In HoN, “Pulling” is the emergent mechanic of leading Legion or Hellbourne creeps out of the lane to attack neutral creeps. Within the tri-lane metagame, Pulling is often cited as one of the contributing factors to the proliferation of turtling and farming-based strategies. However, it is important to distinguish whether Pulling is indeed one of the roots of the problem or if it is simply a symptom. This difference is crucial in how we handle Pulling from a design standpoint–do we nuke it, thereby directly weakening farm-based strategies? Or do we leave it alone, instead attacking other points or buffing other aspects, thereby leaving it as an optional but comparably less effective maneuver?
To understand Pulling, we first have to acknowledge that there are both positives and negatives to the mechanic.
Why Pulling is Bad
- Detracts from PvP interaction
- Is characteristic of passive PvE-based play
- Is technically an unintended mechanic
- Requires a gold investment to properly counter
- Is generally considered “annoying” to play against
Why Pulling is Good
- Is an emergent mechanic in a game that is built on emergent gameplay
- May be countered by investment into blockers
- Provides additional income
- Helps jungling heroes
- Gives wards more strategic uses
In game design, there are two schools. The old school of design built games around mechanics that may not have been completely sound in terms of mechanics, but also built their games to last due to the lack of means for post-release support. Think Super Mario World here–extremely solid core gameplay, but full of glitches that enable skilled players, speed-runners, and TAS players to take the game to another level. The new school of design relishes gameplay control, designing every individual aspect of the game but leaving very little room for emergent gameplay mechanics. The end result of this is not only that the game feels extremely polished, but bugs are always considered to be detracting from the core game. Almost every game made nowadays falls into this category. As a gamer, and not a designer, it’s extremely disappointing to see game design slip into the latter, as the serendipitous nature of the old school has led to countless gaming innovations–much of Starcraft’s competitive play is based on mechanical abuse (Mutalisk stacking for example), while entire genres have rested on the laurels of bugs (combos in Street Fighter 2 were not a result of intentional design). As a designer, I wish to point out that this entire blog, including its tagline “Game design in a genre lacking it”, is a purposeful attack on the old school of design. In the case of HoN, this refers specifically to heirloom mechanics and design decisions that obfuscate balance and confuse players to unnecessary ends. Put quite simply, there are portions of the game that simply are not fun to the player. Following this stream of thought, Pulling ain HoN exists not because it was designed, but because it is an emergent mechanic that arose out of a clever abuse of an unintended interaction between lane creeps and neutral creeps. From the perspective of an old school fan, this is part of what makes the game great. From the perspective of a polish-centered designer, this is part of what makes the game so frustrating.
What would happen if pulling were removed?
Tri-lanes would hypothetically be weakened due to the loss of neutral creeps as an effective form of additional income. Currently, according to competitors, the support heroes that lag behind due to their constant engagement as babysitters are the ones who reap the most benefits from Pulling due to the fact that lane creeps provide higher monetary and experiential benefits to a free-farming carry. In today’s game, it’s customary to buy at least two wards; one to cover the pullable neutral spawn point, and one to cover a rune position. To delve deeper along this path, one could also argue that because an additional 100 gold is being introduced into the team’s resources due to the discontinued need to block a neutral spawn, that 100 gold would likely be spent towards either a courier upgrade or a ward to put somewhere else–likely the other rune spot or a position favorable to the middle solo. This would mean that the routes from middle to the side lanes would be bathed in vision, and ganking could in fact become more difficult. In this case, it would be a matter of whether a shift in map vision would be a worthy tradeoff for a decrease in potential farm for support heroes that choose to engage in a tri-lane.
How do we solve pulling without removing it?
The alternative is that pulling is not touched at all, and instead, alternative options, such as buffs to jungling and dual-laning, or a tweak in XP penalties for tri-laning, are implemented. This path rests on the question of whether pulling would be as effective in non-farming-based strategies. For example, it would be unwise to wipe out your creep wave in an early pushing strategy, though pulling could still be used selectively to build up a double wave for a big push. In this usage, pulling is not necessary to do, but is instead an option with both positives and negatives–a true strategic choice. The positives to this solution are very tangible–a mechanic is preserved, while the number of possible strategies increases. The downside is that this is a much riskier route, and requires multiple solutions to discourage tri-lanes (or rather, encourage alternatives), and given the urgency of a stale metagame, the longer, more explorative approach may not always be the one that will please players the most.
As discussed in the Evolving Metagame series, game design is more often than not about a careful balance between multiple sacrifices. Is the sacrifice of the Pulling mechanic worth the decrease in depth and strategic options? Should we be taking the fast route out of the tri-lane metagame, or should we spend the time and resources in developing a more comprehensive solution?
nomeswisdom.wordpress.com 
uh just copy icefrogs solution
ps what is this font
Whatever the hell the default font is 🙂
Cut a bit of that RICE RICE and make them starve for BLOOD
A more graceful solution would be to cut some pull-able camp EXP given or to expose that camp somewhat more(not easily done as it isn’t an symmetric map)
“Or do we leave it alone, instead attacking other points or buffing other aspects…”
Based on the last patch, I would assume buffing every problem is the obvious choice.
do what ever you can to remove tri lanes…they take out skill and just put hero picks in the front line.
and its also VERY boring to watch
Tri-lanes were once considered risky and exciting. Players only think they’re boring because they’ve become the dominant strategy.
Don’t you think trilanes are considered boring because 2 heroes look how 1 hero farm? Awesome!
I would say be conservative. It’s not worth the risk of completely breaking the game by making too many inelegant changes too quickly. Take small steps forward in encouraging 2-1-2 or some new sort of laning arrangement such that they can, in certain situations and lineups, be viable against the trilane.
How does one accomplish that? I don’t know, BUT THAT’S WHY MALIKEN PAYS YOU THE $BIG$BUCKS$ amirite?
Seriously though…tri-lane has to be nerfed but it’s not worth breaking the game over.
Right. What I don’t want is for the game to be locked into 2-1-2 as it is locked into 1-1-3 right now. To me, a 2-1-2 dominant strategy is almost as bad design-wise. The most healthy metagame is one where multiple strategies can flourish simultaneously.
This was a great post. In most games, what makes it fun and exciting is the parts that arent intented. What made half-life so great (apart from the rest of the game!) was the fact that the engine was so amasingly buggable. Using laser-mines to build ladders, bunny-jumping, the way the laser gun propelled you forward like a fucking rocket etc.
Oh, and the patch that removed wall-walking from wow was also a big disappointment for me.
It pleases me tremendously that you as a game developer talks about this, because, as you say, the trend has been to go in the opposite direction. Its obvious that you know what you are doing (or atleast know the theory, what actually happens when changes are implemented can be hard to predict). I’m happy my favorite game seems to be in capable hands.
I actually considered Half Life vs Half Life 2, but then I remembered that Half Life 2 was also riddled with awesome physics bugs 😀
Yeah, that was also a great engine!
Gravity Gun was the single greatest innovation to happen to games in a very long time. The amount of time i’ve spent doing silly stuff with it is ridiculous.
Just skip to 3.16. Didnt know this blog automatically embedded.
copy icefrog’s solution???
The DotA solution does both good and bad things. Because the XP and gold nerfs affect all hero kills, and tri-lanes were never about hero kills in the first place, that change hurts ganking and teamfighting more than tri-lanes. An XP range increase on the other hand is something we’ve been thinking for a long time–that’s why we put it in Casual Mode. In fact, a lot of experimental ideas have gone into Casual Mode. The XP range increase and Tower Armor Reduction mechanics are the two big ones. To say we’d be copying DotA is quite insulting, as we actually had the same idea first! The big difference is that our playerbase tends to be much less accepting of major changes, so we tend to be more careful when implementing such mechanics changes.
Hey Nome, excellent insight. However, there is an option that no one seems to want to talk about too much:
Fuck it.
As has been the recent trend in the competitive scene, aggressive trilanes and aggressive roaming strategies are becoming more and more common. Every other match, we see EG pulling out a pebbles trilane, or having the trilane support roam so hardcore that you’re not even sure which of the two side lanes the trilane is babysitting. (See unz vs eg in the honor tournament recently.) EG put a dsham solo 3v1? Hmm 🙂
So it may be that with new trends in gameplay, the issue of trilanes being dominant may sort itself out with time. Honestly, the strategy of 2/5 members of your team roaming most of the game sounds very appealing from a viewer standpoint.
The other thing that people have been trying to “counter” trilanes is a 2-2-1 setup, which is also very exciting to see. It hasn’t been successful, in that teams almost always adjust their 1-1-3 to deal with it… but it has been successful in that it frequently ends up forcing two 2-2-1 strategies rather than a 2-2-1 and a 1-1-3.
Even jungling fits somewhat well into the trilane-heavy metagame. Sometimes now, teams just abandon the dangerous “1” that has to sit alone and send him to jungle. I’m not sure how viable it will be as we move forward, but actually *picking* a jungler with the intent of having him not bother with the trilane sounds like a cool bit of variety.
Finally, trilane vs trilane matchups are cool. They always end up in scrappy fights for dominance. What’s better as a spectator than having the teams just fight 3v3 for the entire early game?
If anything, I’d be aiming to buff jungling as a viable strategy and let the trilane metagame work itself out for a bit. It looks like it’s still evolving and may be very cool in the time to come.
That’s actually very close (if not identical) to my personal opinion on the issue. I enjoy tri-lane vs. tri-lane. I also enjoy a well-played solo vs. tri-lane (Aluna will be able to solo against a tri-lane fairly well). The problem with that latter option is that HoN’s hero pool doesn’t support the 1v3 setup as well as DotA’s does, but that issue will resolve itself over time as we diversify the pool. My internally proposed solution was actually to:
1. Increase XP range to what it is in casual mode right now. This would help to resolve the zoning problem that happens in a 3v1 or 3v2. It would have potentially negative effects on other lanes, such as a 2v2 where one team outplays the other but is no longer able to zone them out of range. Positive implications on other lanes would include being able to solo more melee heroes in the mid lane. Basically, it’s a dual-edged sword with lots of unseen implications.
2. Differentially scale the amount of XP received when a creep is denied. Effectively, a solo deny would still give 33% XP, while a deny with two heroes would give 44% XP, while a deny with three heroes would give 55% XP, etc. This would be the “hard nerf” to tri-lanes, but it also buffs players who take on a higher number of opponents.
3. Buff pushing. Can’t give details on this one.
4. Buff jungling. Introduce an item to allow more heroes to jungle, albeit much less efficiently than heroes that are designed to be jungle-friendly.
The immediate problem with the above, of course, is that it’s a whole lot of changes, and we’re currently not a position to introduce a mass of changes. If I had my way though, I’d certainly want to see the results!
I would have thought any item which allowed heroes to jungle could be picked up on a actually jungling hero to increase their ability to do so? Any such item would have so many clauses it would probably massivly complicate things or push Jungling heroes into GPM machines. If it was a neutral creep damage reduction, Legionairre would go through the roof, and if it was a auto attack boost, Wildsould would go boom and if its all damage, zeph would excell.
I would be interested to see how this could turn out tho, the idea in theory sounds good 🙂
Very interesting article, as well as the older ones. I enjoyed reading all of them. I learned quite a few things 🙂
And I do agree with you, if you can somehow manage to make the game have different lanes viable, as well as allowing many different strategies (like pushing, which is almost dead for a long time now) it would make it much more interesting for playing and for watching.
Keep up the great work 😉
Well I love the whole discussion on the present metagame and the previous metagames that were followed. A comeletly different approach could be possible if the game were to be changed in such a way that the importance of each hero is buffed and doesnt reside only in the hands of a sure shot hard carry presently like madman, flint etc…New strats to become more competitive something like pulling of a 5v1 lane and going through to try and finish the game in 10 or lesser seems a distant strat in HoN…but it would be fun to see any of the competitive teams try it out.
Would it be complete impossible to balance an item, that whilst cheap, blocked pulling for a short amount of time in some way? Or perhaps encouraged fighting around the creep spots? Im thinking a non invisible, placeable item that can be countered without a money investment. Off the top of my head, something like:
‘Sentry’ 50g (Limited availability)
Places an imovable sentry at your posistion preventing neutrals from spawning for 300 seconds. All neutrals blocked in this way award 20?% XP in a 400? radius to the team that placed the sentry. A sentry takes / has 1000? hp (or maybe 20 auto attacks) to destroy.
Requires the placing team to be in range to stop it being destroyed, to gain xp, but requires the carry support to take time to destroy it, or support in lane.
I agree with you on many points and vehemently disagree with you on others. One of the big issues that drives negative trilaning is the dominance of carry play. I am a huge fan of trilaning and still see it as an edgy and risky strategy – when the intention isn’t to make it a safe farming lane, but to stick the fight to the enemy from the get go.
The problem here is that the farming trilane dominates the pushing trilane and the killer trilane. We simply don’t see trilane on trilane action anywhere near as much as we do in competetive DotA, and the reasons for it run deeper than the XP advantage offered by shortlane (which help to differentiate between the lanes in all strategies and are so much of a boon to strategic variety that nerfing them would be remiss, particularly since they give extra farm at the right point of the game – early to encourage more clashes early-mid). The primary reason, as I see it, is the overwhelming dominance of carry heroes in HoN play. Simply put, many carry heroes have capabilities they shouldn’t have and it discourages the incredible opportunity cost of running harder push and gank strats. As a continuation of this theme, there is little room to run a killer trilane to counter a farming trilane since the farmer in the trilane typically has excellent mechanisms for resisting the killer trilane. It’s little wonder, for example, that Madman doubles as a passable counter-trilane solo since the hero is so incredibly potent and flexible at all phases of the game.
At it’s core, I can’t help but feel that the trilane and farm metagame is at its core caused by the carries. I have frequently enough posted replies with this sentiment on Forum 13, but it is clearly just a venting point while the real input comes from SBT and competetive players. There are just so many nagging conceptual and numbers problems loitering with so many carry and semicarry heroes in the game that it is little wonder that while the particulars may change, the general story has been farm farm farm for the last year. Whether the farm was taken with a hardcore babysitter, from towers, or at the back of a trilane, the transition has always been into hardcore junglefarming behind a wardfield for >350GPM. The response was to introduce a new breed of midgame gankers who could deal with these conceptually overpowered carries, like Chipper and Deadwood, but in practice they are designed so that broadly they have some semicarry potential with the provision of scaling items for their abilities and we see them slaughtering supports and other gankers – it’s very rare to see well targetted ganks on carries these days because they money just isn’t in risky enemy orientated play. As competetive players comment again and again, it’s just much easier and safer to run a farming orientated strat than to have a strat that relies on taking the fight to the enemy’s jungle and their carry.
While there are a great many contributing factors to the trilane meta, I think the undesirable aspects of it are created by two factors – too much incentive to farm for too many heroes (“support” items, extensive and flexible item choices for gankers and semicarries), and the scaling problematic hardcarry concepts that are nigh ungankable midgame if they farm well early (like in a trilane for example).
There are three key negative aspects of the current metagame. The first is the dominance of farming trilanes which have basically held down killer trilanes – while I like trilaning and the roaming it often leads to, I dislike the safe farm environment they have become). The second is the non-existence of the gank phase and targetted ganks at the enemy carry – it’s simply too hard to gank these carries that have been added – something like twice as much of the HoN hero pool has blinds and pseudo-blinks compared to DotA, and a disproportionately high number of them are on carries. The solution is not to add gankers with these capabilities, because the power creep and arms race mentality has so many knock on effects that are unforseen down the line – this entire problem is caused by it. The gank phase simply doesn’t happen anymore because there’s no one realistic to gank besides the supports who don’t really matter, so ganking is used as a farming tool, not a defeat mechanism. This neatly leads to the final negative area – games go too long – far too long. Ideal game length is closer to 30 minutes than some of the competetive slogs we’ve been subjected to of late. For both fo the reasons above, teams farm far more and far longer than they should – buybacks, once the butt of competetive jokes, are now a staple lategame tool. The metagame, while healthy in terms of some kinds of diversity, is extremely sick in terms of some of the fundamental measures of success for the ongoing development of the genre.
So here’s the thing. I don’t mind trilanes, I don’t mind PvE, and I don’t mind the existence of viable carries. What I do mind is the unhealthy evolution of the role of ganking, the loss of the gank phase, the disincentives to run direct counters to farming trilanes, the arms race mentality that’s crept into hero/item design, the powerful incentives to accrue items on a wide variety of heroes instead of ganking, and the improperly considered development of underused heroes/items.
The driver for the entire metagame for a year now has been overpowered carries who’ve been left in place (as far as can be readily ascertained) because someone always has a cry about how few of XXX particular subset exist, and therefore how leaving YYY overpowered hero is actually good for the game.
If you want shorter games, more tactical variety in laning, more strategic variety in drafting, a more interactive game, and a better (as you put it) “lethality vs tacticality” balance, the rational solution is to begin addressing the fundamental drivers for the negative gameplay trends, which are the high carries most regularly behind the trilanes and wardfields, and the “farm up, item up” mentality that pervades the game.
Or drastically reduce the amount of gold available midgame in the jungle. I could write an essay on that particular topic which has been a negative influence on both HoN and DotA for at least 4 years now.
There are so many spelling and grammar mistakes in my post I nearly cried when I reread it.
I really think pzkw hit it on the head, ganking is just turned into another form of farming, with most gankers having some kind of built in semi-carry potential, just go out and hunt support. Just ask yourself what’s better for the team getting a guaranteed supportkill (and get closer to my portkey,spellshards, sheep stick)
or trying to put out a targeted gank on say swiftblade( who’s going to spin away) or puppet withhis double disable. Since everyone can semi carry there’s less of a risk letting it go late.
“Pulling ain HoN exists not because it was designed, but because it is an emergent mechanic that arose out of a clever abuse of an unintended interaction between lane creeps and neutral creeps.”
Are you serious?
All this is unintended and coincidental ????
Then why “pullable” camps are so close to the lane?
Why do they both have “gates” to the lane?
And how comes the timing for pulling is exactly the same on both ???
Excuse me Sir but this is bullshit.
Also, HoN is “inspired” from dota and so on isn’t it ?
Weren’t there players pulling in dota years ago ???
Anyway, old school power, I hate the “new school” !
I’m so sick of video-games today, cmon designers, keep your polish work for sunday car washing and give us PLAYERS what we need : GAMEPLAY (and yes I do like glitches).
I don’t effin care about trilaning just don’t reduce the possibilities jungle offers.
If you really want to get rid of trilaning supremacy then introduce more pushing items and heroes like you did with ganking, that way we can have a “metagame […] where multiple strategies can flourish simultaneously”.
Last but not least, about the “boring to watch” point… screw it, the focus should be kept on players not viewers! If you really want to please “stream” viewers then don’t give them streams but some kind of “delayed replays” where they can actively control the camera yet some shoutcaster will still comment and viewers can still stick to his camera.
Regards.
ZyV
Hi Nome,
I’m a student studying game design at the University of Technology in Sydney, Australia. I have been following your blog for some time and I think that your comparison between ‘new’ and ‘old’ school design is fascinating though can be at times be paradoxical. Allow me to explain.
As game designers I think we can agree that one of our most quintessential responsibilities is to be an advocate for, and design from, the perspective of the player. Taken in terms of your discourse on emergent gameplay, as a player you feel emergent gameplay is something to be cherished that has produced great innovations in gaming. Yet as a ‘new school’ designer you feel that emergent gameplay is the equivalent of flaws and something that detracts from, rather than adds to, the experience. How do you reconcile the contradiction between being an advocate for the player who wants emergent gameplay and a designer who wants tight control?
To further the conundrum you are a designer for a game that is to a significant extent based on emergent gameplay. The practice of pulling, as you note is an unintended mechanic which was the result of a complex interaction between the player and rules assigned to various objects within the game. In this case, the players ability to abuse the rules regarding the respawn timing of empty neutral camps, the aggression commitment time of neutrals towards the player when provoked and the pathing of faction creeps. Ultimately this unforeseen interaction has lead to the development of a recognised emergent gameplay mechanic that has been allowed to become part of game’s experience. Now according to the ‘old’ school of thought this emergent mechanic is a good thing while the ‘new’ school of game design would “always” consider this to detract from the game. I think classifying all emergent gameplay as bad, or conversely all abuse of unintended interactions as good is highly arbitrary and creates an unnecessarily restrictive design paradigm.
In reality, the fact that pulling originates from an emergent style of gameplay is non-consequential in determining whether or not the mechanic itself contributes positively or negatively to the players experience. Therefore one should consider the mechanic on its own merits in terms of the validity of the mechanic according to basic game design principles. To offer an example, a fundamental concept in game design is the idea that having many objects, properties and behaviours adds to the complexity and enriches the depth of experience of a game by making its outcome less predictable. Taken in the context of Heroes of Newerth, pulling adds to the game by increasing the uncertainty involved in resolving the games uneven outcome. This is because effective use of pulling or countering pulling allows skilled players an additional opportunity to exercise control over the outcome of the game.
Addressing your ultimate question of whether or not removing pulling is the right decision in adjusting to concerns about the current metagame is well beyond the scope of the point I am trying to make. However I think that when you do make the decision of whether or not to ‘sacrifice’ pulling, you should not hold bias against pulling as a ‘new’ school designer simply because it is an emergent mechanic.
And unusually well informed post. I’d reiterate something I said earlier in the thread in support of this:
“… reasons for it run deeper than the XP advantage offered by shortlane (which help to differentiate between the lanes in all strategies and are so much of a boon to strategic variety that nerfing them would be remiss, particularly since they give extra farm at the right point of the game – early to encourage more clashes early-mid).”
i agree with life