End of Quarter Projection: $7-15 million

[zebov]

As an engineer, if I can't know exactly numbers, I like to know bounds. Sooo... taking the fundraising so far, I've done a linear regression as well as a quadratic regression on the data (linear provides a lower bound and quadratic provides a reasonable upper bound [quadratic fits much better than power or exponential]).

This data obviously doesn't mean too much, but it does give us some bounds to look at. I'll try to update these plots every week or so to see how we're doing. My guess is we'll follow along the quadratic least-squares estimate pretty closely.


Plot showing regression versus real datapoints:

http://img86.imageshack.us/img86/965/rpdonationsshort2007102cz8.th.jpg (http://img86.imageshack.us/my.php?image=rpdonationsshort2007102cz8.jpg)


Plot showing regression to end of quarter 4:

http://img81.imageshack.us/img81/3399/rpdonationsfull20071022jk2.th.jpg (http://img81.imageshack.us/my.php?image=rpdonationsfull20071022jk2.jpg)

[born2drv]

i'm an engineer too (well i was, and i quit and i'm a jeweler now) :) but i don't think you can extrapolate that far out with so few data points. just looking at the data points it looks to me a straight line if anything, it might be slightly parabolic but not as much as your high estimate. i'd give the range more like 7-10 instead of 7-15.

which is still very decent ;)

[Original_Intent]

Nah we already had a thread on this, I am not an engineer but I know how to do statistical analysis. An 8 million dollar spread is giving yourself a lot of wiggle room. Your upper end should be 17.8 million. that could be beaten but is a reasonable "high" expectation. If you want an 8 million spread i would say 9-17 million. but if I had to tighten that it would be mostly upwards. Not because I am an eternal optimist, but because the data points to more like 14-17 million.

If I had to lay money on a 3 million dollar range, that's where I would put it.

edit: nice graphs. you may have nailed it. I would still expect it will be toward the upper part of your range. You have more data points than I used, so probably more accurate.

[Van Damme]

Nah we already had a thread on this, I am not an engineer but I know how to do statistical analysis. An 8 million dollar spread is giving yourself a lot of wiggle room. Your upper end should be 17.8 million. that could be beaten but is a reasonable "high" expectation. If you want an 8 million spread i would say 9-17 million. but if I had to tighten that it would be mostly upwards. Not because I am an eternal optimist, but because the data points to more like 14-17 million.

If I had to lay money on a 3 million dollar range, that's where I would put it.

edit: nice graphs. you may have nailed it. I would still expect it will be toward the upper part of your range. You have more data points than I used, so probably more accurate.

It would be interesting to do an analysis on total money raised to date, using Q1 Q2 and Q3 numbers as data points. This, I think, would provide a more accurate quadriatic extrapolation.

[zebov]

It would be interesting to do an analysis on total money raised to date, using Q1 Q2 and Q3 numbers as data points. This, I think, would provide a more accurate quadriatic extrapolation.

I don't think this is the case with... each quarter is kinda it's own beast with it's own life-cycle, slow at the start and sky-rocketting in the end. Whatever the case, it doesn't matter too much because we don't have data for Q1-Q3 :).

[Van Damme]

I don't think this is the case with... each quarter is kinda it's own beast with it's own life-cycle, slow at the start and sky-rocketting in the end. Whatever the case, it doesn't matter too much because we don't have data for Q1-Q3 :).

We have the ending Q1 Q2 and Q3 numbers. It is possible that every quarter is its own beast (especially with the boost week), it would be interesting to see what those data points to the the quadriatic projection.

[rs3515]

We have the ending Q1 Q2 and Q3 numbers. It is possible that every quarter is its own beast (especially with the boost week), it would be interesting to see what those data points to the the quadriatic projection.

How about rather than endlessly attempting to predict what the numbers will be, we actually get out there, talk with more people, help them understand Dr. Paul's message, and make the numbers higher?

[Chester Copperpot]

Im not an engineer.. but Ive played with a lot of Lionel trains.. some Im kinda a model railroad engineer. HA HA HA!

[zebov]

How about rather than endlessly attempting to predict what the numbers will be, we actually get out there, talk with more people, help them understand Dr. Paul's message, and make the numbers higher?

*sigh* I very much dislike people saying this. You don't realize how much work I do. Seriously. If you'd like to compare "Ron Paul Resumes," feel free to send me yours, but don't assume that all I've ever done for Ron Paul is make a plot in excel. You let me know how else I can be of assistance to Ron Paul while spending time at home with my wife (already emailed my Meetup, the campus organization of ~70 students I started at Purdue and am President of, and the ~120 members of the facebook group I am the administrator of, today about the need for donations today) and I'll be glad to help. But DON'T tell me I'm not doing anything.