champs that require more team co-ordination).
• Finally, speaking of the state of the meta, we’ve also seen meaingful increases in play rates of many divers, assassins and burst mages after the end of the ranked season/release of the preseason patch, coupled with lower play rates on tanks and defensive supports in general. Playing a marksman often feels worse when either or both of those things happen.

Moving from talking about context to talking about our interpretations:

• We believe some marksmen are currently strong with appropriate agency and satisfaction. Lucian, Ezreal, Miss Fortune and Jhin being good examples. Some others though are failing on at least one of those axes, crit item dependent marksmen in particular. For players who want to play marksmen who don’t have good options other than crit builds that feels understandably bad.
• We believe that itemization, crit primarily, needs work. A lot of that’s the way the items work more than their power levels. Stormrazor for example is really strong at this point, but still not generating good satisfaction.
• Preseason’s a time when willingness to play marksmen impact has dropped historically and then increased again when season starts. We’re seeing that same trend at present. That reflects more marksmen predators getting played, given players tend to flock to less effective but appealing anti carry heavy team comps during less competitive times. Also reflects that fewer players are choosing to queue as support, resulting in more autofilled support players, who then often try experimental champs/builds and pick very aggressive champs rather than tanks or enchanters, leading to less marksmen/support synergy. There are some issues we should be addressing regardless of seasonal trends like this, but also don’t want to overcorrect when some effects are temporary.
• Marksmen who are always forced to build for late game are more likely to be considered unsatisfying. Their balance is also likely to fluctuate more based off the state of the meta. That makes sense where a marksman’s core identity includes ‘late game hypercarry’, but can be a problem for others.
• Marksmen are potentially too skewed towards almost all offense and no defense itemization wise. Being comparatively squishy makes sense as a class weakness. We’ve arguably pushed that too far though, leaving marksmen too tied to multiplicative damage items and therefore excessively brittle when not well protected or in a favorable meta.

Finally, based off that what are we planning on doing?

8.24b

• We’ll be looking at specific marksmen and putting power into some. Candidates we’re currently thinking about include Corki, Xayah, Ashe, Vayne, Trist, Jinx.
• One tactic we’re going to look at for a number of them is giving more starting MR, reducing their vulnerability to harass in lane and magic damage based burst early/mid game. Not sure yet what sort of mixture of just more MR overall versus pulling per level MR forwards into starting MR that will be.
• For Xayah we’ll be looking at whether she and Rakan are too dependent on each other power wise. Certainly want them to be effective together and feel a bit special, some concern their effectiveness is too low when apart though.

9.1 and/or later

• We’ll be looking at crit items. Details to come soon ish once we’ve got them, a core goal will be to make crit items feel more satisfying to most users, given satisfaction has been a long running issue regardless of balance state.
• We’ll also be looking a marksman offense versus defense. Potential approaches include making partial defensive buys decent choices earlier on (e.g. an early BT or Hexdrinker as a decent choice, rather than too often a trap). Might also involve further adjustments to base defensive stats.
• And we’ll be looking at power curves on some marksmen, in particular whether we’ve got champs who’d be better off if they lost a bit of power late game, which they don’t always get to access, in exchange for a better early and mid game.

---

Gold/XP Funneling

For those unfamiliar with the term funneling here means putting more than one position's worth of gold/XP onto a single champ. We took actions to remove it earlier in the year where it involved low interaction strategies that don't give the opposing team sufficient options to respond to it. The main thing that addressed it was a debuff called Monster Hunter on starting jungle items that reduced lane minion gold heavily if you have the highest minion score on your team. That debuff did the job, but also inappropriately punished non funneling play at times. Intent was to only use it as a temporary band aid solution and we removed it in the 8.23.

Our belief is that funneling should be much less powerful than it used to be, since not contesting a lane can be taken advantage of more now through turret plating gold. Putting that much farm on one person also increases their gold bounty, so focusing a funneled champ is more rewarding than it was.

It’s certainly possible that’s not enough to keep funneling under control though without a more explicit penalty, even if it’s not as strong as it was. We’re working on a backup approach at the moment in case it’s needed. Possible direction is a more focused version of Monster Hunter that does a better job of distinguishing between funneled champs and champs who’ve just had a really good game and are picking up some lane farm while holding a tower, split pushing etc. More details to come if that looks necessary, want to make sure it’s an actual problem again, not just people giving it a try now it’s less obviously a poor choice.

---

Next Week

Some thoughts next week on turret plating, tower stats and tower rewards, including possibility of seeing lane swaps in organized play again. Was going to cover those topics today but this ended up really long already.

http://ddragon.leagueoflegends.com/cdn/6.24.1/img/champion/Ziggs.png