That's only sort of true OF (sorry to be picky) ... but following your logic, the odds of having a defective Ascent if one has never gotten one, is 1 in 10 ... the odds of getting 3 in a row if one has never gotten one before is, as you stated, 1 in 1000, however, once one has received 2 defective Ascents, the odds of the 3rd one being defective is still 1 in 10 if the defect rate is 10% (not saying that it is ... pretty sure its lower than that)
Gamblers fallacy ... your odds don't go up as you keep playing ... they are the exact same on every roll and all prior rolls have no impact on the current one ...
Believe what you want, I'm I think stating the way statistics work. For sure the way I was taught to do SFA (Statistical Failure Analysis). The chances of Da Vinci having such a customer (3 dead in a world of 9/10 being fine) is indeed a thousand to one.
999 of those thousand customers will have gotten at least one good one in the three they get. Unlike the case you cite, the 'failures' are related to each other.
Yes, the chance the next one will be a dud is exactly the same as the first, second of third. But the chances of four random draw from a population of 90% reliability (and I think it's much higher than that for Ascent really.....) are now 10,000 to 1. You just don't meet such people in the real world. At least no where near as often as it seems reading this forum. I seriously doubt DaVinci has anything like that many customers?
We're not talking statistics here on this (or any other) Forum. This is 'self selected anecdotal data' from non certified sources. Interesting, but IMO not something to hang your hat on.
If more than .1% of your customers report having 3 in a row bad, something else is in play. You either have bad statistics for your base, a problem in distribution (such that each example isn't independent), a
customer issue (not the uints) or you're dealing with the unluckiest customer in six western states.
I worked CS doing root cause analysis on failures too long to think it has anything to do with throwing a natural next time at Craps.
That logic has to do with what happens on the next pass, not with whether the gambler goes home broke or rich. It's kind of fun, when you make the rules that you must find the
specific cause of failure in each case, suddenly a lot of your 'failures' can turn out to be not real. Since the early days of Demming and 'Statistical Quality Control' this idea (trusting your history to understand the quality picture) has become 'the Holy Grail'. It really is a proven, powerful tool......and it works that way, not like at 'Vegas'.
Regards,
OF
)
