|
For reference, latest patch has gold decreasing from creep kills and increasing from hero kills.
The general effect of lowering gold income is an excellent thing which the game needed. We've been getting far too many ways of getting gold lately, and that's been messing things up quite a lot. If money stops being a resource (read: a limiting factor), the only real resources left in the game are Roshan and buybacks, and in some of the games we've seen recently, not even Roshan is of much concern. Gold needed to go down, and gold going down was a very good thing.
However, this change only fixes half of the problem. Still remaining is the other half, which is the question of where the gold comes from. In the most traditional form of the game, gold came almost entirely from creep kills. In our current iterations, gold is coming in a much higher from hero kills. These two different sources have a major effect on how the game is played.
To demonstrate the effect, let's run a thought experiment: what would games be like at either extreme? At the one end, where every piece of gold is from a creep kill, we have a situation where players will only stop farming to take necessary objectives: for example, towers controlling critical farming locations, Roshan, and barracks. Kills will be scarce and focused entirely on important cores for the sole purpose of taking them off the map so that they can't farm. Supports will be starved. We have seen the game look very similar to this in the past, in fact. Now, on the other hand, it is possible that we could have farming yield no gold whatsoever. In this situation, the only reasonable occupation for a hero is to roam around the map looking for kills, and look they will. We can theorize about whether this will lead to a level 1 fight at the Rosh pit every game, but the fact remains that all that people can be expected to do is fight. In summary: rewards for doing something incentivizes doing that thing.
To people who enjoy fighting, this may seem to be a good thing. The response is clearly "Why not?" And indeed, there is such a high amount of support for fighting that it seems the most reasonable option to incentivize it. The problem is, however, that killing heroes is inherently incentivized while killing creeps is not. In the thought experiment above, hero kills had obvious value in limiting what the enemy could manage to farm, but in a situation where all gold comes from hero kills, who would ever realistically try to kill creeps? The only purpose for that would be pushing out lanes as an irritating bit of prep work before pushing. That is to say, one of the fundamental features of the game would become nothing more than a distraction.
This is a problem because it turns the game more and more into a "hero arena" style of game. Those games are certainly fun, but they fail to capture the imagination as DotA Allstars did, but why? In a term: resource management. Creep kills and farm priority are issues of resource management, but kills (with or without assist gold) are chaotic things which cannot be effectively managed or planned. This, of course, has other insidious faults built into it: hero composition becomes less important, because the question of which hero gets what farm is harder to predict, for example. By losing resource management, we lose a huge part of what makes the game so fascinating and engaging.
I won't pretend that this choice was made entirely out of caprice. In fact, I believe the exact opposite: it had a very solid and consistent purpose of trying to give the miserable supports in pubs some help. However, this was not the only way. Affecting cores with mechanics intended for supports is, obviously, a big problem. Solutions might include, for example, hero kill gold bounty that is split according to an assist and scales to the difference of net worth between heroes, or perhaps to have some kind of gold timer grant "potential gold" that's added onto creep bounty so that supports can farm their meager amounts quicker (and get everyone used to supports killing creeps). It might even have been worthwhile to change some of the rules for lower level pub games.
However, the end result is to shift the game away from one of its core mechanics and lose part of what made the game so special. It's a huge loss at that, and not one to be taken lightly. If you're unconvinced that it's a problem, then I recommend you go play the Warcraft 3 custom map Battleships Pro on No Traders mode. In that mode, there is no worthwhile source of income besides kills, and the result is a low-strategy tactics-only slugfest that has the additional problem of snowballing. Traders mode, where there is a distinct farming method, lacks all of those downsides but requires some strategic skill. It's a perfect case study for my argument.
To preclude a criticism that I anticipate: a lot of you will say that focusing on creep kills instead of hero kills will unfairly advantage heroes like Antimage over heroes like Slardar. This is entirely true and is why Antimage should be nerfed and Slardar buffed alongside the change I'm advocating. There will be teething problems for sure, but restoring income sources to what they ought to be will allow the game to be better adjusted by simply modifying the heroes. This is not to say that hero kills should be worth no gold at all, but simply that the relative values should be nudged back in the favor of creeps.
Hope this is convincing to some of you.
|
Resource management is lost in your thought experiment of creep kills giving absolutely no gold. In the current form of Dota, this is not true and I doubt that would ever be the case. The 6.84 just changes the risk/reward ratio. It reduces the reward of the low risk activity of killing neutrals. It doesn't eliminate this activity. As long as you can gain something, it will be advantageous to do it in certain scenarios.
Resource management is very much part of the game and you are undervaluing Creep equilibrium management if you think people will ignore creeps. The problem with too much gold was mostly spurred by incomes sources that require no additional action namely passive gold increase and the comeback mechanic (which hopefully will be ultimately removed).
Considering the latter portion of your post, it seems that you are viewing the game at a very shallow angle with low level assumptions. Supports kills creeps very often in higher level pub games. The changes aren't made to "help supports" in pubs. Anti-Mage and Slarder are distinctly different concepts and should in no way be balanced in the way you have suggested.
Kills are part of resource management because there is also the resource of time. It is a riskier activity but it can still be planned and managed.
|
This is a problem because it turns the game more and more into a "hero arena" style of game. Those games are certainly fun, but they fail to capture the imagination as DotA Allstars did, but why? In a term: resource management. Creep kills and farm priority are issues of resource management, but kills (with or without assist gold) are chaotic things which cannot be effectively managed or planned. This, of course, has other insidious faults built into it: hero composition becomes less important, because the question of which hero gets what farm is harder to predict, for example. By losing resource management, we lose a huge part of what makes the game so fascinating and engaging.
Now that i think about it .. i pretty much agree
|
On April 28 2015 22:47 Reson wrote:
Resource management is lost in your thought experiment of creep kills giving absolutely no gold. In the current form of Dota, this is not true and I doubt that would ever be the case. The 6.84 just changes the risk/reward ratio. It reduces the reward of the low risk activity of killing neutrals. It doesn't eliminate this activity. As long as you can gain something, it will be advantageous to do it in certain scenarios.
This was my point, and this is the function of a thought experiment. I am not saying that the game is like this now, but simply demonstrating how changing the mechanic affects what strategies are effective.
Resource management is very much part of the game and you are undervaluing Creep equilibrium management if you think people will ignore creeps. The problem with too much gold was mostly spurred by incomes sources that require no additional action namely passive gold increase and the comeback mechanic (which hopefully will be ultimately removed).
Creep equilibrium is maintained for the sole sake of denying farm to the enemy offlane. In my imaginary world of no-gold, there is no reason to do this, as there is no farm to be denied. In the current state of the game, pre- and post-patch, there is still an excellent reason to maintain equilibrium, but the importance of it declines as creep farm goes down. If you're talking about just pushing lanes out, then I agree that it will still be important, but the reason for pushing lanes out becomes less about farming and more about backdoor regeneration as creeps give less gold. I definitely agree with your statement about passive gold and the comeback mechanic, but they're a different issue that I might address in another blog.
Considering the latter portion of your post, it seems that you are viewing the game at a very shallow angle with low level assumptions. Supports kills creeps very often in higher level pub games. The changes aren't made to "help supports" in pubs. Anti-Mage and Slarder are distinctly different concepts and should in no way be balanced in the way you have suggested.
Higher level pub games approach professional games in playstyle and were not the target of the changes that we've seen coming since after TI3. Those changes were all made with the goal of improving support play for people in lower brackets (say, 3k and below). I think you misunderstood what I was saying in that part of the blog, and recommend you re-read it while considering that Icefrog is balancing for all players and that even just going above 4k MMR brings you into a very small minority of the total playerbase.
Kills are part of resource management because there is also the resource of time. It is a riskier activity but it can still be planned and managed.
This argument doesn't follow, actually. We say that time is a resource. Very well, let's say it is. If time is a resource (a limiting factor), then there need to be multiple things we wish to do that we don't have enough time to accomplish entirely. What activities are there in DOTA2? You can kill heroes, of course. You can take towers and Roshan. You can ward, but that is a brief occupation and is hardly worth mentioning. What is there besides that? Nothing but farming. Farming is the largest investment of time for most heroes by far. When you say that time is a resource, what you're really saying is that farming is important and that you don't want to take time off of it. If farming is not important, then the only resource management involved in getting kills is deciding who to kill. The resource management aspect of going for kills only makes sense in the context of farming.
|
On April 28 2015 23:33 Acritter wrote:Show nested quote +Kills are part of resource management because there is also the resource of time. It is a riskier activity but it can still be planned and managed. This argument doesn't follow, actually. We say that time is a resource. Very well, let's say it is. If time is a resource (a limiting factor), then there need to be multiple things we wish to do that we don't have enough time to accomplish entirely. What activities are there in DOTA2? You can kill heroes, of course. You can take towers and Roshan. You can ward, but that is a brief occupation and is hardly worth mentioning. What is there besides that? Nothing but farming. Farming is the largest investment of time for most heroes by far. When you say that time is a resource, what you're really saying is that farming is important and that you don't want to take time off of it. If farming is not important, then the only resource management involved in getting kills is deciding who to kill. The resource management aspect of going for kills only makes sense in the context of farming.
That's not quite true, one aspect you are missing from time management is defending objectives. If you spend a bunch of time trying to kill one hero and start losing towers/rax/ancient in the meantime you might not be using your time resource correctly.
I still see what you are trying to say though, but personally I don't think it will be a huge issue, since kills aren't guaranteed and it's not like creep gold was decreased massively, resource management wont be gone it'll just be changed slightly.
|
I mean we need to compete with Heroes of the storm so i dont see the purpose icefrog just realizes what he needs to do to stay relevant.
|
Creep equilibrium makes you able to farm safe, close to your tower but outside its attack range. It's not just to deny offlaners.
|
They reduced creep gold by ~7% and boosted hero kill gold by 10%. If you think that'll have a big negative impact that's fine, but bringing up an old WC3 custom map where "creep" gold was totally eliminated isn't convincing me at all. He's making tweaks. The thing that made rubberband so impactful was the fact that certain bounties were getting changed to huge degrees, like 5X the amount of the original bounty.
|
Income needed to be reduced, he's giving it a shot this patch. We needed a step towards heroes who are very impactful with limited items opposed to those who make multiple items synergise (high dmg item + high AS item + lifesteal etc).
|
Reducing income has the indirect effect of prolonging the early game, because your first big farming or snowballing item is slightly further away as a core. Allocating kill bounties that scale inversely with the amount of farm you have, prolongs your early game (as a core) even further, because both your supports and enemy supports get far more bounty from hero kills while the bounty a carry receives is gimped by 25%.
This is indirectly a buff to heroes like Alch and Doom who are able to accelerate their farm much faster. And of course the position 5 supports.
I quite like the change, as it will increase the emphasis on stronger, more offensive supports;.
|
Let me first say that there are some people who are not addressing the thesis of this blog. The thesis is fairly straightforward: it is the title of the blog minus the "why." The change in question is what spurred me to write out this argument, and in the first paragraph I say that I appreciate much of the intent behind the change. I am simply presenting a secondary issue which I have seen become more prevalent in the past two years, and which I believe is appropriate to bring up for discussion here.
The question of offensive supports is a very good one to bring up, and I'll devote a bit of time to it. I'll start with a simple premise: the nature of every role in the game, including support roles, is determined by risk versus reward. If the ratio of risk to reward is appealing enough, the role will take on a given task. If not, they won't. Now, aggression is typically a high-risk task in DOTA2. You have to get to the lane in question undetected and then execute properly without suffering losses. One way of encouraging aggressive support play is definitely to incentivize it highly with huge rewards, like we have seen Icefrog trying to do in this patch, but that isn't the only way. Another method is to simply buff the aggressive potential of supports and nerf the defensive potential of mids and offlaners. This lowers the risk of aggression and incentivizes the behavior without having to mess with the fundamental mechanics of the game. I am a strong proponent of this style of hero-based rather than mechanics-based balancing, I should note, but that's a different discussion. I simply wish to demonstrate that it's not just a binary choice between the change that Icefrog made and forcing us to have passive supports.
|
Dota income still comes from creeps. There is an overall larger amount of gold available at any given point in the game due to the rubberband mechanic (which was nerfed). Pros are better at stacking neutrals than they were before. Many farming abilities are more spammable than before (axe, tide, lina). Nerfing neutral farming is a nerf to certain heroes, but a buff to others. You can't really ignore the experience portion of creep equilibrium.
This patch does not turn dota into a hero arena type game. Creeps will still provide the majority of farm.
|
On April 29 2015 11:20 Bigtony wrote:
Dota income still comes from creeps. There is an overall larger amount of gold available at any given point in the game due to the rubberband mechanic (which was nerfed). Pros are better at stacking neutrals than they were before. Many farming abilities are more spammable than before (axe, tide, lina). Nerfing neutral farming is a nerf to certain heroes, but a buff to others. You can't really ignore the experience portion of creep equilibrium.
This patch does not turn dota into a hero arena type game. Creeps will still provide the majority of farm.
Nothing you say refutes my thesis, because my thesis was not that this patch was the end or something like that. I'm simply bringing a concept to people's attention that I have never seen expressed before and which I think is incredibly important to introduce into their understanding of how the game functions at its core.
|
Your thesis is already true - the main source of gold in DOTA2 is creeps and will continue to be creeps ins 6.84. Most people who play dota already agree with you that last hitting and denying are critical parts of distinguishing DOTA from other similar games. At least I definitely agree with you.
values should be nudged back in the favor of creeps.
I don't agree with this point though. Creep farming is incredibly strong and is one of the reasons certain heroes are significantly weaker in 6.83 (especially in terms of supports). 6.84 plays with the risk/reward of farming or ganking. Farming is still better for making money and getting EXP.
|
All right, it's good to see we are in agreement on the main principle. I'd just expand it a little bit beyond last hitting and denying: it's also about when heroes farm and where. There was definitely a revolution in terms of map exploitation thanks to players such as EternalEnvy and teams such as Alliance, but I don't buy that it's made it necessary to enforce all these adjustments that we've seen since after TI4. As I said in an earlier post, I find the best solution is simply to rebalance heroes. The crazy TI4 push meta was mostly because the bulky independent cores were both overbuffed and balanced for the pre-TI3 laning setup, for example, and the nightmare scenario with the TI3 midlane was mostly because OD was far too strong (and happened to have one extremely hard counter and one pretty great counter).
I think there's one further concept I'd like to introduce, although it's getting a little too in-depth for the original blog. The primary concern is not so much where the majority of a team's farm comes from as where the majority of the farm that the enemy team is not getting comes from. To express it more clearly, imagine that each team is farming their midlane and safelane at a 100% rate. That is, every last hit coming in. They are currently even. Let's say one team starts farming their offlane as well at whatever rate: all of their farm advantage is coming from effective use of the offlane. Let's say it's the jungle instead: their advantage is from use of the jungle. Let's say it's from towers instead: their advantage is from pushing. From kills: from fighting. The math gets a lot more complicated when you look at a real game where every resource is being exploited inefficiently and one team is exploiting one better than the other, but I think you get the idea. Now, different sources of farm advantage indicate different game states. The simplest example might be a difference in midlane farm, which quite clearly indicates that one mid is beating the other mid (whether or not that's thanks to support rotations). Difference in tower bounty indicates that one team is better at taking objectives than the other (thanks to draft, play, or both). These are very straightforward. Lane creeps and jungle are a little more complicated, but they generally show a draft designed to take advantage of the jungle and ancients as well as good map movement. Kill gold advantage, however, is the most complicated, because it indicates several things which are very hard to disentangle. It can indicate that one team is better at ganking, better at teamfighting, better at staying in safe areas, and possibly even just better at being lucky. The reason all of these are intertwined is that kills happen as a natural consequence of every other event which takes place in the game. The enemy players are trying to accomplish something, and quite obviously you wish for them to not accomplish it. The solution is through direct interference, which will quite often result in kills. In other words, people are rarely just trying to get kills (except in pubs and with certain inexplicable players). They are trying to accomplish another objective, such as slowing down farm or making a key pickoff before pushing, and the process necessitates killing. This confusion is an excellent reason why income based on kills should be limited to simply pay for the process so far as offering gold to the two supports is concerned. It's the same as tower and Roshan bounties: they should simply be enough to pay for some of the time the heroes spent in going to the objective, because the objective itself has value outside of the gold. Problems tend to arise when you have players aiming for objectives as a major source of income, like with the nightmare at TI4. To conclude: gaining advantage over the other team through other sources tends to show some kind of strategic advantage, while gaining advantage through kills can mean anything.
Oh, it's worth noting as well that the entire issue with heroes taking too much from the map is partially a legacy from the post-TI3 offlane changes where it became far too easy to farm the offlane and partially thanks to heroes with the potential to farm the jungle and ancients being extremely strong in 6.83. If that change (which I derided at the time and which lent itself to the ugliness of TI4) were reversed and the farming heroes nerfed so that there was a real decision of whether or not to invest in that kind of income (just as it is and was a serious question of whether to invest in an Antimage pick), you wouldn't see both teams getting such stupid lane and jungle efficiency.
|
I agree with most of what you say.
I think a big point is the "resources", in this case GOLD being less and less scarce, making it less valuable, and when you make creeps give less gold you are also turning the game to the position you want it to go.
I'd actually like a comparison of NW per team from idk, TI1, TI2, TI3 TI4 and TI5, at given times.
|
Did you take into account that there'll be more lane creeps overall to farm than before? "Extra melee creeps additions spawn time changed from 17:30/34:00/50:30 to 15:00/30:00/45:00".
|
On April 28 2015 22:11 Acritter wrote:
In a term: resource management. Creep kills and farm priority are issues of resource management, but kills (with or without assist gold) are chaotic things which cannot be effectively managed or planned. This, of course, has other insidious faults built into it: hero composition becomes less important, because the question of which hero gets what farm is harder to predict, for example. By losing resource management, we lose a huge part of what makes the game so fascinating and engaging.
I completely agree with this. However, I think it is obvious that we are not losing this kind of resource management from the game with this patch, the changes are way too minor for that. And personally I don't fear that we will lose it in the nearby future, if ever. So while I think you make a very good point why creep bounty should not be removed, I think that the reason many of the above posters disagree with you is that in the context of this patch, your argument is more or less irrelevant  Still, if I at any point need to convince someone that creep bounty is important, I'll refer them here^^
|
|
|
|
|
|
EPT
