[bhpfaninblr]

Originally Posted by ajayclicks

Purely out of curiosity, do you have numbers for the numerator and denominator, and would you care to share them?

I could write a script to get that information but not sure I will be able to make the time anytime soon nor worth the effort. I have nothing to prove against XUV or otherwise. Seeing the plight of arvindb4 and similar other experiences on the net and on team-bhp searches, I arrived at a heuristic estimate. My distaste for Mahindra comes from the poor safety of their vehicles plus the horrid engine that they put in the KUV petrol.

That is why I used the term heuristic - meaning when I had searched similarly for comparable manufacturers, the numbers came out lower. If you could show me otherwise with the real numbers, would love to see them. Do search and tabulate for comparable competition to the XUV and even a segment lower, and you will see that the ratio works. Of course, if you could prove otherwise, I would be happy to take back my words on the XUV. My heuristic measure from my searches clearly showed the XUV as the culprit

[ajayclicks]

Originally Posted by bhpfaninblr

If you could show me otherwise with the real numbers, would love to see them.

I do not have to prove or disprove anything. You made that point, and I was simply curious as to how you reached that conclusion.

Now, heuristically speaking, when I search, I come across an incident where Toyota was successfully sued when airbags in its Fortuner failed to deploy, leading to the death of the driver. There are also threads here where airbags in eitos have failed to deploy, some incidents of tire burst in the Crysta, a news report of atleast one dealer in US suing Toyota due to safety concerns in Pirus.

But then, I have nothing against Toyota, or Mahindra, or Maruti or whoever....

Cheers

[bhpfaninblr]

Originally Posted by ajayclicks

Now, heuristically speaking, when I search, I come across an incident where Toyota was successfully sued when airbags in its Fortuner failed to deploy, leading to the death of the driver. There are also threads here where airbags in eitos have failed to deploy, some incidents of tire burst in the Crysta, a news report of atleast one dealer in US suing Toyota due to safety concerns in Pirus.

I gave you the heuristic but you seem to have misunderstood. The numerator and denominator have to be used to get the ratio. Reading some random events is not the right way

1. Do a simple search for XUV airbags on google. Filter out airbag issue
2. Do the same on the team - bhp site
3. Tabulate the results
4. Look at sales of XUV
5. Compute the ratio at both google and team-bhp site
6. Repeat 1-5 for Jeep Compass, Creta, ..any other. Or for that matter a Brezza or an Ecosport - you will find that even Brezza and Ecosport that are a segment lower and sell much more have a lower ratio than the XUV

See the numbers for yourself. Your post makes it seem that I am quoting random instances. It is a formula based on heuristics my friend!

If you repeat above for broken alloys, quality issues,...you will be surprised at what you get out of the XUV! I will try to make some time to prove this point. I do not want any more team-bhpians to face the same fate as what arvindb4 has faced.

[SDP]

Originally Posted by bhpfaninblr

I gave you the heuristic but you seem to have misunderstood. The numerator and denominator have to be used to get the ratio. Reading some random events is not the right way

When I asked you the original question, I wasn't looking for what method/process (heuristic or otherwise) you used to come to the conclusion. I was (and still am) looking for hard undeniable data.

1. Do a simple search for XUV airbags on google. Filter out airbag issue
2. Do the same on the team - bhp site

You are assuming that TeamBHP or even the broader internet provides a realiable sample for deriving statistical inferences. In reality, it does not. Do all the ~3 lakh XUV owners have access to Internet and forums like TeamBHP (posting privilege) to share their positive or negative experiences? Even if they have that access, does everybody care to share their positive experiences with the same zeal with which a negative experience is shared? Assume that there are 1000 owners of a particular car model and all have access to forum like teamBHP. If 998 are extremely happy with the car, but the 2 have faced nothing but problems, do you think there would be 998 positive reviews/posts/threads about the car in question and just 2 negative threads? I am pretty sure the ratio of positive v/s negative threads/reviews would be completely skewed towards the negative and the 2 threads about problems will be teeming with comments, blames, unsubstantiated conclusions and what not ..and mostly from people who have zero experience with that car. So if I were you, I would be very careful while deriving inferences from any heuristic or overly simplified methods.

Read about your 'distaste' for Mahindra vehicles in another post. Heard about 'confirmation bias'? If not, please do Google. That would explain why it's so easy for you to believe that all Mahindra products are of poor quality and unsafe.

See the numbers for yourself. Your post makes it seem that I am quoting random instances. It is a formula based on heuristics my friend!

You made the outlandish comment about ALL Mahindra products being unsafe. So I am expecting the corresponding numbers from you and not Ajay

If you don't have the numbers to prove your claim, just say so. It's a free world and everybody is entitled to an opinion. All that I request is - please don't present your 'opinions' or 'perceptions' as 'facts'. We will still hear and respect your opinion, provided it clearly includes the disclaimer "in my humble opinion" and does not try to overgeneralize based on unrepresentative data samples or no data at all .

[Biraj]

First, I hope and pray that Arvind's son recovers quickly.

Not sure if this has been posted earlier in the thread but while searching for details around Mahindra XUV and airbags, I came across this http://www.team-bhp.com/forum/road-s...e-upgrade.html. As per the recall, all XUVs manufactured on or before July 2014 were supposed to undergo a software upgrade for side-airbags. It was mentioned in the service request - "There is a possibility of an intermittent fault code that might inhibit deployment of the side curtain airbag."

Arvind should check whether his car was recalled and if yes, did Mahindra make the changes. It can be a case of serious negligence along with airbag malfunction if the recall/repair didn't happen with this fateful car.

[nav2111]

Originally Posted by SDP

When I asked you the original question, I wasn't looking for what method/process (heuristic or otherwise) you used to come to the conclusion. I was (and still am) looking for hard undeniable data.

While I generally agree with you that one should not present their opinions as if they were facts backed by some sort of data analysis, I feel that the example you gave of comparing +ve vs -ve threads somewhat misses the point. How many people have had no issues with their cars is of no significance here. What bhpfaninblr wanted to compare was negative vs negative threads across all brands/models. I agree that collecting data here or on the internet is not going to give a good sample but I feel that collecting data on failures is of more importance regardless of the number of positive experiences.

[ruzbehxyz]

Originally Posted by Biraj

First, I hope and pray that Arvind's son recovers quickly.

Not sure if this has been posted earlier in the thread but while searching for details around Mahindra XUV and airbags, I came across this http://www.team-bhp.com/forum/road-s...e-upgrade.html. As per the recall, all XUVs manufactured on or before July 2014 were supposed to undergo a software upgrade for side-airbags. It was mentioned in the service request - "There is a possibility of an intermittent fault code that might inhibit deployment of the side curtain airbag."

Arvind should check whether his car was recalled and if yes, did Mahindra make the changes. It can be a case of serious negligence along with airbag malfunction if the recall/repair didn't happen with this fateful car.

This is a nice point that you have made. However in many cases, manufacturers recall cars in the name of airbags to rectify some other faults. Well this may not be the case with the XUV but it can be.

[SDP]

Originally Posted by nav2111

....I feel that the example you gave of comparing +ve vs -ve threads somewhat misses the point. How many people have had no issues with their cars is of no significance here.

Good point and I did think about this aspect when drafting my post. You are absolutely right that the fantastic experiences of the 998 do not matter to the 2 who might have gotten into some serious trouble with the same car. I still included the point about positive experiences because possibly only 3 out of the 998 would bother to share those positive experiences/reviews and then when someone takes a shortcut of doing 'research' on internet, its very easy to come to the conclusion that "2 out of 5 people had serious issues with their car". In my opinion, that's the real reason behind gross-generalizations like "Mahindra has been making unsafe cars and is putting lives of people at stake".

What bhpfaninblr wanted to compare was negative vs negative threads across all brands/models.

I understood the intention. Was just asking for data.

...I feel that collecting data on failures is of more importance regardless of the number of positive experiences.

Again, where do you do your research and what data sources you use is extremely important.
For example, a good friend AnilKumar MO shared the following link:
https://indiankanoon.org/search/?for...did+not+deploy

Some concrete data on court cases between automobile manufacturer and consumers impacted by non-deployment of airbags in an accident scenario. Right from Mercedes and BMW, to Hyundai, you will find all major manufacturers there. What happened with Arvind's son is still a major issue and I am not trying to defend Mahindras by saying it happens with Mercs and BMWs as well. No doubt, there should be a proper investigation to assess the root cause(s) and an analysis to figure out if its an isolated issue, or an issue with specific batches or a basic design flaw. Until that investigation and analysis is done, I wish respected members could hold their horses and maintain self-restrain instead of jumping to conclusions, passing judgments and arriving at gross generalizations like we have seen in bhpfaninblr's post.

[selfdrive]

Originally Posted by SDP

Until that investigation and analysis is done, I wish respected members could hold their horses and maintain self-restrain instead of jumping to conclusions, passing judgments and arriving at gross generalizations like we have seen in bhpfaninblr's post.

I agree with this point. Let us not arrive at either assumption that M&M builds :
- unsafe cars, or
- safe cars

I don't know if we have any organisation/ agency that compiles such accident data and analyses it. As we have ARAI for pre testing/ certifications, is there any agency that deals in analysis of accidents? Obviously with the intention of eliminating the root cause and reducing fatalities/ serious injuries.

Roads and drivers could be some of the causes, but what if there are other contributors?

[suzuki san]

Hi Arvind,
Really sorry to hear about your son's accident. Wish him a full and speedy recovery.

I know this post or the pictures are not really going to help you in any way but posting nevertheless.

This accident took place about a week ago. The black Honda City driver claims his brakes failed and he pulled the hand brake after which his car swerved. His car jumped over the divider, smashed through a solid steel barricade and hit our (my relatives) car - a Maruti Suzuki 2010 SX4 Zxi. The airbags deployed in the SX4 even though the damage is nothing compared to your XUV500. There was a driver and 2 rear seat passengers in the car at the time of the accident which took place in broad daylight in Kandivali West. No one was injured though the driver suffered a bit of whiplash and was quite shocked by the airbag deployment.

The driver of the Honda City was also not injured at all though I am not sure if the airbags deployed in his car. It looks like they did but I am not sure.

My point in sharing the pics is to highlight the fact that even with relatively light damage the airbags still deployed.

Regards,
SS

[samaspire]

Originally Posted by suzuki san

Hi Arvind,
- a Maruti Suzuki 2010 SX4 Zxi. The airbags deployed in the SX4 even though the damage is nothing compared to your XUV500

Something's not right here. The airbags are supposed to deflate immediately. Why is one still inflated?

[mayankk]

Originally Posted by samaspire

Something's not right here. The airbags are supposed to deflate immediately. Why is one still inflated?

I don't think it is. It's just propped up, or just the stiffness of the material holding it, or a combination. The picture from the front clearly shows the vent holes.

[bhpfaninblr]

Originally Posted by SDP

When I asked you the original question, I wasn't looking for what method/process (heuristic or otherwise) you used to come to the conclusion. I was (and still am) looking for hard undeniable data.

Please do update the thread when you find hard undeniable data either way. My method/process is based on the heuristic tested against various cars including those a segment lower like the Ecosport and Brezza. Not to forget SUVs in the same segment
Not sure you or I have the means to commission a survey that comes out with a report that XUV airbags are indeed unsafe. By that yardstick, no statement can be proven anyway. For me, the heuristic is a good enough benchmark as that is the way most decisions are made. The team-bhp data or the google data is sufficient data to come up with a conclusion statistically. See my quote below

Originally Posted by SDP

You are assuming that TeamBHP or even the broader internet provides a realiable sample for deriving statistical inferences.

What I proposed is a realistic sample. I am sure you already know having used the word 'statistical inference'. Anyway, the details are below:
All statistical inferences are based off samples. The standard deviation of different sample standard deviations is called the standard error. The mean difference/standard error is the t-ratio and all statistics are based off different samples.
In this particular case, we would use the the samples drawn from team-bhp for different cars. Once we get the values, we can do a t-test with an assumption that the
H0: The mean airbag failure rate of XUV, Ecosport, Brezza, Creta are the same
We could do a t-test with unequal variances or equal variances based off another F-test. So we can get statistically significant inferences from the team-bhp sample
We can even take it a step further and try it out with Google

Originally Posted by SDP

I wish respected members could hold their horses and maintain self-restrain instead of jumping to conclusions, passing judgments and arriving at gross generalizations like we have seen in bhpfaninblr's post.

It is based off the measures I used to say XUV is poor. I admire your passion but I have outlined my approach and method. If you have a better method, please let us know - happy to change my opinion. If you need more data and analyses to conclude, do not conclude. For me and many others, the framework and methodology posted here would suffice. So let us "not generalize" what amount of data/methodology is conclusive to "generalize". I do not mean to hurt your sentiment

[ferrarirules]

First of all Arvind, wishing a speedy recovery to your son and may you get the justice you deserve.

Originally Posted by bhpfaninblr

What I proposed is a realistic sample. I am sure you already know having used the word 'statistical inference'.

I dont think you are getting the point here. The approach that you have taken is right. But the assumptions for gathering the realistic sample is way off. You are basing your statistics on a sample which doesn't take into account alot of factors. I will mention a few

1. How many people who own a Maruti or Ecosport have access to internet?
2. How many of these people reported their airbag failures on the internet?
3. How many pages of google search results did you use to gather a substantial sample. Not every incident will be present on the first few pages of google search as the number of Maruti cars is way higher than Mahindra or XUV?

Answers to these questions will add major deviation in the sample collected by you. So no one is challenging you on the maths of it. What everyone is challenging is that the sample set which you used for analysis or "Statistical inference" is incorrect/insufficient.

Hence it is wrong to generalize that XYZ manufacturer doesn't make good cars based on weak sample data.

[SDP]

My method/process is based on the heuristic tested against various cars including those a segment lower like the Ecosport and Brezza. Not to forget SUVs in the same segment
.....

What I proposed is a realistic sample. I am sure you already know having used the word 'statistical inference'. Anyway, the details are below:
All statistical inferences are based off samples. The standard deviation of different sample standard deviations is called the standard error. ....

We could do a t-test with unequal variances or equal variances based off another F-test. So we can get statistically significant inferences from the team-bhp sample
We can even take it a step further and try it out with Google

All that is fine. Its very easy to formulate a method to calculate. I am reiterating, I am not interested in the method, just the outcome. Going by your statements which include a lot of "we can" "we could", I am sure that all that has been said is just theory and that has not actually been used before you made the statement.

Originally Posted by bhpfaninblr

Please do update the thread when you find hard undeniable data either way.

Since you made a statement, logically the responsibility lies with you to prove that statement or take it back. Please don't expect others to do the leg work to prove that your statement was wrong. Its as simple as that. I have already explained that twice. I am pretty sure, you don't have any data to back up your claim and that's possibly the reason for all the beating around the bush and jargon like heuristic.

Anyways, I am out of here. Can't afford to spend any more of my time and effort towards a discussion which doesn't seem to be going anywhere. Good luck.