Hearth, Heroes & CCL
My Builds My Blog My Tiers
Fixing the Meta - Snowballs, XP Changes, and Avoiding the Passive Meta

View my last Fixing the Meta post here.
View all of my previous blog posts here.

Been a while since I've done one of these, eh?  Truthfully, articles like this take a lot of creative time and effort, and with me being so busy lately, it's been tough to make them happen.  But with all of the controversy surrounding the upcoming XP changes, and the possibility that they may be rolled live as soon as next week  with some (possibly untested) changes, I felt that I had to put out an article on this topic as quickly as possible in an attempt to advocate what I feel is the correct approach to this patch: not rolling it to the live servers yet.  As such, this article may not be the utmost quality that I usually try to provide, and I apologize if that is off-putting.

In this article, I'd like to explain why I believe that the XP changes are somewhat misguided and how they may be targeting an aspect of the game that isn't as much of a problem as it seems.  I also want to explain how these changes will not encourage the sort of play that they are intended to create.  Furthermore, I'll offer my thoughts on some alternative changes that can help correct the problems that this patch is intended to address.  But before we can do that, let's rewind a bit and discuss why these changes are being made, and what the devs are attempting to address.


Let's Talk About "Snowballing" and the XP Changes

A long-standing complaint about the game (and indeed, about any competitive game) has been the idea of "snowballing".  A common staple joke in cartoons is the image of a small snowball rolling downhill which grows bigger and bigger until it becomes impossible to stop, and often creates calamity for the unfortunate souls caught in its path.  The term has a similar meaning in video games.  I"Snowballing" refers to individual or team mistakes that continue to pile onto a player or team to the point that their loss becomes an inevitable outcome that can't be stopped, much like the aforementioned snowball.  Several maps are considered more "snowball-y" than others, such as Tomb of the Spider Queen and Sky Temple, where getting an advantage on the first objective can quickly slide into multiple objectives and a clear map advantage.

In Heroes of the Storm, an effort has always been made to reduce the amount of snowballing that is possible.  Player kills were made less significant in the early game to prevent teams from getting lots of early kills and running away with a game.  The game was built on a framework of "underdog XP", awarding more XP to the team further down in level and further encouraging aggressive moves to regain the lead.  But in many cases, cracks have emerged in that plan.  Top teams will often rotate for kills in the early game if they can use it to leverage value elsewhere, but because early kills have very small death timers and low XP, the "value" in such a move is often questionable and is passed up in favor of soaking minion waves.  This encouraged a bit of passive play that the devs didn't really want.  Furthermore, because structures grant a fair amount of XP, some maps which provide objectives that dish out heavy structure damage (like Sky Temple) can become "snowball-y" because of the XP advantages gained by taking the first objective and shredding an enemy's towers and forts.  Also, while pro players will often capture mercenaries strategically to split an enemy team's focus, regular players would often ignore their value, to their own peril.  And finally, Blizzard has been making an increased effort to respond to some complaints that the "laning phase" of HotS feels somewhat meaningless.

And so, we have the new XP changes.  Here is a list of the changes:
  • When a team destroys a Fort, they’ll now gain a Catapult in every 3rd Minion wave.
  • Removed all Experience earned from destroying a Fort or Keep
  • Decreased Experience earned from destroying Towers by 50%
  • Passive Experience gain increased by 15%
  • Increased Experience gained by defeating defending Mercenaries by 100%
  • Boss Camps, Sentinel Camps (Hanamura Temple), and Support Camps (Volskaya Foundry) will now be flagged as Elite.  Elite Mercenary Camps are Unstoppable & Unbribable.

The idea behind these changes is to try and address these issues.  The idea is to increase the value of laning by providing a clear "reward" for taking forts, without contributing to the possibility of a snowball.  To that end, forts would provide no XP, instead rewarding catapults every few waves.  This, in theory, gives players a reason to keep pushing forts despite their lack of XP by allowing them to gain additional pressure in that lane.  Combine this with a boost to experience gains on camps, and you effectively encourage players to grab camps more often and put more emphasis on lane pressure.  At least, that was the idea.


Reactions to the Changes

These changes were first proposed at Blizzcon 2018, back in early November.  At that time, players had an opportunity to test the changes out, and reactions were rather varied at the time due to the difficulty in assessing the changes in an environment like Blizzcon.  One standout exception was Maksym "Mopsio" Szczypa, who voiced his concerns across multiple tweets, quoted below:
I really really really dislike xp changes from structures. Heroes was supposed to be a brawler but not being able to create XP lead by outrotating enemy and gathering structures. It hurts the game design a lot.
After playing a couple of games on catapults no xp patch client my guesses were correct. Our loved game loses more and more aspects to outplay your opponent. Getting xp was slow and getting lvl 20 took us about 25 minutes. @azgaz please consider taking feedback from top players

Mopsio also provided a thorough analysis of the changes at that time, and gave his own suggestions for improving the game in a Twitlonger post, located here if you'd like to review it (or didn't get to see it previously).

Accepting some of this feedback, slight alterations were made to the proposed XP changes, namely, the dev team removed XP gains from defending against mercenaries.  However, all of the changes outlined above were rolled to PTR on December 3rd, 2018.  And the reactions were....widespread disapproval.  Now, some individuals have argued with me that this is untrue, or that it's not possible to gauge such a thing.  So a sample of some of the community's reactions have been provided below to help prove my point.

For the sake of not stretching this article out to be longer than it already is bound to be, I'm only going to link to the tweets that I could find on my own Twitter feed to showcase the community reaction to the changes, and I'm going to avoid linking multiple tweets from the same individuals over and over (mostly for my own sanity).  This list also covers a fairly large range of individuals, such as pro players, streamers, content creators, and casters, all listed in no particular order.
Rich McIntyre Arcaner
Horsepants Justing ChaosOS
CavalierGuest Lutano Equinox
ishb00 BBJ DarkChimaera
TrikSlyr Lauber Fury
TBKZord Droplets Nurok
Halorin Cris Goku
Tetcher Drated mewnfarez

I trust that this list is sufficient to confirm the fact that almost all of the competitive community views these changes negatively.  And while some were generally pretty tame in their critiques and offered alternative suggestions to fix the changes, others were....not so gentle.  But why are there so many negative outlooks on these changes?  To put it simply: because they didn't accomplish the goal they set out to do.  As it turns out, many of the predictions made by the competitive community were accurate.  The patch fails to make the changes that it sets out to do, and in some areas creates even worse problems.  What are these problems?


Problem 1: The Value of Forts and "Map Control" in General has been Greatly Reduced

Put simply, the loss of XP on forts and the catapult changes can actually make it more valuable to let your forts be taken early than to defend them.  Obviously in any MOBA, the end goal of any game is to reach the enemy team's core, and structures represent the main obstacle in your way to doing that.  Originally, capturing structures not only got your closer to that end goal, but also provided experience to your team, allowing you to build a lead or recover from a deficit after a big fight.  Unfortunately, these changes have removed the inherent XP value from forts, which means that the vast majority of your XP will come from soaking lanes.  And that becomes much harder to do if your lane is pushed up into an "unsafe" zone, contrary to what the devs may have intended.

Allow me to demonstrate with an example. Cursed Hollow is a large three-lane map with multiple twisted corridors that affect the time and safety of rotations, so it's a great example of the problem with these changes.  Let's look at the map as it normally works with all forts and walls standing (keeping in mind that this is only a relative example of the map's function in normal play and not the be-all-end-all, this-is-how-the-pros-do-it view of the map).

The blue and red lines indicate the "safe soak" line for their respective teams, with the space between them representing the veritable "no man's land" in which you could potentially be rotated onto and killed.  The red arrows indicate some possible rotations that the red team could take to try and get behind a blue team player in that situation.  As you can see, these arrows aren't too far out on the map, and if a rotation leads to a fight that the red team doesn't want, it's not a far distance to get away.  Now let's look at the same map with the assumption that a fort has fallen, and that walls are either weakened or gone, a typical "mid-game" scenario.

As you can see, the "safe zone" for blue players has retreated, indicating the increased map control for the red team.  However, this did not mean that the red team gained an equal amount of territory as "safe" territory.  In fact, the "no man's land" gap between safe zones actually got bigger, and the red team is still mostly restricted to their half of the map in terms of safe territory.  This is because even though the red team has taken a fort, they are not necessarily able to just go anywhere they want within that territory.  Remember, the blue team is closer to all of these locations than the red team, including the "no man's land" territory.  As a result, blue team can reach these spots faster when traveling from any of their safe locations.  You can see the red arrows indicating the new paths that the red team now has to take in order to get around the blue team's "safe" line and kill a blue player in "no man's land".  They're all awfully close to the blue base!

In the current patch, this is fine for two key reasons: First, because the red team gained an XP lead from taking the fort, which helps counterbalance the much more difficult task of earning rotation kills.  Second, because if the blue team decides to hide in their "safe zone", the red team has total freedom to go after the other forts and build an even larger XP lead.  This is what you'd normally call "map control".  It forces the blue team to go outside of their comfortable safe zone if they want to keep the red team from running roughshod over the rest of their territory and build their lead even more.  In other words, taking forts reduces risk of making rotations for the red team, and increases that risk for the blue team.

However, in the new patch, both of these advantages are gone.  The red team no longer gained an XP lead from the fort they took, and if they threaten to take the other forts, they won't build their lead any more than they already have, and all they'll do is make it harder for themselves to soak the lanes that are now going to be closer to the blue base.  Because of the risk associated with such rotations, there isn't very much for teams to actually do now.  A team which loses a fort can sit further back and safely soak their lane since losing another fort doesn't make the situation any worse.  And since the lane now has periodic catapults, it will almost always be pushed against them, ensuring that the wave will frequently fall within their safe areas or at least be confined to a lengthened "no man's land" in which their rotations are safer than the other team's.  By comparison, the winning team has to play risky in trying to soak those same waves now because of how much larger and further away the "no man's land" is, and how much closer the other team is to its base.

All of these factors together tie into what I'm calling the "passive meta".  Many pros have indicated that they see greatly decreased value in taking an early structure lead now because of the way these changes reduce the meaning in having map control, and that these changes make stacking heroes with late-game power considerably stronger because of the fact that both teams have little incentive to push each other's forts down.  As a result, the best move is....to be passive.  Soak your lanes, capture your mercs, and only fight when absolutely necessary.  The dangerous but captivating rotations that the HGC boasted will be fewer and further between, because the value from such moves will be greatly reduced.

Of course, this map is only one example.  This problem is greatly reduced on smaller maps and two-lane maps, where the safe zones are closer together and rotations can be made more safely.  And on maps where the changes are mitigated, such as Towers of Doom, there is little change from how they played before.  But the end result is the same: we now have many pros saying that they foresee the meta becoming more passive instead of more aggressive, as the changes were seeking to achieve.


Problem 2: The Changes Mistakenly Conflate "Snowballing" with Ordinary Mistakes

As previously mentioned, a "snowball" effect is when a simple mistake continues to build and build until the team that made that mistake no longer has an avenue for victory.  For example, Tomb of the Spider Queen is a map in which the first objective can often provide significant value and make it easier to pressure for another turn-in, creating a sense of "snowballing" as that one mistake leads to a cascade of success for the other team and a big XP lead.  But this is often mistakenly combined with another scenario: games in which a team makes several consecutive mistakes which lead to a large XP lead.  That isn't snowballing, that is simply a series of errors, and sequential errors like that should be punished accordingly.  But the current design of the patch doesn't distinguish between these two things at all.

One great example showcasing the difference between "snowballing" and plain old mistakes is the now famous game from Zaelia's stream in which his team was down all six forts and 40 kills on Infernal Shrines.  Losing a game this decisively requires making not one, but several major mistakes, mistakes which would ordinarily lead to a significant level lead, likely 3 or more levels.  And yet, under the new system, the difference was only a single level.  This is a massive difference in how the game is being handled.  And that's not necessarily a good thing.  You may be wondering "why wouldn't it be good for level gaps to be closed?" and the answer has multiple parts.  Let's break them down individually.

My first issue with this approach of treating every lopsided level lead as a "snowballed" game is that excessive forgiveness of player mistakes can lead to a game where the majority of the match doesn't really matter, so long as you're doing everything you can to soak experience.  This could result in matches where the old split-push strategies come to bear again, granting excessive match forgiveness to teams that soak away any mistakes they've made throughout the match.  When a player gets caught out, or when a team fight is lost, or when the team throws on the boss, that should always matter and it should always lead to immediate consequences for the team that made the error.  You don't want that one error to end the game early, certainly, but you do want repeated mistakes to have clear consequences, or players will not be able to learn from them.

My second issue with this approach is this belief that simply closing an experience gap will dramatically alter mentality and give both teams a better chance.  The reality of the situation is that sometimes, players are going to have matches that you can sort of tell aren't going to end well.  Sometimes you're going to have a game with a 40 kill gap.  Sometimes you'll have that player who first-picks a hero without regard for team strategy, raves at his team, then gets regularly killed in lane.  And honestly, it's okay if games like that end early.  One of the advantages of HotS is that matches are shorter to begin with, so a match that is particularly "stompy" is even shorter and it allows you to shake it off and move on quickly.  But if that same match is artificially changed to keep both teams close in level, with only a level or two separating them, it's very likely that such matches will continue to drag on due to the fact that there's no significant advantage on either side and teams will be afraid they're going to be the one to make the big "late-game mistake" that loses the match.  As anyone who has been in a 25-minute game that should've ended at the 10-15 minute mark can attest, these sorts of games are infinitely more frustrating because they feel like your time is being wasted.  It's one thing to want to give both teams a fair shake, but it's quite another if teams are stuck playing out a game whose results are pretty much a foregone conclusion.

With all of the debate surrounding how much of the community's lack of education is the fault of the game vs the fault of the community vs other factors, it's interesting to me that we should see a change which, for the most part, seems to be aiming to make matches far more forgiving of player mistakes.  The most direct result of excessive mistake forgiveness is that the players making those mistakes aren't given clear feedback as to what they did wrong, and they will continue making those mistakes as a result.  The game should seek to avoid making changes which could potentially teach players incorrectly, and the obscurity of how this new patch works (at least to the average player) will make it harder for them to pick out their errors and improve upon them.


Problem 3: The Changes Make Experience and Level Leads Far Less Intuitive than Before

Heroes is a game which has mostly removed the sense of resource gathering, in that the game has chosen to avoid gold and an item system.  However, one resource still remains: experience.  This resource is often undervalued by many players and is commonly misunderstood because of the game's general struggles to adequately convey how experience is actually gained.  There is no real indicator to tell when a player is "soaking" correctly outside of the small XP text that appears when a hero claims experience, which is not very easily visible and doesn't give a player any clear indication of where soak range begins or ends.  This is especially problematic on heroes like Abathur or Azmodan who have some non-traditional means of soaking.

But in general terms, the game never really needed to convey this information too clearly in the past because most of it was self-evident.  The most common activities in a game, that being killing enemy heroes and taking their structures, all awarded XP and directly contributed to the team's experience lead.  So a team which has a large lead in kills or a large lead in structures almost always had a clear lead in experience, reducing the need for the game to give clear explanations about the range and function of soak in this game.  In other words, the relatively simple nature of the game allowed it to "get away with" not properly explaining the mechanics of soak to its players.

This also extends to the game's "jungle".  Heroes has greatly simplified the process of jungling by making mercenaries into a form of super-powered laning minions that can be captured to help push out lanes, rather than being a large buff and XP farm to limit the amount of XP "stolen" by players visiting each other's lanes.  Often these camps go undervalued by players who don't realize that mercenaries provide experience on top of their general lane pressure.  And again, the only indication of such is the small "+XP" text you get when you capture one, which could just as easily be confused with your team mates's progress in lane.  Nothing in the game right now makes it clear that the mercenaries you just captured gave your team experience unless you can read the aforementioned messages, which aren't very large or obvious and are very easy to miss.

This all feeds into the problems with the new patch.  With forts no longer providing XP, the most clear indicator of a "lead" in each game is now meaningless.  A lot of experience is coming from activities that don't convey themselves very clearly (such as mercenary camps and minion waves), and this will leave teams will be confused as to why their structure lead or their much larger number of kills isn't translating into a clear advantage in their levels.  Of course, we'll know that the other team is compensating by soaking up minion waves and capturing mercenaries, but most players won't understand that.  As a result, leads will become even more difficult for players to understand than they were before, leading to increased frustration at the game in general.


So What Needs to Change?

For starters, the XP changes should not go live next week, in any form.  Rolling the reworks, the armor changes, and the radii changes alongside the new event and its content already makes for a fairly significant patch as it is, with plenty of changes to mix things up.  They are more than enough to satisfy the general population for a while.  I understand the passion that the devs have for this change....after all, they spent months working on it....but there is no need to rush it out for the holidays, especially when there are concerns that it might not be ready yet.  The XP changes should remain on PTR for no less than an additional week, if not through all of December, so that the devs can slowly iterate the changes and we have ample time to test them before they go live.  There's no shame in waiting to thoroughly test these changes and roll them out in January after they've been rigorously refined by the community.

As for my own personal suggestions, I have several thoughts on this topic.

First, forts and keeps need to provide XP in some form.  Some have suggested more or less than I'm thinking is needed, but I think 50% of current XP is a good middle ground and there's room to tweak it from there.  I also think the devs should add Mopsio's suggestion of allowing structures to award "comeback" XP in the same way that hero kills do right now.  This gives more opportunities to teams that are behind in XP to catch up, but unlike the current XP changes, this would only help those teams if they are doing the right things to try and catch up.  Personally, I'd like to also extend that bonus to mercenary camp captures (in exchange for a reduction in their new XP gains, of course), as this will also encourage players to make proper plays while they are behind.  Adding more ways to obtain comeback XP is an excellent way to encourage good play-making from a team that is behind without being too forgiving of their mistakes, as the current system is.  Of course, with the expansion of comeback XP, the passive XP that was added in this patch should be removed.  This is in part because it's no longer necessary, and also because it's part of the reason that leads became more obscure to begin with.

Secondly, I agree with several of ChaosOS's suggestions, as listed here.  There was no real reason to change how catapults are scaled, and they should be returned to their original levels.  And we are also in agreement that it would be a great idea to alter mercenary timers so that they are popping up more frequently, which will increase the number of ways that teams can remain active during a match and might help to improve the "jungle" aspects of this game, as well as providing a needed opportunity niche for melee assassins.  I don't think they should be halved, however, as that feels much too frequent with the XP bonuses they're being given.  Perhaps closer to 75% of their current timers, again, with the understanding that they can be further tweaked from there.

Third, I believe that if the devs want to encourage aggression, there is one sure-fire way to do so: adjust the scaling of hero kills so that early kills are worth more, and try to keep late-game kills similar to where they are now.  The early game is played largely safe by professional players because of the low value in kills and the much better value in structures or minion waves.  Higher early value from kills will inevitably result in more aggressive plays, breeding the sort of exciting team fights that the HGC is known for.

Fourth, I believe that the devs need to begin work on clearer indicators for things like experience so that they are more understandable to the average player.  I think that when you perform activities that gain the team experience, there should be a sort of "XP feed" (similar to the kill feed) at the top of the game panel which tracks the experience that you're giving your team through your own activities.  When you soak a minion wave or take a merc camp, the XP you gain from being near those minions would appear in large and visibly clear numbers at the top of the screen right below your XP bar, allowing players to make the obvious association that "oh, this thing gave my team some XP".  This will help players understand when they are doing things that gain XP for the team and might help encourage them to do these things more often.  It might also be helpful to include additional ping options so that other players can ping a minion wave and say "hey guys, we're missing experience right here, can someone get on that?", which will also help players to guide each other towards making better decisions.

Lastly, and most importantly, I think that the devs need to commit to making significant and sweeping changes to the maps in this game to further reduce true snowballing.  While some maps are generally pretty good about this, like Towers of Doom, other maps are absolutely terrible on this front.  A few examples:
  • Tomb of the Spider Queen suffers from a lot of this snowball-favoring design, for example, as an early turn-in can quickly result in easier future turn-ins.  Furthermore, the smaller map size and design of the lanes don't allow as much aggression on rotation, and offer little to nothing in the way of "jungling", which will lead to more focus on soak and gem collection.  I think that each following turn-in needs to cost more than just 5 extra gems, and I think the map needs some design tweaks to allow for more room to rotate and to make space for additional merc camps for more jungling options.
  • Sky Temple is easily one of the most snowball-y maps in the game right now.  The map is very open due to the centralized temples, which leaves lots of rotation room, but the top lane is further away which makes it harder to rotate up in time.  In my opinion, part of the problem is the fact that temples don't give you any control over the targeted lanes, so it may be beneficial to tweak the temples in some way to allow more control over where they fire.
These are not the only maps, of course, just a few examples of changes that might help to reduce map-based snowballing and player frustrations.  There are numerous other maps with changes needed here or there.

This last point about maps, in my opinion, is the change that is needed the absolute most out of any of the changes listed here.  I think that the devs need to be willing to make a massive time commitment to revisiting their map designs this year to refresh them and make them all work better and with less snowballing.  I also think that the devs should work closely with the professional scene and with the community to decide which maps need to be revisited the most and what sorts of problems they are experiencing on those maps.


I appreciate those of you who took the time to read through this article.  I'd like to thank the members of this community for coming together to discuss these changes and managing to do so in a relatively civilized manner.  It's easy to resort to flaming or negativity with massive changes like these, especially if the perception is largely negative.  But I have faith that the development team will work together with us to come up with a satisfactory conclusion to this debate.  I would like to also thank the development team for putting in those months of effort to try and come up with ways to spice the game up, and despite what my comments may have sounded like in the past, I really do look forward to working together with them to make these changes work better for all of us.

As I mentioned earlier, this article was put together in something of a rush due to having a limited amount of time available to me, so I apologize for any unclear parts, typos, etc that may have resulted from the long nights I spent writing it.  Please let me know if you find any issues and I will try to clarify or fix them in a timely manner.  Feel free to let know what you think of this article in the comments section below, on Reddit, or hit me up on Twitter at my handle @CriticKitten.  I look forward to hearing your thoughts, and to hearing more information next week about what sorts of changes the devs have in mind to adjust these changes.
There are no comments for this post.
Any thoughts on reducing late game death timers?  I feel like a lot of games come down to a single team fight after minute 20 and the team that loses that fight can’t respond, game over. It almost makes the rest of the game not matter so long as you win that last team flight. 
I'm of two minds on that, honestly.  On the one hand, it definitely gets annoying to be isolated from the game for long period of time.  But on the other hand, it's part of what makes those late-game deaths so meaningful in the first place: the effective "penalty period" it creates on the team that loses someone.  You want the early game to matter too obviously and I think some of my suggestions can help with that, but you do still want room for teams that fall behind to rally and recover.  Death timers help to enable that.
A very good piece, thanks for taking the time to write down your thoughts! 
Thanks for taking the time to thoughtfully summarize and assess the community feedback. I think you and CavalierGuest have made some very good points without straying from the categories of constructive criticism and usable feedback.
good article. i agree with everything here, except the suggestion that forts should give comeback xp. they absolutely should not. because there are only a limited amount of forts to be taken, this means the losing team will simply have access to more XP than the winning team and the rubberbanding would be incredible. There is already more than enough rubberbanding in this game, if they want to decrease snowballing they should do what spiderqueen or blackhearts does and make objectives harder to take the more you have activated it (maybe)