Tomorrow is the “absolutely final can’t-miss” deadline to finish our house, and Bob has not made it. In fact, like every completion date he has set, he has not so much “missed” it as “crashed through it with no signs of stopping”.
Guessing when a construction project ought to be finished is notoriously difficult. I’m not the first person to be frustrated with a contractor that is constantly missing the deadlines he himself set. Back when the U.S. Navy had commissioned the first Polaris submarine, they decided that rather than rely on the crystal ball of intuitive guessing based on past performance, they would turn some of their vaunted engineering skills toward creating some sort of due-date algorithm, some way to take the educated guesswork of all the various contractors and subcontractors and scientifically estimate when they will REALLY be finished. They invented the PERT chart.
So I decided to plug in my own data and guesses and see when PERT thinks our house will be done.
The Optimistic Guess: the soonest it might be finished (1% chance of happening) = 29 days from Friday Sept.14. This assumes that Carlos finishes the scratch coat this Friday, and during the 28-day stucco process, everything else manages to be completed… plus a single day to remove all the scaffolding. This finish date works out to October 13.
The Realistic Guess: human opinion as to when it will “most likely” be finished. Without going into all the complexities, and without listening to Bob’s guesses (his estimates are consistently sooner than the 1% guess, which is why he keeps blasting past his deadlines so dramatically) I’ll hazard 40 days from this Friday. This finish date works out to October 24.
The Pessimistic Guess: the latest it might be finished (1% chance of happening): it’s painful to say this, but based on the calendar distance from Bob’s previous confident deadlines since we made this time-critical (Christmas 2006, March 2007, June 15 2007), it will be a long time. Mitigated by how much has been accomplished since those deadlines, I’d put the Pessimistic guess at 120 days. This finish date works out to January 12, 2008.
The Expected Guess: Here’s the magic of PERT’s prognostication algorithm: none of these other guesses are reliable. The Expected Guess is (O + (4xR) + P) divided by 6 = (29+160+120)/6 = 51.5 days! This finish date works out to November 5.
Thus begins our test of PERT. I sure hope it is right. November 5th would be a good time to be able to move into our house.
Anytime after Thanksgiving and I think my poor wife will have a breakdown.
Well, it’s after Thanksgiving, and she did not have a breakdown. I think God prepared her for the letdown somehow. And neither of us think we’ll be done before Christmas.
PERT is only as good as the data you feed into it, I guess… not a magical-technological oracle that tells the future.
So I want to run the numbers through it again, and see what its new guess will be. But not now, I have to get some sleep before fetching the Sparrow from Rick Smith tomorrow.