Drive unit replacement - getting better?

[vgrinshpun]

This forum reports a much higher failure rate than 5 percent: Drive Unit Problem / Fixes - View Poll Results

I suggest brushing up on your statistics.

This is not proper statistical data as this poll is clearly going to be biased toward people who **had** problems with the drive unit.

The Consumer Reports annual questionnaires are sent to *all* subscribers; their data based on the questionnaire voluntarily returned by the subscribers.

 

[hpham007]

I suggest brushing up on your statistics.

This is not proper statistical data as this poll is clearly going to be biased toward people who **had** problems with the drive unit.

The Consumer Reports annual questionnaires are sent to *all* subscribers; their data based on the questionnaire voluntarily returned by the subscribers.

The TMC poll had a sample size of 136. Can you tell me the sample size of the CU questionnaire for Tesla?

 

[mspohr]

This forum reports a much higher failure rate than 5 percent: Drive Unit Problem / Fixes - View Poll Results

If you believe this poll... I have a bridge to sell you.
People who have problems are highly likely to respond to this poll. Those with no problems are not likely to respond to the poll. It is worthless.

 

[Lump]

Just this morning I met a former owner who had so many problems Tesla agreed to buy his car back, his VIN was 4xxx delivered in Feb. 2013 & I doubt he would bother filling out a customer satisfaction questionnaire from Consumer Reports.

 

[mspohr]

Just this morning I met a former owner who had so many problems Tesla agreed to buy his car back, his VIN was 4xxx delivered in Feb. 2013 & I doubt he would bother filling out a customer satisfaction questionnaire from Consumer Reports.

Why not?

 

[vgrinshpun]

Just this morning I met a former owner who had so many problems Tesla agreed to buy his car back, his VIN was 4xxx delivered in Feb. 2013 & I doubt he would bother filling out a customer satisfaction questionnaire from Consumer Reports.

This is clearly highly pertinent statistical data.

If anything Consumer Reports subscribers are much more conscious about the reliability of the products they buy than population at large. this is exactly the reason they pay yearly dues and spend time providing information on reliability of the products they buy, and thoroughly review Consumer Report data before buying.

 

[ecarfan]

This forum reports a much higher failure rate than 5 percent: Drive Unit Problem / Fixes - View Poll Results

"This forum" does no such thing! That is not a statistically valid sample of Model S owners. It means nothing.

 

[ecarfan]

The TMC poll had a sample size of 136. Can you tell me the sample size of the CU questionnaire for Tesla?

The TMC poll is not a statistically valid "sample". It is meaningless.

 

[jerry33]

The TMC poll is not a statistically valid "sample". It is meaningless.

Correct. Many people join an automotive forum because they have problems and are looking for information. Go to any other forum and you'll see the same thing. If people were to base a car purchasing decision on forums, no cars would ever be sold.

 

[vgrinshpun]

The TMC poll had a sample size of 136. Can you tell me the sample size of the CU questionnaire for Tesla?

First, it is Consumer Reports (CR), not CU (Consumer ????).

Secondly, if the poll is biased toward people who had problems with the drive unit, the size of the sample is irrelevant to the statistical validity of the poll results.

The data for CR is collected and presented by the team of professionals, including automotive engineers and statisticians. Based on proper statistical analysis they decide if the sample is large enough for the data to be valid. If the sample is not big enough to draw conclusions they, instead of presenting data, mark the cell with an asterisk indicating "insufficient data". See couple of snapshots with the reliability data for MB Class S - CR did not include any data for the 2011 column because of the "insufficient data".

BTW according to this statistical data the MB flagship has reliability that is -126% worse than a group of peers, with CR reports categorizing it as "much worth than average". MS is categorized as "Average" and has reliability 6% worth than New hybrids/EVs.

 

[hpham007]

First, it is Consumer Reports (CR), not CU (Consumer ????).

Secondly, if the poll is biased toward people who had problems with the drive unit, the size of the sample is irrelevant to the statistical validity of the poll results.

The data for CR is collected and presented by the team of professionals, including automotive engineers and statisticians. Based on proper statistical analysis they decide if the sample is large enough for the data to be valid. If the sample is not big enough to draw conclusions they, instead of presenting data, mark the cell with an asterisk indicating "insufficient data". See couple of snapshots with the reliability data for MB Class S - CR did not include any data for the 2011 column because of the "insufficient data".

BTW according to this statistical data the MB flagship has reliability that is -126% worse than a group of peers, with CR reports categorizing it as "much worth than average". MS is categorized as "Average" and has reliability 6% worth than New hybrids/EVs.

View attachment 94018
View attachment 94019

So what was the sample size used by CU again?

 

[vgrinshpun]

So what was the sample size used by CU again?

Large enough to have statistically valid data, as explained in my post.

 

[Lump]

This is clearly highly pertinent statistical data.

At no time did I claim "pertinent statistical data"!

But Tesla did buy back a car from an early adopter who like many had high hopes & was repeatedly disappointed with his car & the service he received, after his car shut down on the freeway Tesla assured him everything was fixed then when it happened a second time with their small children in the car that was the last straw.

 

[vgrinshpun]

At no time did I claim "pertinent statistical data"!

But Tesla did buy back a car from an early adopter who like many had high hopes & was repeatedly disappointed with his car & the service he received, after his car shut down on the freeway Tesla assured him everything was fixed then when it happened a second time with their small children in the car that was the last straw.

This is a real bummer.

What was the reason his car shut down on a freeway two times?

 

[rxlawdude]

This forum reports a much higher failure rate than 5 percent: Drive Unit Problem / Fixes - View Poll Results

Without knowing the DENOMINATOR, all you are recording are responses from owners with the problem (at least once).

You further assume that TMC posters are representative of the MS owner population.

Fatal flaws in your statistical methods make your conclusion less than reliable.

 

[Lump]

Poor guy had every problem imaginable, mis aligned panels, leaking pano, door handles replaced mutable times etc...combined with poor service.

He mentioned battery cables replaced but sounded like the 12V battery caused the shut downs, received the "car is shutting down & you need to pull over immediately", this happened to them twice. Their 12V was replaced 3 times in total & their car was given a clean bill of health & returned to them but soon after the second shut down occurred...they were done.

 

[hpham007]

Large enough to have statistically valid data, as explained in my post.

So you don't know. It could be 10 or 1000, it doesn't matter to you because the results fit with your narrative. Ponder this for a second, Motor Trends and Edmunds long term Tesla both had their motors replaced (Edmunds had four drive units in 30k miles). Can you think of any other car that had to have it's motor replaced in two separate long term tests?

 

[vgrinshpun]

So you don't know. It could be 10 or 1000, it doesn't matter to you because the results fit with your narrative. Ponder this for a second, Motor Trends and Edmunds long term Tesla both had their motors replaced (Edmunds had four drive units in 30k miles). Can you think of any other car that had to have it's motor replaced in two separate long term tests?

Just in case you missed what I posted - the sample used by Consumer Reports is large enough to have statistical significance, otherwise, they would not publish information, noting "insufficient data".


This innuendo that CR data can't be reliable is outright laughable: you demonstrated ignorance on the subject by claiming that rate of the drive unit problems is higher than 5 percent based on statistically irrelevant data, and yet you feel confident enough to challenge published data from an organization that is employing professionals and doing this type of surveying and statistical analysis since the 1940-ies (most likely longer than you are alive)...

I have nothing to add here.

 

[rxlawdude]

Just in case you missed what I posted - the sample used by Consumer Reports is large enough to have statistical significance, otherwise, they would not publish information, noting "insufficient data".


This innuendo that CR data can't be reliable is outright laughable: you demonstrated ignorance on the subject by claiming that rate of the drive unit problems is higher than 5 percent based on statistically irrelevant data, and yet you feel confident enough to challenge published data from an organization that is employing professionals and doing this type of surveying and statistical analysis since the 1940-ies (most likely longer than you are alive)...

I have nothing to add here.

+1. You can't fix stupid.

 

[kennybobby]

CR for subscribers only

CR indicates requiring a minimum of 100 responses in order to publish the data for any model in any model year.

"Our sample sizes tend to track quite closely with market sales. Individual sample sizes vary from year to year and range from a hundred to several thousand for the more popular models."

The survey is only open to subscribers of either the magazine or the website, as are the results.

Their survey is scientifically designed by experts and they consider it to be both valid and reliable in the data and results.

So it's hard to argue with science and the facts--tesla DU problems are really very minor and insignificant, if they even exist at all. If it were otherwise the CR survey and reports would have clearly identified it--since they didn't it's not an issue.