no you see them, sure, but he's saying you still don't know where their base is, you have a rough idea of where they may be but say you know they aren't adjacent to you because they didn't find you immediately, you still have 3-4 locations to pick from, sure if they find you in less than a minute, you know they are adjacent, but if they take a bit longer, that means they have an advantage over you, because you have to check multiple locations far away, when they no longer have to spend time checking, and even if you know they are next to you, are they to the left or right? you don't know, flip a coin. Next they know the structure of your base before you know the structure of theirs, this is all do to the fact that they luckily found you first. They got an advantage due to luck. You can't argue that it isn't luck because it is. You rolled the dice wrong and they rolled them right for the early scouting advantage, end of story. Sure 1-5 minutes later their advantage will be mitigated by you scouting them out, but the fact that it was there is do to luck, once again, I stand by the fact that you don't know what luck is. Do I think they should go luck on everything? No, but luck has been a factor in every game out there. For example, an fps. You know the enemy might be in that general direction, but you don't know exactly where, you lob a grenade, you get a kill. That wasn't pure skill, there was luck involved, but it doesn't make it any less satisfying. The thing is that luck has to be implimented int he right ways. If you just put flat stats on everything and make the game static and calculatable fromt he beginning, as much as I hate to agree with innociv, it becomes boring and repetative, because you will know exactly what is statistically better in all situations. You might have 3 builds with 1 item variations at best since you don't have those, well what if I do this this game options to do every once in a while for fun. From a probability perspective, they odds of someone winning due to luck is very, very low, assuming the luck based trends in the game are designed low. You have luck from the getgo in ranked, MM is known as luck, there's something for you to chew on, but no, statistically if you play 2000+ games you will find your true MMR, which is hidden in league, assuming it never shifted during those games. Luck is embeded into video games and has added to their playability and function for years. I will agree with you that it does feel bad to lose to luck, but when it happens 1 in 1000 games, it doesn't bother me that much.
Originally Posted by Kreamsoda
Edit: I do agree that luck shouldn't be the dominant focus of the devs, but it shouldn't just not exist.
Last edited by Haarenfang; 06-30-2013 at 03:02 AM.
It makes the game more interesting!
The luck factor also allows for comebacks in the game for example if one team just gets a parasite advantage then they can just utilize their gold advantage and theoretically not make any mistakes and win the game. Luck factor helps to make the game more fun instead of just statistical nonsense.
Are you seriously that ignorant? There a much more elements to it than that.
Originally Posted by Kreamsoda
Reacting to this just means you played into the situation. If that luck factor wasn't there it wouldn't of forced you to make that choice therefore bringing another element into the game.
Last edited by MazeDota; 06-30-2013 at 04:24 PM.
There's also the fact that if you can see the luck element, say it is on an item, you can counter it by building something that negates it, which in turn means a wider variety of items, a wider variety of possible abilities for characters, a wider variety for spells. So luck elements should never be on perks as you can not clearly see what perk the enemy has.
I'm incredibly surprised and relieved to see how many people here actually understand how some luck plays an important part in most competitive games.
GlacielA's point as well.
Last time I played Lycan, I got really unlucky never bashing and always missing on two people that had evasion. In the end I realized I probably would have won that gave if I had gotten truestrike, but I made a mistake in not thinking about that at the time.
I could have blamed luck, or my team, but it was all my own fault.
I've said all along that all sports should have a RNG. Sports are so boring without randomness outside of the players control.
"The referee has rolled a 20 and the following series will have 5 downs." Sweet!
There's two big discussion points in this thread I want to provide some of our thoughts on. Why the Hunter role specifically uses a random aspect, and our philosophy on randomness in a competitive experience in general.
The first implementation of the role had this effect (a small guaranteed amount of Vim for every creature killed). Ultimately, we ended up switching from a guaranteed bonus on every kill to a chance for a larger bonus. Here's a little bit of a viewpoint into our reasoning.
Originally Posted by Kreamsoda
Why isn't the Hunter role's vim bonus a consistent boost to every creature's bounty?
Having the effect be a consistent, always present, effect felt fine for Hunters, but it made anyone who wasn't using the Hunter role feel really bad about killing jungle creatures. When it was a modification to every creature's bounty that could be counted on 100% of the time we found that player's began to internalize the wealth of a creature as the combined sum, i.e. for small Mushrooms instead of being worth say 4, they were instead considered to be worth 9, with the bonus being factored in. This meant that when anyone else went to kill a creature, such as someone from a lane while they were waiting in-between waves, or a jungler using the Predator role, they felt penalized or that it wasn't worth it. Instead of being seen only as a bonus for the Hunter role, it was instead seen as a penalty for the other roles whenever they were in the jungle quadrant.
That's great and all, but how does that point relate to the Hunter role?
Similar to the above point, the fact that people began to internalize the bonus bounty when it was on every kill as the base bounty, it meant that it was no longer perceived a bonus, and as a result the Hunter's didn't feel like they were being rewarded, they were simply getting the perceived baseline, with other roles getting less than that.
The difference here is subtle but important. A great example of this physiological effect in game design in a much more well known system is the system of 'Rested Experience' in World of Warcraft. In very, very early beta for WoW the developers needed to slow down the rate that players leveled to prevent people from exhausting content too quickly. They decided to solve this by making it so that when players first logged in, they gained 100% experience, but over time as they continued to play in one session, they would gradually occur a debuff that reduced their experience gain down to 50%. Players hated it. The response from the beta testing community was over-overwhelmingly negative. Why were they arbitrary debuffing player experience rates just for enjoying the game and playing it for long periods of time!? So, blizzard simply doubled all of the numbers. The amount of experience needed to level was doubled. The amount of experience you received when you first logged on was doubled from 100% to 200%, the amount of experience you received when you had been playing for a bit was doubled from 50% to 100%. The system is identical mathematically, literally nothing had been changed. The only difference was that players now perceived it as being rewarded instead of being punished, and they loved it. In fact, it was so well received that pretty much every single MMO since WoW has incorporated the rested experience system.
We needed to do a similar presentation shift with the Hunter role's bonus, so that it felt like a bonus for the Hunter role, without feeling like a penalty for people who didn't take the Hunter role but still killed jungle creatures. Making all of the roles feel like bonuses is very important, since otherwise (and we experienced this a lot as we were testing different roles internally) it caused a since of claustrophobia, where players felt heavily punished for participating in sections of the game outside of their designated 'area of play' dictated by their role. This is at the core of all of our Roles now, so that if a Gladiator is buying, Tacticians don't feel bad about taking last hits, or in a team fight a Gladiator doesn't feel bad about taking kills away from a Predator, or a Predator doesn't feel penalized for clearing creatures in the jungle, etc.
The random chance, as you pointed out, is systemically equivalent to distributing the bonus across every creature, but it keeps it feels like you're getting greater than 100% of the bounty to the Hunter, whereas other roles still feel perfectly content about only getting 100%, since it is the full amount of the perceived base bounty.
This reminds me of another thing I've been seeing recently, which is:
Originally Posted by Kreamsoda
Surely there were other implementations that accomplished this goal other than randomness, what about every 4th creature gives the bonus instead of a 25% chance?
It's absolutely true we could have solved the perception issue in a non-random way, but we ultimately decided to go with a random chance on the role because the chance for a 'jackpot' to be won is quite fun. When we were experimenting with the Mastery stat it initially was going to given basic attacks a consistent percentage increase, similar to Spell Overload, so that with say a 33% crit chance you would hit for 133/133/133 instead of 100/100/200. What we found is people simply had less fun with it. They missed the chance for seeing a really big number and missed the fun feeling of 'maybe just the next hit' will do something big and awesome. It's not a great analogy, but I think of it as what's more fun, pulling the lever on a slot machine, or investing that $20 into a savings account? Chance of big returns vs consistent small returns. The random nature of it is simply quite fun, as long as it doesn't create some of the negative aspects that RNG can create, which leads me to my last point.
What about the downsides of randomness?
Randomness certainly can have really negative aspects to competitive play. Generally the way I look at it is that if the outcome of a single random event is extremely impactful it is probably ill-advised for a competitive game. An example of this is Madred's from LoL where the difference in that proc occuring or not sometimes meant if you were going to kill Dragon or a buff objective or not. That's a lot riding on a single proc chance, and I agree with their decision to remove the random nature from that item. Another example where I'm not a huge fan of it is the basher from DoTA, where the difference in a proc or not can easily mean the difference in killing an enemy Hero or not. Low amount of sample sizes, with really high returns, is a form of randomness I would like to try to avoid, generally speaking.
Places where I feel randomness is alright is in areas where the number of sample sizes are very high such that the statistical nature of randomness wins out. The three areas where randomness is used in Dawngate for example are all areas with high sample sizes: critical strikes on basic attacks (number of auto attacks fired), the health return on the Consume passive (things near you dieing), and the vim bonus on the Hunter role (the number of creatures killed).
In conjunction with the above, Dawngate does not use a purely random system. We use a psuedo-random system that removes the possibility of getting unlucky while retaining the ability to get lucky. Every single time a proc does not occur, the chance that the proc will occur on the next attempt is increased until eventually it reaches 100%. This helps, along with large sample sizes, to further mitigate the random nature while retaining the fun aspect to it where players can go "well I just got a proc on that small mushroom, maybe this next small one will proc it too!".
I hope your first hit chance is still the stated amount in your PRNG.
I'm not a fan of Blizzards PRNG implementation, where something that states 25% chance is actually like 8.5% chance on the first hit.
This is one I was playing with a long time ago, but since the "memory" of it starts when you attack some target, it averages out worse over hundreds of hits despite how good it is on a small sample size. Though, that wouldn't happen in a moba, it still is less ideal than it could be and I was meaning to work on a better one.
I wanted something that works more like how a child understands it. "25% chance means one in four" sounds like it should happen every 4 times. But you also need a 1 in 4 chance on the first hit. So it keeps at 25% chance, but then adjusts for guaranteed nth hit procs, but double procs still occasionally happen.
I had been meaning to make one that'd work for a more general thing, perhaps with a short "memory" of the past few cases, but haven't gotten around to it.
Anyway, when it comes to basher and monkey king bar, one reason those are good is that people with them wouldn't even bother to attack someone who is teleporting out if they didn't have a chance to stop them.
If there are two people, and one is running away and further away, and another is teleporting out, it makes you quickly decide whether you try your luck at getting a bash on the person TP'ing, or if you go and chase the other person.
The person who TP'd out was also betting their luck on not getting bashed instead of running further. If there was no such thing, there would have been no risk-vs-reward decision to make in the first place. TP'ing or not would have been a clear choice.
If that random chance didn't exist, the decision would have been clear and easy as can be. Luck like that can make a game FAR harder, but only unfair if it's something really silly like "1% chance to instantly kill".
(some bashes in dota also have cooldowns so you can't perma-bash)
I definitely don't thing basher needs to be in DG. A 1.4s stun like that wouldn't fit. But I don't think something similar is out of place as long as the effect isn't too huge and the chance isn't below 25%. Like a chance that a non-ultimate ability that already has a low cooldown instantly cools down, for example. When that happens, the player has to react to use the instantly cooled down ability.
Last edited by innociv; 06-30-2013 at 07:43 PM.
Everybody has luck, sadly someone has to have the bad side.
I remember reading an interview with TSM Reginald I think it was about something similar, and if I recall correctly he was basically saying 'It all came down to who's Phage proc'd first' - which came down to luck.
If you have two shapers shooting each other at the time with the same health, armour, DPS, items etc theory says that you should both die at the same time. However with a % to crit thrown into the mix you don't know who will win; hell it could even be a stalemate again if no one is lucky enough to get a crit.
What I've noticed, especially with Mobas is that you create your own luck. The amount of times I've thrown a blind random Pudge hook or a Blitzcrank hook and been successful is unreal. So I'm increasing my luck by doing lucky things, if I miss...meh bad luck! But all those things count in the long run.
But to my point, you will probably find you have had more 'lucky' moments then 'unlucky' moments without even realising it. And to highlight a point Innociv said -the amount of times I've played as Sniper or Spirit Breaker and never seem to stun is unreal BUT when I play against them I seem to get stunned every other hit; its just one of those things but it shouldn't be enough to deter you from a game.