test content
  • Games
  • Social
  • Arc News
  • Support
  • Zen
What is the Arc Client?
Install Arc
Forums > Home› Power Discussion

is teddied really only 1% chance?

Archived PostArchived Post Posts: 1,156,071 Arc User
edited November 2011 in Power Discussion
I ask this because I have been playing an Inventor, and teddied is firing off left and right at rates way too high to be that low of a percent. It's been a running joke that every single boss in the game i have fought i have teddied at some point, and I'm curious if the math is borked.
Post edited by Archived Post on

Comments

  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    stagcutlery wrote:
    I ask this because I have been playing an Inventor, and teddied is firing off left and right at rates way too high to be that low of a percent. It's been a running joke that every single boss in the game i have fought i have teddied at some point, and I'm curious if the math is borked.

    Sometimes the RNG ain't so random...
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    No, it's still random. It's not busted in the sense of I get back to back teddied's or death rays, just the frequency feels more like 10% than 1%.
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    I play a lot of Gadgeteering toons - I have 107 levels worth of Experimental Blaster experience :)

    I love the Experimental Blaster - it is kind of funny when you teddy Mega-D or Tako - just yesterday I teddied a Lemurian submarine; that was a bit weird :)

    As for your question: It's a trick of the RNG - Random is Random. Every shot has a 1% chance of teddying the enemy. Sometimes it teddies two or three enemies in as many pulls, sometimes you can play a whole evening without seeing the little teddy once.

    One percent isn't that seldom when you fire your EB about ten times per pull, the average would be a teddy every ten pulls - which seems to fit with my experience. Ten percent would be a teddy on every single pull, and that's clearly not so.
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    Perhaps you don't understand probabilities. It's perfectly possible for you to teddy someone a hundred times in a row. Not very likely (in fact, very very unlikely), but nonetheless possible. A 1% chance does not mean every 100th time is guaranteed to do it and the other 99 are not; it just means that on average out of a hundred tries, one will succeed. But average is just that -- your results will not necessarily conform to the probabilistic expectation until your sample size is sufficiently high. I recommend looking up statistics (statistical analysis, probability theory) on an encyclopedia for further information (it is useful, in general, to understand probabilities better). Don't take that as condescending for it is not intended as such -- I myself find the need to revisit the subject somewhat frequently because it's quite easy to get it wrong.

    The nature of randomness, and indeed that of probabilities, are something most people could brush up on for they have unforeseen implications for life in general -- attributing credit and blame, for instance, evaluating causal relationships etc. One interesting book on the subject that I can recommend is The Drunkard's Walk - How Randomness Rules Our Lives by Leonard Mlodinow.
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    I understand statistics. I am saying in the amount of time I have leveled, I am not seeing average teddies of 1%. Probability is only meaningful with a real sample size, not some indiscriminate one. 1% is a very low average. I should not be seeing as many teddieds as I am given the realistic amount of zapping I am doing. What you are talking about is an academic fallacy where probability is not attached to any real sample size, so you can push back numbers infinitely if they don't reach the desired percent.
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    Uh..I ain't no math whiz but isn't that why it's called "chance"?

    There are times I get sick of having to beat on teddy bears that were formerly mobs that would have otherwise been one-shot. Then there are times I never see one teddy.
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    If you're seeing them in streaks then there could be something wrong with Cryptic's RNG. I don't think it would be the first time such a thing has happened.
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    Galactiman wrote:
    If you're seeing them in streaks then there could be something wrong with Cryptic's RNG.
    Streaks are part of any random sequence. That's why you have to do a very large sample to say with any confidence what percentage chance teddying really has. I don't recall my statistics, probability, and hypothesis testing formulae well enough to be able to calculate the exact number of tries needed, but I'd say it's well over 10,000 to tell if the chance is 1% as opposed to 2%.

    That it's not 10% should be clear in just a few short minutes of play though, since that would mean a teddy on nigh on every pull.
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    stagcutlery wrote:
    I understand statistics. I am saying in the amount of time I have leveled, I am not seeing average teddies of 1%. Probability is only meaningful with a real sample size, not some indiscriminate one. 1% is a very low average. I should not be seeing as many teddieds as I am given the realistic amount of zapping I am doing. What you are talking about is an academic fallacy where probability is not attached to any real sample size, so you can push back numbers infinitely if they don't reach the desired percent.

    If you're seeking to challenge the assumption that the probability of teddying is 1%, then you need to perform an empirical experiment with a large enough sample size. Simply stating that your experience differs from the average expectation proves nothing and is no basis for a generalization.

    So, where is your data? If you're here genuinely to find out if there's anything to your doubts, surely you must recognize that nothing will be achieved until you bring us some numbers to back up your suspicions. And surely you don't expect anyone to take your personal account that basically amounts to "it doesn't feel right to me" as anything but anecdotal, so I ask -- what is your goal here?
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    Shmn727 wrote:
    Simply stating that your experience differs from the average expectation proves nothing

    It proves they were luckier (more lucky? which one is correct?) than most at that point, I guess?
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    Something else to consider here , is we don't know what the 1% means. Or rather, what it's based on. I know that's not really helpful to the OP but it's some food for thought. We don't know if that's 1% of 100 or 1000 or 1,000,000 etc so yea...
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    We know exactly what the 1% means. By definition, 1% means 1 in 100. A 1% chance means that, averaged over time, 1 in 100 shots will teddy the target.
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    The reason I asked was to get more data than my experience. In other games it's easy, you just download and run a parser to parse the chat log. This i dont think is possible, and I was curious if others noticed any irregularities.

    the thing about teddying on every pull is that I dont use a lot of experimental blaster shots always. The pets do the bulk of damage, and also experimental rifle is more efficient to pull with, because you can aoe. There's enough randomness to cause doubt-if its only 1%, its conceivable i can go more than 100 shots without one teddy, and thats not been my experience at all. 1% means only on average 1 per ten pulls by your criteria, and I am not seeing that either.

    I can try and hit the training dummies but tbh i dont think i can do the sheer number of zaps needed for a sample size without a parser or going mad.
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    MistressBria wrote:
    Something else to consider here , is we don't know what the 1% means. Or rather, what it's based on. I know that's not really helpful to the OP but it's some food for thought. We don't know if that's 1% of 100 or 1000 or 1,000,000 etc so yea...

    *facepalm*
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    stagcutlery wrote:
    The reason I asked was to get more data than my experience. In other games it's easy, you just download and run a parser to parse the chat log. This i dont think is possible, and I was curious if others noticed any irregularities.

    the thing about teddying on every pull is that I dont use a lot of experimental blaster shots always. The pets do the bulk of damage, and also experimental rifle is more efficient to pull with, because you can aoe. There's enough randomness to cause doubt-if its only 1%, its conceivable i can go more than 100 shots without one teddy, and thats not been my experience at all. 1% means only on average 1 per ten pulls by your criteria, and I am not seeing that either.

    I can try and hit the training dummies but tbh i dont think i can do the sheer number of zaps needed for a sample size without a parser or going mad.

    If you can't do the sheer number of zaps, then it's obviously 1%.

    10% should be readily recognizable.
  • Archived PostArchived Post Posts: 1,156,071 Arc User
    edited November 2011
    stagcutlery wrote:
    The reason I asked was to get more data than my experience.
    I offered my experience (109 levels of Gadgeteering and Experimental Blasting) in my first post. I don't know if it's a 1% chance or a 2% chance, but it's not 10%, that much is clear as day by just using the EB in-game.
    stagcutlery wrote:
    the thing about teddying on every pull is that I dont use a lot of experimental blaster shots always.
    Then take a session and only use EB. It should be readily apparent that the teddying chance is much lower than 10%.
    stagcutlery wrote:
    There's enough randomness to cause doubt-if its only 1%, its conceivable i can go more than 100 shots without one teddy, and thats not been my experience at all. 1% means only on average 1 per ten pulls by your criteria, and I am not seeing that either.
    1% means only that each shot has a 1 in 100 chance to teddy the enemy. It does NOT mean that if you fire your EB one hundred times you'll get exactly one teddy. You might get none, ten, fifty, even a hundred (although that is so extremely unlikely as to be practically impossible).

    The only way to statistically prove that it's a 1% chance is to do a lot of EB taps - as I said, I'd venture a guess that you'd need upwards of 10,000 to see any statistically significant result.

    On the other hand, a 10% chance should be easily disproved by a lot fewer EB taps.
    stagcutlery wrote:
    I can try and hit the training dummies but tbh i dont think i can do the sheer number of zaps needed for a sample size without a parser or going mad.
    As I said above, to informally disqualify your hypothesis of 10% you probably wouldn't need all that many zaps at all.
Sign In or Register to comment.