PDA

View Full Version : Simple Decision Modelling


bingie
01-28-2008, 04:55 AM
I am looking at taking a trip and would like to employ some sort of decision-making equation to decide the trip with the best value for the dollar.

I have 9 options for my trip, with varying durations. The trips also have a different amount for the cost of the flight. The only constant is the trip_cost_per_day. This is approximatly $100.

So, the total cost for my trip would be calculated as total_cost = $_flight + #_days * trip_cost_per_day.

Now, here is my delimma. If the trip is (for example) 4 days long and the flight is $600 then the total trip cost is 1000, or if the trip is 7 days long and the flight is $300 then the total trip cost is 1000. Clearly the second trip is a better value as it is 3 days longer.

Clearly here (i think) I can just devide the total_cost by the #_days and determine the cost per day - the lower of which is a better deal (1000/4 = 250 vs 1000/7 = 142). This works for trips with the same number of days.

But, how do I employ this tactic on trips with DIFFERENT number of days?

Here is my data:

option #_days $_flight
3 8 489
10 8 750
4 9 553
7 9 489
1 10 383
6 10 695
8 10 383
2 11 413
5 11 498
9 12 489

I can easily determine the best option within the groups with the same number of days, but then to determin the best option between the results.

I was thinking along the line of:

$_flight2 - $_flight1 >= (#_days2 - #_days1)*trip_cost_per_day

and set rules based on if the equation evaluation to > < or =.

But if this is the correct method, i am unsure how to interpret the results.

Any thoughts?

Thanks =)

HallsofIvy
01-28-2008, 12:16 PM
What you are saying sounds perfectly reasonable to me. Calculate the total cost of each trip, then divide by the number of days. By your criteria, the "best trip" is the one that gives lowest "total cost per day".

(I say, "by your criteria" because one can imagine other criteria- for example if the trip MUST be at least a certain number of days or NO MORE THAN a certain number of days. If the trips were at different times of the year or to different places, that would certainly change the criteria.)

bingie
01-28-2008, 08:38 PM
Yes.

The thing that is playing on my mind is how

trip a -> flight is 500 and is 6 days long, total cost is 500+6*100 = 1100

trip b -> flight is 300 and is 8 days long, total cost is 300+8*100 = 1100

so, here trip b is a better choice. But, daily_cost_A = 1100/6 = 183 and daily_cost_b = 1100/8 = 137.5 so that indicates as well that b is a better choice....

Correct? =)

Identity
01-29-2008, 09:06 AM
The thing that is playing on my mind is how

trip a -> flight is 500 and is 6 days long, total cost is 500+6*100 = 1100

trip b -> flight is 300 and is 8 days long, total cost is 300+8*100 = 1100

so, here trip b is a better choice.


Your sums relate to 'value for money' and in that regard you are spot on. I don't see what's playing on your mind. When you initially look at the equation you see 'trip a' is a more expensive flight and you stay at your destination less days than 'trip b' which has a cheaper flight and where you stay more days, but they both amount to the same cost, so 'trip b' obviously looks more appealling.


But, daily_cost_A = 1100/6 = 183 and daily_cost_b = 1100/8 = 137.5 so that indicates as well that b is a better choice....

Correct? =)

When you then work out the cost per day value, all you are doing is showing the working for what made it look good to you in the first place (if that makes sense).