These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

EVE General Discussion

 
  • Topic is locked indefinitely.
 

Objective measurement for risk?

Author
Tippia
Sunshine and Lollipops
#81 - 2012-09-05 20:12:05 UTC
Mal Ishos wrote:
Too many words.
That's ok. I believe in you. Just try to push through and see if you can read it all the way to the end. It's kind of important that you understand the difference.
Telegram Sam
Sebiestor Tribe
Minmatar Republic
#82 - 2012-09-05 22:10:51 UTC
Telegram Sam wrote:
It seems that the risks in EVE can be divided into just a few broad categories:

-The risk that your ship will be destroyed. This can only happen to the ship you're in, not to a ship in your hangar. This can happen through PVE or PVP.
-The risk that your pod will be destroyed (causing the loss of a clone and your implants). This can only happen through consensual or non-consensual PVP.
-The risk that you will lose isk on the market (because of unexpected changes in buy/sell prices).
-The risk that you will lose isk or items to another player through a scam.
-The risk that you will lose isk or items to another player because of theft (intra-corp theft or ninja salvaging).

Any others?

No takers, eh? OK, I'll just quote myself. Like univ profs do. Smile

In theory, each of these broad categories of risk could be refined to take into account any number of variables. Example: A player is considering soloing the mission "World's Collide," Level 4, solo. The Risk vs. Reward (isk/hour) analysis model could factor in the capabilities of the player's toon and ship capabilities (DPS, EHP, targetting speed, targetting range, cap recharge rate, etc.) and compare those to similar factors for the mission "rooms" (overall DPS, placement, range and speed of the NPCs, warp scrambling abilities, etc.). That would yield the risk figure for "losing a ship to NPCs" category.

Then, the risk of being ganked by another player, killed as a wardec target, etc. while in the mission could be factored. That would alter the figure for "risk of losing a ship" and also yield a figure for "risk of losing a pod." Risk of losing isk on the market or being scammed should be 0 for this activity. If the player is interacting in those areas while on the mission, there would be separate analyses for those.

Then, to calculate the reward (isk/hour) for the activity, the model would take into account the isk for the mission reward and NPC bounties, and the estimated time required to complete the mission. (Looting and salvaging wrecks would be a separate isk/hour activity to be analyzed, with the risk of ninja salvaging or loot theft factored in).

Finally, the "intangible" factors of the player's experience, intelligence, and skill would have to be accounted for. (Here is where a lot of the real world risk:reward analyzers fail, I think. The model is based on hard mathematics, but the intangibles are fuzzy and have to be guessed at intuitively, or estimated on the basis of past performance. Which is a pretty inexact indicator, as anyone betting on horse races or sporting events would know).

In the end, some Risk vs. isk/hour ratio for the Level 4 mission should be produced. In theory, a similar analysis could be performed for any other activity in the game. So one could compare the relative Risk vs. isk/hour ratios of different activities.

(But note: The Level 4 mission example shows that there's a problem with defining Reward as just isk/hour. LP are also earned, but those could be counted as interchangeable with isk. But the example player may not care much about the isk. He may be grinding missions only to get higher standing with an NPC corp or faction. But I suppose standing gain and loss could be factored into Reward as well).
Mara Rinn
Cosmic Goo Convertor
#83 - 2012-09-05 23:14:40 UTC
Mal Ishos wrote:
subjective


You keep using that word. I do not think it means what you think it means.

Here's the rub: objective refers to statements without emotional bias. subjective deals with emotional responses to situations.

Here is an objective statement about exploration: in an evening, pilot X probed down 30 sites. From 27 of those sites, pilot X obtained no materials of value. From 3 sites, pilot X obtained materials worth approximately 300M ISK at Jita lowest sell prices.

Here is a subjective statement about exploration: I wasted two hours probing down a bunch of useless sites, but stone the crows if I didn't make bank on the last sites! Why is it always the last place you look?

What Tippia has been trying to tell you is that it is perfectly possible to objectively measure income from activities such as gambling (and thus, invention or exploration). These numbers will be expressed in terms of averages, with standard deviations, or other statistics-type jargon. One might say, for example, that the ISK/hr of mining is 11M ISK/hr mean with a standard deviation of 1M ISK/hr (thus implying that mining is a low income stream, but a very reliable one). One might also say, for example, that exploration has a 5M ISK/hr mean with a standard deviation of 50M ISK/hr (i.e.: it's just likely that you will waste an hour making no ISK, but you might end up making 300M ISK/hr). The statisticians will know the appropriate terms (probably things like "normal distribution" or "gamma distribution" and something to do with the difference between activities which involve chance for fixed reward per attempt, and activities that involve variable reward per attempt)

But in the end, it is very possible to have an objective measure of risk for certain activities. That measure would make most sense expressed in terms of ISK/hr. Thus the cost of failure is ISK, and the chance of failure must be measured in some proportion per hour. You get these units using dimensional analysis (which is simply the analysis of the dimensions used in a particular expression: ISK is a dimension, hours are a dimension, thus the dimensions of the equation to get a result in ISK/hr must involve X ISK and Y/hr).

As an example, I fly a tanked hulk. My risk is the cost of failure (let's call it 200M ISK for a replacement hulk) multiplied by the chance of failure (per hour). For a tanked hulk, that chance of failure per hour is going to be something like 0.1%/hr. That is, for all tanked hulks that were lost, they lived for an average of 1000 hours). For a max-yield hulk, that chance of failure[/] will be closer to 1/10hr (that's not 1/10th of an hour, that's a dimension expressing a mean of 1 event every 10 hours, so perhaps a more descriptive unit would be splodes/hr or failures/hr).

As another example, if I fly a mission running or plexing ship such as a Tengu, the cost of failure will be about 700M ISK (tengu, subsystems, Caldari Navy ballistic controls). The chance of failure per hour will be in the order of 1/100 fails/hr for the average nullsec ratter. The [i]chance of failure per hour
rises significantly if you fly that Tengu in hisec, and rises even faster if you fit deadspace or officer modules to that Tengu, and rises even further if you fly that offcer-fit Tengu in Osmon or Motsu (from second hand experience, the chance of failure per hour for an officer-fit Tengu in Osmon is about 1/48 fails/hr).

But what about measuring fun? Did you know that psychologists can actually measure how much fun you're having? Of course, they will measure it in terms of how excited you are, or how aroused you are (aroused is the clinical term for excited, which can include sexual arousal, but usually refers to adrenaline levels, pupil dilation, heart rate, or other physiological measures). Interestingly enough, how much fun you are having is not related to how much you enjoyed an experience, assuming "fun" is defined as "excitement induced for effort expended".

And the cruncher: we can also measure how much you enjoyed something.

These are all objective measures. They are all used by the gambling industry to tailor games to suit the players. This is why poker machines are so dangerous: they are specially designed to trick people into putting more money in. Different machines use different strategies, and the players will gravitate to the machine that gives them the rewards they feel that they deserve. The machine always makes a profit, and the gambler always ends up choosing the machine that extracts money from them at the optimal rate.

Your opinion that mission running is boring is a subjective measure. There is no scale you can measure your opinion against. How much fun you are having is an objective measure. The risk of an activity is an objective measure. The reward of an activity is an objective measure.

There is no magic left in the world.

Well, except for sunsets.
Mara Rinn
Cosmic Goo Convertor
#84 - 2012-09-05 23:19:33 UTC
Now here's a new word I want you to meet: variability.

You have stated multiple times in this thread that income is subjective. I think once you get to know variability, you'll realise that it's your new best friend, and subjective just got relegated to the "words that I used to know" bench.

Mission running income is variable, rather than subjective. Thus objective measurements of mission running income (or indeed, any other variable) take the form of statistical measures with uncertainties (i.e.: things like "means" accompanied by "standard deviations"). But statistics is a black art, and I do not grok it at all Sad
Kethry Avenger
Ministry of War
Amarr Empire
#85 - 2012-09-05 23:31:06 UTC
Individual Risks Only

Hi-Sec Risks

War Decs
Suicide Ganks- Revenged by Concord
Thieves
Ninja looters/salvagers
Can Flippers
Death by NPC
Market Losses

Low-Sec Risks

War Decs
Ganks (Some protection on gates and stations)
Thieves
Ninja looters/salvagers
Can Flippers (Theoretically they would most likely just shoot you)
Death by NPC
Market Losses
Cyno-Hot Drops

Null-Sec Risks

War Decs
Ganks - no protection
Thieves
Ninja looters/salvagers
Can Flippers (Theoretically they would most likely just shoot you)
Death by NPC
Market Losses
Bubbles
Cyno-Hot Drops
Bombs

This is just a start and not complete or broken down into all the possibilities. But even just enumerating the individual risks doesn't really capture the imbalance in risk versus reward on an individual player level. I say that cause if your an organization that can manage to hold a tech moon then the balance is different.

Wouldn't it just be easier to measure isk/gained versus isk/lost in each section of space because when it comes down to it your really trying to say that for the extra risks in lower security space and the loses that come with operating there your not being rewarded in proportion to the amount you would need to keep up with some alt in highsec over the same amount of time.
Tippia
Sunshine and Lollipops
#86 - 2012-09-05 23:58:11 UTC
Mara Rinn wrote:
As another example, if I fly a mission running or plexing ship such as a Tengu, the cost of failure will be about 700M ISK (tengu, subsystems, Caldari Navy ballistic controls). The chance of failure per hour will be in the order of 1/100 fails/hr for the average nullsec ratter. The chance of failure per hour rises significantly if you fly that Tengu in hisec, and rises even faster if you fit deadspace or officer modules to that Tengu, and rises even further if you fly that offcer-fit Tengu in Osmon or Motsu (from second hand experience, the chance of failure per hour for an officer-fit Tengu in Osmon is about 1/48 fails/hr).
…in fact, for T3s we can potentially do something ridiculously circular and fun. We an define risk in terms of rewards, and normalise the rewards to be our unit so its value is always 1. This gives us a relative measurement that is normalised for each player and each activity.

Why do that? Because if you lose a T3 you lose something else as well: time (or SP, which is much the same thing). How much is time worth? Well, we know how much you bacon you bring home with that reward figure and since we measure that in ISK/h, we have a worth for your time right there. So let's take that hypothetical Tengu and assume that you casually very lazily get 50M ISK/h with it. One hour is now worth 50M ISK. So when you lose that Tengu, your loss equates to 700M ISK = 14 hours worth of rewards. You also lose 78 hours of training time to get back that lvl V subsystem skill you also lost — a total of 92 hours. And then, of course, there's the opportunity cost: while you train back those lost SP, you're no longer getting the full 50M ISK/h, but rather only, say, 45M (10% less), so those three days of retraining also mean you miss out on, oh, let's call it 45M ISK (3h play time per day for 3 days at a reduced income of 5M ISK/h each) — another 0.9 reward-hours worth of loss. In total, the loss is worth 92.9 hours, and using the likelihood of 1/48h mentioned above, that gives us a total risk of 1.94 reward-units.

…in other words, the example Tengu in question is not worth the risk it generates. Its risk vs. reward ratio is 1.94:1 — for every hour played, you will statistically be set back 1.94 hours worth of ISK and training time — and you really want that ratio to be less than 1:1.

P
Herr Hammer Draken
#87 - 2012-09-06 00:01:28 UTC  |  Edited by: Herr Hammer Draken
Another measurement not considered in this risk/reward is fun vs boring.
Do you mine your own ore for production or is that too boring and you buy your minerals for production?
Do you pay extra high prices to buy stuff because it is right at your stations or do you go 7 jumps to save a few isk.

Fun vs boring plays at least as great a role if not even more so than risk/reward does.

What I am suggesting is that the reward for a lot of stuff is set by its isk value which in turn is determined by its fun vs boring level.

So fun vs boring trumps risk/reward.

I just had an odd thought. If we had women running the households in eve then they would be out shopping for the best deals they would be doing that 7 jumps for the best price all of the time. Oh my what would happen to trade values? I think EVE prices would be less than 10% from high to low all over eve every trading post would have similar prices. If one tried to raise prices the women would all shop elsewhere. Gender plays a huge role in eve prices!

Herr Hammer Draken "The Amarr Prophet"

Ryoken McKeon
Caldari Provisions
Caldari State
#88 - 2012-09-06 00:13:43 UTC
Mara Rinn wrote:
Mal Ishos wrote:
subjective

Awesome explanation from an overly caring soul.




I'm pretty sure that guy was trolling. Sorry :/
Roime
Mea Culpa.
Shadow Cartel
#89 - 2012-09-06 05:38:57 UTC  |  Edited by: Roime
This is actually fairly easy, risk related to PVE activities:

No risk: Empty local
Low risk: CONCORD protection and people in local
Medium risk: No CONCORD but people in local
High risk: No CONCORD and no local at all, empty dscan
Certain death: No CONCORD and no local, Proteus on dscan

(the middle three define the risk levels in relation to reward, first and last the limits of risk)

.

ugh zug
Aliastra
Gallente Federation
#90 - 2012-09-06 06:38:52 UTC  |  Edited by: ugh zug
losses > or = income
annoyances (new space) > or = annoyances (old space)
(new space) income < or = income (old space)
if true old space > new space
too much risk no profit
if false new space > old space
risk worth while.

Want me to shut up? Remove content from my post,1B. Remove my content from a thread I have started 2B.

Mara Rinn
Cosmic Goo Convertor
#91 - 2012-09-06 10:17:11 UTC
Ryoken McKeon wrote:
Mara Rinn wrote:
Mal Ishos wrote:
subjective

Awesome explanation from an overly caring soul.




I'm pretty sure that guy was trolling. Sorry :/


I'm pretty sure that Mal Ishos was just stuck on some mental rut. At worst if he was trolling I'm sure someone else reading this thread might have found my post has cleared up their thinking on "risk". At the very least, it helped me clear up my thinking, and thus was worth posting for my own benefit.
Tyrton
Imbecile MIiss Managment and Disasters
Intergalactic Interstellar Interns
#92 - 2012-09-06 12:02:45 UTC
Din Chao wrote:
I'm thinking someone doesn't know what "subjective" means.



or just like the sound of it .... either case did anyone do a word count for mal's use of the word "subjective"
Vincent Athena
Photosynth
#93 - 2012-09-06 17:06:03 UTC
On the subjective side of risk/reward, different people have a very different view of PvP. Some get a big reward from it, an euphoric feeling from the adrenaline released. That feeling can be so strong that they are willing to risk the entire contents of their wallet to get it.

For others that same exact experience boils down to a huge block of stress that leaves them feeling shaky and ill, even if they in fact "won" the PvP encounter. For these players, placing themselves in a situation where this can happen is a huge risk in and of itself, even if nothing happens.

For some adrenaline is a reward, for others, a risk.
On the other side, for some watching the ore pile up in the hold gives a reward, a gentle feeling of accomplishment. For others its a risk: boring tedium.

Given the above, most say risk/reward of in-game activities is all about in-game assets.

Know a Frozen fan? Check this out

Frozen fanfiction