Simple Maths Problem

sansnom

sansnom

sansnom

Associate
Joined
14 Feb 2004
Posts
1,103
Location
London
I feel really dumb asking this but for some reason I just can't see how to do it.

I'm trying to value used cars a simple math formula that takes into account year of production (as a factor with 2014 being 1), mileage and price.

I don't know why but it's confusing me. Ideally the formula should be something like price per mile multiplied by the factor but it isn't working right.

Can one of you math gurus help?

 
I'm not sure it's possible to do that without making mileage an index as well...lemme have a think about it.

The problem is taking price and dividing it by mileage will give you nonsensical ratios.
 
Last edited:
op, you need to say what your expected milage per year is, then you can figure out a constant to make things work:

Define

Value = (BasePrice / Price) * (Mileage/BaseMileage)

Where basePrice is your 2014 price * factor and base mileage is X * age where X is whatever you want to take as acceptable mileage per year. You still have to define a constant to change the weightings - e.g. what should the 'value' be for price as expected, but mileage is double X*age?

NB this wont give linear changes...
 
Last edited:
Agree with the above. My recommendation is to pick a base expected mileage (say 12000/year) and then experiment with various example mileages from various ends of the spectrum (low/medium/high) to come up with a suitable weighting.

One challenge you will have when looking at such modern cars is that you don't have enough granularity in terms of age. In other words a 2013 car could be brand new or nearly a year old, so any expected mileage you come up with could be massively different in ratio terms from the actual mileage (e.g. say you pick 5000 as expected mileage for a 2013 car, a brand new car with 5 miles on the clock will be 0.1% of the expected mileage). So I wouldn't weight things too heavily in that regard, think about how much such a car would be worth in reality compared to say an early 2013 car on 10000 miles, maybe 15% difference max(?) despite a massive difference in mileage ratio, If you've got information for month of registration, or even plate (6-monthly) I would use that instead.
 
Last edited:
HangTime said:
Agree with the above. My recommendation is to pick a base expected mileage (say 12000/year) and then experiment with various example mileages from various ends of the spectrum (low/medium/high) to come up with a suitable weighting.

One challenge you will have when looking at such modern cars is that you don't have enough granularity in terms of age. In other words a 2013 car could be brand new or nearly a year old, so any expected mileage you come up with could be massively different in ratio terms from the actual mileage (e.g. say you pick 5000 as expected mileage for a 2013 car, a brand new car with 5 miles on the clock will be 0.1% of the expected mileage). So I wouldn't weight things too heavily in that regard, think about how much such a car would be worth in reality compared to say an early 2013 car on 10000 miles, maybe 15% difference max(?) despite a massive difference in mileage ratio, If you've got information for month of registration, or even plate (6-monthly) I would use that instead.
Click to expand...

Why so complicated, use a well tested method of...

Guessing!
 
Just add 25% to the price you paid for it? Surely that gives you wiggle room to negotiate when re-selling?

I mean you would only pay 1k for a high mileage banger, try sell it for 1250... get 1100 and you are happy?

Conversely a top exec motor you buy for 28000, so you list it for 35000 and can sell that for like 33000? Thats a nice tick.

idk how cars work, but that seems simple enough to me.
 
If you're making some sort of attempt to identify which of a series of cars you find on autotrader are worth looking at this entire idea surely comes unstuck because you're completely ignoring several other big factors that affect value such as optional extras and number of owners.
 
I think it would be useful to know the overall objective too, i.e. is it a hypothetical problem, is it target prices for a used car dealership, is it trying to weigh up possible options for a car to buy etc. The first option is the only one where it makes any sense to even be attempting this with so few variables.
 
sansnom said:
glad im not the only one who hasn't figured it out yet though!
Click to expand...

That's because it cant be done.
Your figures have no relation to each other at all.
Look at the 1st and 3rd lines for example:
Code: 
Year    Mileage    Price    Estimated value
---------------------------------------
2009    56000      21739    17846
2009   110000      21739    17846

You cant have a simple formula that includes mileage but still gives the same output when you enter different mileages.
 
touch said:
That's because it cant be done.
Your figures have no relation to each other at all.
Look at the 1st and 3rd lines for example:
Code: 
Year    Mileage    Price    Estimated value
---------------------------------------
2009    56000      21739    17846
2009   110000      21739    17846

You cant have a simple formula that includes mileage but still gives the same output when you enter different mileages.
Click to expand...

Isn't op just showing us that his spreadsheet makes no sense?
 
Amleto said:
Isn't op just showing us that his spreadsheet makes no sense?
Click to expand...

Could well be, there isnt a lot of info about it tho, so i'm making a few assumptions.

It doesnt look to me like any of the fields shown are the results of a formula he has tried?
'Estimated market value' seems to be read from the 'base index' table at the top (for some reason?).
I was guessing that the 'value ratio' field is where he wants his formula.
 
You must log in or register to reply here.
Back
Top Bottom