Lens Defects
There have been 2535 survey responses so far
There have been responses about Canon lenses from 2017 users
There have been responses about Sigma lenses from 961 users
There have been responses about Tamron lenses from 586 users
There have been responses about Tokina lenses from 224 users
There have been responses about Nikon lenses from 377 users
There have been responses about Pentax lenses from 103 users
There have been responses about Minolta/Konica/Sony lenses from 123 users
There have been responses about Other lenses from 214 users
The data presented here is based on the Lens Defect Survey. No survey is 100% reliable and as someone once may have said 'There are lies, damned lies, and statistics' (Mark Twain quoting Benjamin Disraeli). So be aware that:
- First the survey doesn't sample random users. It samples users who were directed here from one of two or three other websites.
- Second it doesn't sample all users, only those motivated to fill in the survey.
- Third, some people like to mess with surveys and give bogus data. While efforts have been made to detect and eliminate obviously bogus data, it can't all be eliminated.
- Fourth, it's possible that people with complaints are more likely to be motivated to leave their data than those who are happy with their lenses.
- Fifth, there is no differentiation here between cheap lenses and expensive lenses. You might expect cheap lenses to have lower quality control and perhaps ahigher numbers of defects, but on the other hand you might expect buyers of expensive lenses to be more critical and so find more defects!
My requirement before publishing any numbers is that there should be data on at least 100 lenses by any single manufacturer and that data must come from at least 25 different users. With less data than that the data would be even more unreliable. The idea is that the more reports received, the better the statistics will be and no single user has a huge influence on the averaged data. I'd really like to see data from at least 100 different users on at least 300 different lenses before I feel reasonably comfortable with the numbers.
The probabilities assume that the distribution of defective lenses is completly random, i.e. doesn't depend on which lens you chose or where that lens is shipped from. This may not be true, in fact it probably isn't. Some models may be more likely to show defects that others. There could also be bad batches of lenses and if the store you buy from has a bad batch, your chances of getting a defective lens in exchange for a defective lens would be higher than if the distribution was totally random.
The bottom line is therefore that these numbers should be treated with some skepicism. When there are enough entries I may be able to do some more statistical analysis, but the results are only as good as the input data, and I have no direct control over that. As they say in the TV ads for 'psychic advisors', this data is for amusement purposes only. It's interesting, but don't read things into it that may not be there.
Survey Results
Note: Values are greyed out if there are less than 300 lenses sampled
Canon lenses - 9040 with 696 defects
The probability of getting a good lens is 92 %
The probability of getting 5 good lenses in a row is 67 %
Sigma lenses - 1823 with 382 defects
The probability of getting a good lens is 79 %
The probability of getting 5 good lenses in a row is 31 %
Tamron lenses - 904 with 125 defects
The probability of getting a good lens is 86 %
The probability of getting 5 good lenses in a row is 48 %
Tokina lenses - 310 with 41 defects
The probability of getting a good lens is 87 %
The probability of getting 5 good lenses in a row is 49 %
Nikon lenses - 1672 with 119 defects
The probability of getting a good lens is 93 %
The probability of getting 5 good lenses in a row is 69 %
Pentax lenses - 355 with 25 defects
The probability of getting a good lens is 93 %
The probability of getting 5 good lenses in a row is 69 %
Minolta/Konica/Sony lenses - 433 with 42 defects
The probability of getting a good lens is 90 %
The probability of getting 5 good lenses in a row is 60 %
Other lenses - 576 with 59 defects
The probability of getting a good lens is 90 %
The probability of getting 5 good lenses in a row is 58 %
Note for a random distribution of defective lenses the cumulative probability of getting 5 good lenses is a row is lower than that of getting one good lens, but that does not mean that if you have 4 good lenses the chances of the next lens being good are lower. It's like tossing a coin. There is always a 50% chance of it showing tails. If you toss it 9 times and it comes up tails each time, the chance of it coming up tails on the 10th toss is still 50%. However the cumulative chance of getting 10 tails in a row is only 1%, because the chance of getting the first 9 tails in a row was only 2%.