Show HN: Car Cost Calculator – my weekend project | Hacker News Reader

Show HN: Car Cost Calculator – my weekend project

car-cost-calculator.com

48 pointskhet45 comments

You might add a parking cost field, because for anyone in a city, parking's unlikely to be free

Where did you get a 20-30% average depreciation on cars? Depreciation is not linear on vehicles. Most of the depreciation occurs in the first few years and then tapers off. Is this calculated into your algorithm?

scroll down. He explains his algorithm : "The caclulator uses exponential depreciation is used to calclate the cost of depreciation and the asset value of a car over time"

That sounds like a fair way to do it.

It looks like the 20% is 20% of the value that year, so it looks like value is shown with exponential decay.

So if a car starts out at $20,000 it will be worth $16,000 the next year, $12,800 after the second year, and $10,240 after the third. Notice that around the third year, the decrease in value slows considerably.

Yup. And I'd say that's pretty accurate.

I'm using an exponential depreciation function to calculate the cost of the car in time. The car loses most of it's value during the first few years.

While I'm fine with exponential depreciation, the rates are very high (or maybe just variable by car). My RAV4 has closer to 9-10% depreciation (assuming 15k miles/year) per the Edmunds TCO mentioned elsewhere. May be constructive to look up your individual car and input it into khet's calculator.

Kudos for giving us the option. May be helpful flag inputs that are highly deterministic of the output and suggest look-ups.

Yeah I actually have an outlier that has negative depreciation. Got it used for 26k with 44k miles. 6 Months later my insurance company informed me that they had to put my car value closer to 32k due to replacement costs and that this may go up over my ownership of the car.

That's actually not true, at least not always. I did some analysis on KBB for a car I was looking to buy. I found that for my particular model, depreciation between years 1, 2 and 3 were far less than between 3, 4 and 5. There's a graph on http://www.undiscoveredfeatures.com/2011/09/kelly-blue-book.... that shows the relationship.

Like someone else mentioned, there are better tools to calculate static ownership costs. This is definitely a good start.

I'm working on https://www.maintenr.com and we allow you to calculate your service and fuel expenses dynamicly.

Could you explain what your planned features are & who your target audience is? The 'tutorials', will you have any generic ones for people to use or a repository of them? Can the tutorials be made public vs private? This is interesting & I signed up just to see what it's about. I don't do my own work on my car but I knew a few mechanics & I understand the 'process' of how to do certain things with my vehicle. I also keep a pretty detailed record of the work I do on my car. Neat!

Target audience will be shifting towards general market. I focused on enthusiasts at first as I know that market very well.

Tutorials can be private and could be shared if you wish; we offer 'export' functionality.

Down the line, as we collect more data (makes things interesting) there is a plan to introduce some sort of crowd sourcing aspect.

Feel free to bug me if you have any questions! Thanks for checking it out.

The other week I was thinking about buying a car to save money on commuting... And I'm sure I am not the only one.

I definitely see value in this tool. I have 0 experience with cars, and I am sure I have no idea how much does it cost to maintain, repair. Again, I don't think I am the only one. So is £20 or so worth paying to get accurate numbers, tweaked for a specific country, with nice explanations why and how? I think so ;)

I would really consider experimenting to see if this could be monetized, whether people are willing to pay. It won't make millions, but as a passive income source? That might work.

Don't the maintenance costs go up after a while. Otherwise people would only buy 10 year old cars as they'd be like 20-30k cheaper :D

The problem there is, I have no idea how to model maintenance costs! The field is basically a hack. If there is a better way, I'd like to do that.

It looks like AAA releases some data about that:

http://newsroom.aaa.com/2013/04/cost-of-owning-and-operating...

That said, I would never endorse AAA to provide money to. They were horrible towards me as a claimant.

It's also model-contingent and sometimes luck. My SAAB has required about $1,600 a year for repairs and maintenance in the past three years I've owned it, but my wife's Honda Element has required only new tires ($500) and brake rotors and pads ($200, DIY install) since it rolled off the assembly line.

Depends on how deep you want to go, but you could set up a table with common considerations and let users set frequency...

Most maintenance is mileage-dependent, so if you base your schedule on their monthly mileage, it could look like:

  Repair           Cost           Frequency
  ------           ----           ---------
  Oil Change        $30             5000 mi
  Tires            $500            45000 mi
  Battery          $150            30000 mi
  Timing Belt     $1500            60000 mi
  Misc             $200            15000 mi

Do the math in impact/year and input that cost into your table...

To complicate matters, some of those items are time or distance items. Oil and battery, for instance, need to be changed after a certain amount of time regardless of the miles driven. Not too hard to work into the pricing algorithm, though.

If you wanted go get even more creative, you could get users to crowdsource scheduled maintenance tables for their vehicles. It's all in the owner's manual and just a matter of transcribing it.

What if you were to partner with someone like truedelta.com?

>'If there is a better way, I'd like to do that.'

I gave this a little thought a few years ago, it's an interesting problem in.

There are lots of variables generated (estimates) and validated (bills) by disparate sources (dealerships / independents) which are not at all inclined to share. Though some vehicle owners do share I'm not sure I've ever seen anything more prone to exaggeration and selective reporting.

Folks like Mitchell [1] surely have loads of great information obtained via the usage of their shop-facing products, but as far as I know they don't offer a data service.

Also, standard maintenance items like brakes, fluids and tires aren't terribly interesting or costly. The really noteworthy maintenance concerns are the brand and model specific time-bombs like plastic cooling system bits in various BMWs.

Coming up with a way to collect or somehow infer these things would be a killer feature for any service of this sort.

1: http://en.wikipedia.org/wiki/Mitchell_International,_Inc.

I take contention with your first sentence as if you find a good enough mechanic, you'll be able to ask them to see what their cost is with parts. They'll be very level with you & show you what they get as far as dealer pricing. You're right that this is still a variable in & of itself as Joe Schmo might get completely different pricing than STS Tire or (Insert Chain Here).

But, I digress. You can easily look up the cost estimates for a lot of stuff like this, but it still depends on the "skill level" of a given mechanic. I've pulled the time required to do certain jobs on my car saying it's a 1.3 or 4.5 hour job when the mechanic I was talking to said it's double that, or half that given you know some tricks.

Lastly, model specific time-bombs are becoming more obvious as people complain about a given part (see carcomplaints.com) - it's just a shame it's hard to get access to the relevant TSBs for a vehicle unless you have access to the right networks.

>'I take contention with your first sentence as if you find a good enough mechanic, you'll be able to ask them to see what their cost is with parts.'

Sure.

I think we're talking about totally different things.

The question isn't whether you can get one mechanic to share with you or even drive/call around and get dozens to share. It's how to get large numbers of them from all over to share continuously in some hands-off fashion so that you can present recent, accurate data to a site user regardless of their location, year, make and model without loads of legwork.

Obviously, narrowing the scope could help with that.

>'You can easily look up the cost estimates for a lot of stuff like this...'

Again, the issue isn't how to fact-check an estimate or haggle for a deal as an individual with a car or two. It's how build a service out of it in a way that scales for large numbers of people and all manner of vehicles - in the US that's 200-300 models per year.

That implies automation and building any sort of business from it implies using legitimate, reliable, legal sources.

I think that's very very difficult to do to be honest, though. There are variables such as prices based on location, and within that location, what the shop or store wants to charge you. The same person with the same car can go to the same location twice for an oil change & get two different prices.

I know because I've had it happen to me. The price differed by a few dollars, but still. Take it one step further & any shop that's established will have tools built specific to them where they can charge whatever they want on a sliding scale. Go to STS Tire in a nice, fancy area. Ask for a quote on a particular service. Note the price. Ask for the same work to be done in an impoverished area & ask for the same work to be done. The price will be radically different, especially if the shop owner isn't greedy.

I think you're on to something, I just think it would take a substantial amount of work but figuring out how to do that in the most efficient way to collect all those data points makes everything easier. Smells like a project to me. :D

As a short-term stopgap, maybe create a "appreciation" field for maintenance to go up by a certain percent each year?

In any case, thanks for the calculator! I'll probably be using it alongside others to help me buy a car soon.

This is a great tool! It'd be awesome to add lease information into it as well, to help with the lease-vs-buy decision.

I too would like to evaluate lease versus buy.

The default gas price is shown as $3.40/gallon. Where are you getting gas that cheap??

That's pretty typical for Houston at least.

Wow, that's amazing. Was $4.17 in Chicago when I filled up a few days ago.

Ah, the 8$ per gallon price in Europe. And I find it affordable…

It's $1.97 where I live (Dubai). You should see what that does to the traffic!

Just paid $4.99 in LA last night.

Texas, probably. It's been hovering between 3.40-3.60 the past couple months.

Thanks!

Feature suggestion: Metric units (for every other country that is not the USA)

I completely agree. This tool is useless to me in Australia, unless I convert all the units.

Even within the same measurement system different countries express units in different ways (e.g. Imperial Gallon vs US Gallon; and MPG, L/100km & Km/Litre).

I would be curious to see this versus an electric car cost calculator

You could approximate by saying that a gallon is a kWh of energy. set up the Fuel Price per Gallon to electricity cost $0.1(a), 3(b) MPG, and a much lower maintenance cost...

(a) 10 cents per kWh (depends on where you live), (b) 3 miles per kWh (or 333Wh/mile, depends on the actual EV)

> and a much lower maintenance cost...

As long as your figures include replacing the batteries every so often....

Good idea, unfortunately barely usable on my ipad - entry text fields are too slow

Nice idea, but I actually think Edmund's TCO is a bit more comprehensive and uses real world data for each car.

http://www.edmunds.com/porsche/911/2011/tco.html?style=10138...

Nice aspirational example there with the 911...

Didn't know about TCO, but I like the idea. First glance impressions of the Edmunds tool:

  - Mileage assumes 15k/year, which may not be accurate for a
  large part of the population (I only drive 10k/year, which
  affects depreciation, maintenance, etc.)
  
  - The math governing financing is opaque, as is maintenance
  and repairs. I'd like to know what they're assuming will
  break, need service, etc. Esp. because it's non-linear. Not
  sure how I'm going to accrue $500 in maintenance in the first
  year on my 2012 Rav4 - all it needs are oil changes (3 x $30).
  
  - Insurance also strikes me as astonishingly high (over
  double what I'd pay/year).

All of this is to say, I see the value of a more flexible and transparent calculator to meet individual needs...

There's a reasonably high chance that you will need to replace the tires fairly soon after buying a used car (which is how I read it working). That's most of that $500.

> Nice idea, but I actually think Edmund's TCO is a bit more comprehensive

It would be astonishing if that weren't the case: one is a weekend hack while the other is a major product in the space.

A good Show HN discussion should engage with a project at its own level. When a comment is too critical, or raises the bar too high, it tends to shut the discussion down (i.e. why bother). Our goal in these threads should be to foster further creativity.

I'm speaking generally—don't mean to pick on one comment (and certainly not to impugn your intent). It's more that this is on our minds a lot right now, and we'll probably post something much more detailed about it.

such caclulator. very caclate.

Reply on news.ycombinator.com