Progress

Trevor - "you are awful....but I like you...".



Actually, you did mention WORK....in terms of estimating the kill-rate of 1 pigeon per shooter per minute. Only by this could you have arrived at a duration. You could, of course, have done a reverse calculation to "backfit" the kill-rate in order to fit some random guesstimate of a duration (perhaps the client said that he wanted it cleared within 4 days)...thus assuming 2 shooters; therefore a kill-rate of 1 per minute (+contingency). Such a kill-rate doesn’t sound unreasonable.....as would 1 pigeon every TWO minutes (but that would instantly double the work, double the duration and double the cost).....’cos you’ve halved the efficiency. In this case, doubling the resource would halve the duration......but in the case of the car journey, two cars driving at 50Mph does not enable you to reach your destination in 1 hour.



Back-fitting is a favourite hobby of PMs/Directors. In fact, we had a project that did precisely that {a summary level costing and duration was quoted). When I inherited the programme, I decomposed it into it’s individual deliverables...and arrived at a quantity of work that was totally unrealistic e.g 7 hours to write a Works Assembly Schedule.....yeah, right!!!



Trevor...I like jousting with you....you are a worthy opponent in the nicest sense of the word.



James.

Oh, BTW, I arrived at 4 days, 2 shooters, 3000+ pigeons etc without having to mention Work or Cost.

They both tumble out of the output side of the calculation as a consequence of the stuff injected at the input side.

64 Hours, and at $50/hour is $3200.

James, don’t imagine opponents where there might not be any. I’m on your side. We are on the same page. I am sure that you know that I know that when the plan is done, after shaving off all the spikes, bumps and hollows, it all has to stack up and balance.

You say "Ultimately, it is the WORK.." and I could not agree more, but initially, I think (just because it works for me), it is the duration.

I agree that bottom-up is essential at some point, but not necessarily as a starting point.



Re the pigeons, assuming they are flying around in the town plaza, I would have estimated 1 day (or 2, it doesn’t matter). Then I would have assigned 2 guys (or 3, it doesn’t matter). Then I would have checked that a reasonable expectation (from TV) might be 1 dead pigeon per minute per shooter, ie 8 x 60 x 2 = 960 dead pigeons.

Seeing this is a bit less than 1/3 of the number of dead pigeons required, I would then have tossed up the options re more days, more shooters, faster guns or a combination.

There’s no rush. The pigeons aren’t going anywhere. Best to proceed in an orderly fashion, especially when using firearms in a crowded town plaza in daylight. No need for high tech equipment. Let’s call it 4 days.

I build the plan as I go, and do not wait until I have a perfect set of numbers in my head before I start.

Just because the job could be done in 37 minutes, does it have to be, and does it have to be planned to do it this fast?

During WW2 Kaiser built a Liberty ship in 4 days, but only one of them as a demo. The others were built pretty quick too but longer than the shortest possible.

Trevor,



Thank you for your enlightenment. You have now just confirmed why so many projects fall behind schedule. I agree wholeheartedley that anyone can guess the duration...and that’s where most people start...and that’s why people aren’t planners and are totally suprised when they fall behind schedule, miss their milestones and go over-budget. It’s all ’cos they completely forget about the fundamental element called WORK, and the fact someone has to DO IT. Ultimately, it is the WORK that drives the Duration (even concrete curing can be defined as WORK (chemical work)).



It is fine to estimate the duration if you have a reasonably good grasp of the work involved...but as you noticed with your "pigeon" project, I made the estimate from the bottom-up and arrived at a "duration" that was so short that it surprised even me. However, that was ONLY because I defined the scope of work beforehand. Had the pigeons been free-flying then it would have been a whole different ball-game. The big problem generally, is that a PM/Director gets it into his head that just because the client wants the job done in six weeks, it doesn’t mean it can be done in six weeks. Maybe that’s true if one ASSUMES that we have 250 people.....but when you have only 5, then the equations don’t quite stack-up. The arithmetic is so elementary that you’d think that "intelligent people", like engineers and PMs, would have no problem in understanding it......but they do. They often live in cloud-cuckoo-land and when they get so wrapped-up in their little worlds, I use the analogy that, if you want to drive 100 miles in 1 hour, and your car does a maximum speed of 50Mph, then it’s going to take you two hours....regardless of much you WISH for it to take only one. It just won’t happen.



The moral of the story: Duration is only ever a result of Work, Resource and true Logic. They must all tie-up. If they don’t, then that’s because one or more of the variables has changed.

Anoon, Trevor, Rap James,



I can’t believe you guys discussing the process of duration estimates



Have you not read the basic of PERT



D = (a+4t+b)/6



t is the mean of >>>>



This is what the polaris missile and du pont all about that why we were able to come up with PERT/CPM.



The problem now is really how to do this with 25,000 activities or 5,000 activities with planners ???? housewives, labourers, planning jockeys



So so sooooo, they came up with guesstimates from managers, etc, etc. sometimes teaboys, office cleaners....



OH MY where are we going?????

i believe you were also starting with "work" Trevor (even if you said you don’t), how can you start or type Task Names if you don’t know the work? I think the real problem lies always with the initial estimation of Duration(s) which only the first Planner who made the program knew and not a collective estimate of the team who will do the real work, but who is doing a collective estimate in the initial stage of the project anyway? I guess you are right, i even encounter a project (team) who cannot figure out what resources to hire.

Anoon, James,

That’s my point. Only my POV. Duration estimate, however you come up with it, comes first. Before Resources, if any. Before Work, if any. Before Cost, if any. You can come up with an estimate of duration "scientifically" or you can estimate it. The first pass estimate doesn’t have to be the final estimate, and it probably won’t be, but it gives us a way to get started. The whole of the planning process is an iterative process which converges to a balanced solution after a few passes. After some progress has been made, we can estimate the remaining duration the same way as the first pass. Anoon, you said before, if I understand correctly what you said, that you start with the Work. I don’t. I have seen people try to start with the Work and it just causes confusion.

Trevor,



"after typing the name of the task", of course next is Duration, you can put whatever figure you want, but it doesn’t mean that it is correct and you can nail it down (as Fixed Duration), without figuring out the probable resources and productivity rates. I always believe that a realistic task duration must come from a calculated productivity rate (applicable and practicable) and not the one that comes first in mind after typing the Task Name or the one dictated by a drunk PM.

Gentlemen

What most folk are missing is that we are attempting to estimate the duration for a task, given a scope of work and in the perfect world the likely resources to be aloocated to that task. From this data we are able to estimate the duration.



Trevor is quite right with his very able attempts to get people to understand the estimating process.



We are Planners and Schedulers and to be able to understand how a duration is estimated. We should know where a day starts and ends especially when reporting progress.



Project Managers talk in Elasped time and not duration and believe me there is a huge difference.



The way Primavera, MSP and Open Plan work is by actually assuming that the duration is an elapsed time. MSP further complicates it by being manupilating the calendars and hours per day which are worked and as I have found to my chagrin I can have 422 working days as a duration in one calendar year, because I was working a five day week with a ten hour and the global calender assumed a forty hour week.

Try working out the progress then.

Of course one doesn’t have these sort of problems using Micro Planner X-Pert.



Raf

Trevor,



I really don’t know what you’re trying to get at. What you input into the programme is just the result of the aforementioned exercise ("...the hard, slow complicated way"..."). Are you saying that you DON’T have to figure how you’re going to execute the activity prior to inputting into the programme? What have I been doing with my life, all this time???



All MSP allows you to input are the variables of Work, Duration and Units (in whatever configuration you consider best suitable). The software itself is totally ignorant of how the activity is to be executed, and it won’t tell you that you can’t fire/reload simultaneously whilst using one gun or just one person. It will do whatever you programme it to, and you can amend it to show that you can actually achieve the impossible (and this is where the practical experience of the programmer comes into play).



Whether you care to input the data using a Fixed Unit, Fixed Duration or Fixed Work can actually depend upon how you think the activity will evolve. Personally, I never use Fixed Duration....it tends to screw-up everything during programme husbanding.



C’mon, Trevor...I’m bored. Put us all out of our misery, and give us the benefit of your razor-sharp intellect.

OK, that’s the hard, slow, complicated way, but let’s say that’s how you do it.

After you type in the name of the task in MSP, Which bit of data do you put in next, and then what?

If I use my imagination - just for fun, I can assure you that it won’t involve brick-walls and pigeons. Maybe green fields, sheep, ribbed wellington boots and rubber wetsuits :-)



OK, if you press me. Pigeons. Scenario 1 Assumptions: All cooped-up, all on one horizontal plane standing beside each other. 12 bore shotgun, double barrelled; gunner is lying down firing in the same plane as the pigeons (maximum kill rate per shot - estimated 10 per barrel.). Net placed over them so they can’t fly away. Just me.

Activities: Aim-Fire: Re-Load: Aim-Fire. Estimated timefor activities 5 secs per set of barrels, 10 secs reload.



3000/20 = 150 "sets of barrels" but only 149 "Reloads". Therefore: (150 x 5)+ (149 x 10) = 750 + 1490 = 2240 seconds. 2240/3600 = 37.3 minutes. This would equate to both Work and Duration, owing to the fact that I’d be occupied 100% of my time.



Uncertainties: Kill rate might be 20 per barrel or only 5. Barrels might overheat at the firing rate of every 10 seconds. Never having done this before, I really have no idea.





Pigeons: Scenario 2. Assume whatever you want (still just me, doing the shooting, though). The Work and Duration will be equal to howsover long it takes. The only thing that will change drastically is the kill rate per shot (output per unit effort. Therefore, you will need more shots and reloads and cartridges....thus it all takes a lot longer). However, as long as you are working 100% of the time, then W & D will be equal.



Pigeons: Scenario 3. I start talking to them about planning. All dead in 5 seconds.



James.

Just for fun, use your imagination. Imagine you know the scope. You’ll never be an estimator if you can’t make an estimate.

Trevor - did the "bricks" question catch you out? The reason for the "how many layers" question was because of the fact that the configuration of the bricks defines any major constraints on how it can be done.



Reference the pigeons: Over what area are those pigeons spread? The reason for this question first: It defines the scope, major physical constraints - thus defining the method of completing the scope, therefore the work/duration, therefore the cost. If your pigeons are all cooped-up, then the approach of executing the work will be entirely different than if they were spread over a large area.



As you will note: we are reverting to Anoonimous’"define the scope" first.



Feel free to debate.

Forget the bricks.

Call it "shoot 3000 pigeons".



It makes no difference to the correct approach or the answer to the question.

First of all: How many layers of bricks?

Consider a simple case of just one simple task, "lay 10000 bricks".

How would you approach populating the rest of the duration, work and cost information, in what order and with what data?

i believe that you can never nail down Duration first, why? Duration must come from the guesstimated amount of work, so it is always Work first or the Scope. What always happens is, you got a guesstimated Duration from a guesstimated Scope with an added guesstimated Cost.

I think I will have to agree with Trevor Rabey.



Most of the time, during a project, Duration, cost and work will be control and monitor seperately, either using planing software or spread sheets.



I had been using MSP and Primavera and realized that both planning software have a different way of calculating the duration, cost and work. Anyway, I think that this is not important but we need to understand the basic of project management and control and make those as basis to measure and update the progress. As long as we are answerable to the client, who cares?

Hhhmmmm.... thinks for a week, sound of cogs turning, then says:



Duration probably isn’t all we want but it may be all we have, and even if we have more it is still a good place to start when doing the tracking, monitoring, updating etc.



I like to get the Duration nailed down first and then proceed to Work and Cost, actuals and remaining.

At least with Duration, there is a chance of less argument over whether yesterday is the past, now is the present and tomorrow is the future, although apparently that is not so obvious to many people. Having no unused Duration on the left of the Status Date and no progress (ie actual duration) to the right of the status date is fairly indisputable.



Any task can (must) have duration (even zero counts) but need not have Resources, Work or Cost, or they might be in the real project but not modeled in the project plan.

A Task (or more usually the model of the Task) can have Cost with or without any Resource or Work (fixed Cost in MSP).

A Task can have Resources with or without any Work or any Cost.

A Task can have Work with or without any Resource or any Cost.

But it always has Duration, and if we are going to control the times, dates etc it has to be about Duration.



Sure, as you say, proper planning means going the whole way with Resources, Resource Costs, Fixed Costs, etc etc.

And it is always amazing no matter how many times we see it that many people think they have a plan even if it has tasks with Durations and nothing else.



You equate Cost with Work. It need not be so.



I prefer to think of Duration, Cost and Work as all more or less independent.



I say "more or less" even though, of course, more Resources and/or more Work from the same Resources usually is the way to shorten Duration, and if this does not change the total cost it probably changes the cost side cash flow profile.

Hmmmm.......



Trevor, you are quite correct in the principal application of calculating duration. However,I can only partially agree with your overall posting...because ...



Duration is all very well if that is all you want. However, Duration does not give you the workload profile....which is absolutely essential. How many times do I ask: "... the duration is 10 days, but how much of your time is actually going to be spent on it....". If the reply is 100%, then the resource is going to be fully occupied and is going to cost 10 x Qty of Working Hours per day) = X. If the answer is only 50%, then the resource is going to be under-occupied etc.etc.etc.



Planning becomes a whole-different-ball-game when it’s done properly (which includes the workload profile).....and I’m sure you know that. Independently, Duration and Work will each answer only one question...and that is start/finish dates (for Duration) and cost (Work). To be remotely accurate in your plan, the two MUST tie-up as closely as possible. It really is just a shame that an awful lot of PMs and Business Managers fail to make that connection at the time they give the client a quote. However, it really is fun to watch them squirm when they have "reduced the price" and I ask them which elements of the scope have also been reduced. If they say that no scope has been removed, then you get into yet another ball-game...and that is output-per-unit-of-work (efficiency)..blah, blah, blah.





James.

A Task can be tracked even if there is no Resource assigned to it and even if there is no Work associated with it. Of course, there probably is both a Resource and Work and if they are included in the network model then that is a bonus, more data more information etc, but neither are essential to the model or necessary to track progress.

You ask how the task is going.

Someone tells you that there has been 2 days of actual duration and an estimated 1 day remains. The initial 4 days estimate is irrelevant at this point. Those are the facts, that’s the picture, all that’s needed is to put the right data in the right slot.

Moutaz,

please read the first sentence of the MSP Help for the % Complete field where it defines % complete as about duration (ie actual/Total) only.

James, thanks for the support.

But why not just do what I said?

Why mess with the Work when the update is all about the duration?

Moutaz,



My dear friend, Trevor, is trying to tell you that, in MSP, the "% Complete" column refers ONLY to the task DURATION. It does NOT represent the task’s physical % complete.



Unfortunately, this is a trap into which many users fall into if they are unfamiliar with MSP. Remember, you can have completed 75% of the task, but be only 50% through the task’s duration. Therefore, the best way to represent this is by using "% Work Complete". By using WORK (hours) you may correctly represent whatever level of progress you so desire. You also have the opportunity to spread-out the REMAINING WORK so that the task finishes on the day that you desire.



HTH.



James.

If you really "need MS Project to calculate the remaining duration", stop using it because it doesn’t and it won’t.

Do you agree that Wednesday, which has not occurred yet, cannot possibly be part of the actual duration, yet that is what you are saying if you put in 75% Complete?

In MSP it is very easy. Try it. It works. It paintrs a picture of reality. It is very logical.

You just put in the actual start, actual duration, then estimate the remaining duration.

Why do you expect software to estimate the remaining duration? How can it know how long the task will take?

Does it estimate the initial duration? No, a human does that.

Let me assume you are putting 75% into the % Complete column.

Unfortunately, this is wrong and I will explain why.



Suppose your 4 day Task has a planned start on Monday, and the 4 days are Monday, Tuesday, Wednesday and Thursday.

Suppose your Status Date is end of day 2, ie Tuesday.

Suppose the Task started, ie the Actual Start Date, as planned on Monday and was worked on continuously up to the Status Date.

The Actual Duration is the maximum Actual Duration, is 2 days, Monday and Tuesday.



If you use View, Table, Tracking, and type in the Actual Start Date and the Actual Duration, MSP will calculate the % Complete as 50% because % Complete is always Actual Duration divided by Total Duration, ie 2/4.



However, if you type 75 into the % Complete column, MSP will assume, reasonably, that the task started on the planned earliest start date, ie Monday, and you have just told MSP that 3 days of actual duration have elapsed out of 4 days of total duration. The last of those 3 days must be Wednesday. Please tell me how tomorrow can possibly be part of the actual duration when it has not occurred yet?



If on the other hand you think that 2 days represents 75% of the total duration, then the total duration, ie 100%, must be 2 x (4/3) = 8/3 = 2.6666666666666667 days.

2 days of actual duration are gone, therefore Remaining duration is 0.6666667 days, ie you expect the task to finish 2/3 into tomorrow, mid afternoon.

Just type in the actual start date, the actual duration and the estimate of remaining duration. There is no need to type anything into % Complete because MSP will always calulate it for you.