Ronin

(The fact that your Valor head piece has not dropped is) Not as improbable as it seems. You have a 1:8 chance of seeing it no matter how many times you run the instance. That doesn’t mean you’re going to see it at least once in 8 tries, just that there is 1 chance in 8 for you to see it each run. You may never get it. As someone who has run Scholo as many (if not more) times than you, I understand your pain but urge you to keep it in perspective.

It does, however, remind me of the coin-flip scene in Rozencrantz & Guildenstern Are Dead.

I am aware of the statistical probabilities, although I do not claim to understand even the fundamental mechanics of probability theory. As an aside, I don’t think that my hat having a 1 in 8 chance of dropping is appropriately stated as 1:8 odds, because a coin toss is 1 in 2 and is expressed as 1:1 odds. You read this and tell me what the odds are, as your math comprehension exceeds mine: http://en.wikipedia.org/wiki/Probability.

I have run Scholo over 20 times now, and over half of those included a shot at the Hearald. All told, my chances of seeing valor boots or a helm have been closer to 1 in 6, which makes the fact that I have run it 20 times without success improbable. Of course, it is possible to run it forever without seeing your drop, but all I am saying is that I have exceeded the probability curve on this one. If I ran six times and something dropped, that would be typical; twelve runs and one drop would be improbable, but not highly improbable; eighteen runs and one drop would be highly improbable. I have run twenty with no drops, and I stand by my statement that it is highly improbable.

You’re right. The probability should be expressed as 1:7.

Based on your number of runs you would think that the chances of it dropping have increased, but they haven’t. The problem is that human experience is tied to time and therefore linear, which makes statistics tough for people. We hear “1 chance in 8” and linear perception says that if we do it 8 times, it should happen once. That, however, is fallacy. You have done the run 20 times, with odds of 1:7 each time. Cumulative total = 20:140 or 1:7. You’re no more likely to have the drop you’re looking for occur in 300 runs than in one. You should find, however, that the more times you run the instance the closer you will get to a true statistical 1:7. I wouldn’t expect to really be able to see the trend until you’ve run the instance 100 or so times (I would expect to see valor drop around 12 times in that 100 runs).

Trending over so few attempts (i.e. 20) is akin to trending a coin toss with 4 flips. “Heads is 3 times more likely to occur than tails, according to my research”. You sort of need to flip that coin several hundred times to see any sort of distribution. I would find it highly improbable if you had seen any single piece drop 10 times out of the 20 that you’ve run. I’ve seen a few Beaststalker, a couple of Shadowcraft, Magisters, and one Devout. No Lightforge or Valor yet, though. But hey, I’ve only run it 10 or 15 times...

You don’t seriously think I don’t understand that linear fallacy, do you? The point is that a 1 in 8 probability is a 1 in 8 probability. It is improbable for that not to occur, and the more times you run without the occurrence, the less probable the outcome. Another way to state it is that the chance of running scholo 20 times without seeing a valor drop, given the two random occurrences, is something like 1 in 30? I think that 1 in 30 compared to 1 in 8 is “highly improbable.”

You can’t combine those 2 random events and think that it changes the statistical probability. They are independent. What it comes down to is that you have rolled an 8-sided die 20 times and have yet to see an 8. That isn’t weird. Adding a second 8-sider and not getting an 8 on that one either doesn’t make it any weirder, just more frustrating. You will eventually roll an 8 (probably) on one of the 2 dice. Until that time, I expect you to be totally intolerable.

For the record, I haven’t seen the LF helmet drop at all but people complain on Thottbot that it is the only thing that EVER drops for their group (probably all hunters and druids).

Of course you can combine the two probabilities. As in any statistical problem, the model ultimately defines the answer. I can set up the model to state that there are two random events, each of which will occur in succession for each trial. The chance of each random event happening in a certain way are 1 in 8 and 1 in 14 respectively (boots have a lower drop rate). Therefore, the likelihood of performing the trial 20 times without either event occurring a single time is something like 1 in 30. Each time the random event is tested, the chances are 1 in 8 or 1 in 14, but that is not how I have set up the model; that test is boring. I am measuring the likelihood of one specific overall outcome from a series of trials where each event occurs. The answer to this problem under this model is that the result I have randomly generated (20 trials; 20 events at 1 in 8; 12 events at 1 in 14; zero valor occurrences) is highly improbable.

The chances of flipping a coin four times and having tails come up three times is less that 1:1. The chances of flipping a coin 20 times and having tails come up zero times is less that 1:1. Each of the coin tosses alone bear those odds, but that is a different question. Regardless of how the first 19 flips go, your chances of tails coming up on the 20th toss are still 1:1. That is your point, which I have always understood. I have never stated “it’s got to happen for me soon because I have run Scholo so many times” because I know that is false. My point, as stated above, does not refer in any way to the likelihood the DMG will drop my hat on Friday, which we both agree is 1 in 8.

I will be intolerable regardless of the outcome of my future Scholo runs, just for the record.