Can you see what’s wrong with this image?

That’s a trick question. There is nothing wrong with anything in the image. It’s all fine. You could get nit-picky and say that the brick color is wrong (but you didn’t know that). Or you could say that the windows and doors should have been modeled with the Curtain Wall Tool (they weren’t). Or you might think the trusses shouldn’t be Complex Profile Beams with a length of an inch and a half (I prefer this solution to using the trussmaker). Probably what you noticed is that the main floor is Story -1 instead of Story 1 (we’re using my template which is based on the USA template, so no story 0). But if you called out any of those things, you’d be incorrect. Because there is nothing wrong with anything in this image.

It might hurt your latent OCD tendencies to see -1 Main floor. Or to have the Elevation of the main floor not equal 0′-0″. You might be deep into converting your template to be COBie compliant, or working hard to make sure exchange with Revit users is streamlined. But for this project, none of that matters. And in fact as I built the site mesh (crudely off of a scan of a hand drawing from 1985), having the main floor set to 312′ made my life much easier than if it had been 0′-0″. Having it be the real number was better. 312′ matters infinitely more than -1. Especially for this project. It is important that you understand how to use Project Location and Reference Levels, because that offers more robust solutions, and would make working with the mesh just as easy. But you also need to understand the difference between knowing the formally correct answer and knowing what are the acceptable solutions for the given task.

The generic brick Surface might hurt your eyes, but the real brick color is an even more hideous tan—and absolutely getting replaced. The client doesn’t need to pay us to get the existing color right. They know what the building looks like now, including the mansard roof we are ripping off. So why bother modeling what they already know, and what is definitely getting removed? Plus we have the original thirty year old construction documents; if they really need to see a drawing, we can show them that. What they need to see, if anything at this point, is what’s remaining. What we need to model to start designing are the essentials of what we are working from. The starting condition for this project is not in fact the existing form, but after a pre-defined series of demolition (including all interior walls). On a bigger fee project, or a more advanced BIM project we might model what’s getting demolished to understand quantities, or to show options of different demo solutions. But we know what’s going. The bulk of what is getting demoed is a known quantity that can be easily documented with a few dashed lines and notes, if that.

This model is ready for us to design

All I need for nowKnowing what to ignore is as important as knowning what not to ignore. Here’s a reply to a conversation I was having recently: “I agree that ignoring index numbers seems to be the best was forward. Managing a perfectly chronological index would be a challenge.” For some users that is crazy advice. But for others it is perfectly sound and logical. I’ve talked about managing complexity before. All of these examples are about complexity and about information management. If you don’t see index number issues showing up, don’t stress about them. Forget they exist. If the story number doesn’t matter for the project you’re working on, don’t worry if it’s goofy. To spend time fixing that sort of thing is a waste of your efforts. Who will benefit? Who will care? Who will give you an award for perfect alignment? I sure won’t.

If you are jumping up and down right now screaming at your computer, I’ve probably listed something that does matter to your work. If that’s the case, then don’t ignore that piece of data. Don’t let it slide. Use this article to explain the importance of doing those things correctly. For everything I listed above that I say to ignore, I could easily give you examples of when one shouldn’t ignore those issues. And I could give you dozens more examples of other things that always have to be right. Well for my workflows they always need to be correct.

All these scenarios hinge on my favorite Donald Rumsfeld quote. If you understand the importance of your decisions and their relationship to the known and unknown, you will be fine. Maintain awareness of what is understood, what is required, and then don’t get trapped by silly things like treating unimportant decisions as errors that need to be corrected. Just because BIM and ARCHICAD are super powerful and light years ahead of the old ways, that doesn’t mean you need to harness all that power on every project at all times.

For more, read the follow up post:

Definition of WrongAre you following GRAPHISOFT North America on Twitter? Click Here to keep track of all the latest ARCHICAD News in North America (and beyond). For another perspective on this concept of fighting the wrong battles, check out my recent article Basic Unchangeable Decisions. Remember, just because sometimes things that look wrong aren’t actually wrong—or they are wrong, but don’t need fixing—, there are also plenty of times when things definitely are broken and need fixing. Understand the difference and stop wasting your time fixing unnecessary stuff


  1. eduardo

    The only argument for not fixing what is obviously wrong that I can use here is that not caring about decisions now (or doing it in a logical way) is that somewhere down the line you might get bitten by not remembering why you did an exception in the project.

    The first commandment I teach for using AC in a multiuser and/or production environment is to never change anything that you assume is wrong if you don’t understand why it is that way. Assume that there was a reason it was done that way (Main Story = -1) and that either you don’t remember (YET) the way or have access to the person that did it. This works as long as you don’t make mistakes on purpose or don’t fix them when they happen.

    One thing I like about AC is that most of the times doing something wrong takes more time than doing it right the first time. A good AC user learns that the phrase “We will fix it later” means twice/four times as much work later. Also if you have a good template there should not be “unnecessary” stuff that needs fixing just exceptions because of project decisions.

  2. Jared Banks

    Great advice. I definitely agree. Not rushing to fix ‘errors’ is so important. I probably should write another post that just focuses on how to understand what is an error that needs fixing. And what is ignorable. Because I know I say this above, but I am absolutely not condoning ignoring problems. I’m just aiming to get people to realize that it’s not a black and white issue of wrong or right; of fix or pay the price.

    So questions that should be asked:

    1) Is it superficial and internally focused?
    2) Will it affect the work of others: Internally? Externally?
    3) Do you understand why it is wrong? Are you sure?
    4) Did you cause the error? Do you know who did? Have you talked to them?
    5) What are the known consequences of fixing it?
    6) What are the known consequences of not fixing it?
    7) What are the potential unknowns of fixing or not fixing it?
    8) What is the potential scope for unknown unknowns as a result of fixing or not fixing? Seriously I love that Donald Rumsfeld quote.
    9) How long will the fix take? How much work will it disrupt?
    10) How easy will it be to revert the change if you mess up fixing the mistake? How confident are you in your backups?

    Okay now I definitely need to write that post! What else am I missing? How do you evaluate these decisions?

  3. Brian Spears

    Did you notice the Story (or is it Storey?) Settings in Rob Jackson’s post the other day on COBie workflow? His ‘Ground Floor’ (or Main Level?) was Index #5. I was intrigued. I’m sure there is a finite number of different ways to handle Stories and how they relate to Project Location and Altitude and Sea Level (and probably IFC and COBie Standards), but I’m sure it is quite a few…

    • Jared Banks

      I have had that post open in a tab for 2 days, but haven’t had a time to read it—today hopefully. I have talked to Rob about that though. His site is always the zero story and everything goes up from that. He never has negative stories. That way 0 is 0 is 0 across all projects. It’s an interesting/good solution. It would be nice if we had a zero story. Or the option for it.

      What I really like about his solution is that it completely separates story number from what we call the floor plan we are working on. The story number is a bit arbitrary, and his system puts more distance between that internal number and what matters (the name of the story). I have already started a post on story settings and vertical location of projects, so I’ll have to investigate more.

      It also has me thinking that in my ARCHICAD 19 template, I’ll have to think harder about stories. Should we in the USA start at 1 and go up, in a similar fashion? I have a hunch the answer is yes.

  4. Cary Westerbeck

    Fun to see our little restaurant project become fodder for an informative post. To be fair to Jared, I was the one who started the as-builts for this and accidentally set up the first floor as -1. I’m usually better than that, but I was in a hurry, so I made a mistake. Jared made me feel ok with it though, and as he said, it won’t matter. It will be more important that we work with actual site elevations in this project, as it often is working on hilly sites in the northwest where site grading, civil work, landscaping and other engineering is part of the scope. In those cases an artificial 0′-0″ is actually incorrect and unhelpful, as we need to know real world elevations for everything.

    I admit I don’t use reference levels, so I just use real elevations for everything and build the mesh with the site contours from the survey. Seems very straightforward I guess, and sort of WYSIWYG.

    As always, thanks for an informative post, Jared.

    • Jared Banks

      Don’t throw yourself under the bus! Teams take ownership of mistakes and awesomeness collectively. Check out the link to the post Brian and I are talking about. I think the result of all this conversation is that while neither -1 or 1 is right or wrong, I think we’re going to update our standards for 19. If 1 is always site plan, there are no negative stories, and whatever is the ground floor = the proper elevation, I think we’ll be setting ourselves up for some good stuff. We can talk more tonight at the user group! I kind of like things only growing up from the site.

  5. Link

    It’s not the story number that hits my OCD as much as it is the elevation! Arghhh! 🙂

    • Jared Banks

      Ask Rob about that. Or I will. I think I’m about to be converted to 0′-0″ being sea level, almost. Just need to find the time to experiment. I don’t think I will be working on any live projects for a few weeks because of travel, so it might be a little bit.

    • Cary Westerbeck

      So, is the elevation being correct relative to sea level just a northwest U.S. or west coast thing? That’s how I’ve done it at Seattle area firms and it seems to be pretty standard around here if you look at other local firms. We use *actual* elevations. Sure, we dimension things floor to floor and such and show 10′-0″ floor to floor, for example, but major reference datums for each floor are called out as actual elevations. This helps with all the grading work we have to do on the many steep sites around here. But it’s just always been convention in the firms I’ve worked for. Is this unconventional?

      • Daniel Lindahl

        Its not just the US northwest, this is a convention worldwide as far as I know. In Sweden we measure from sea level, and here in Australia we measure the levels from AHD (Australian Height Datum) which is an adopted notional sea level. Of course with global warming this might have to be adjusted down the track 🙂

  6. Patrick May

    My preference has always been to keep the main floor at 0′-0″ and deal with actual altitude/elevation in the reference levels. By setting section/elevation markers to reference the datum you get accurate marker settings, plus the altitude is set for site analysis in the project location, so no need for actual altitude modeling.
    But as you say, its not a mistake, so no need to jump all over correcting it. I do have a job that is set up exactly this way and is just starting up. It is the only project in the office set up this way, so I’ll probably “fix” it, just for consistency between projects for anyone else joining the project later.


Submit a Comment

Your email address will not be published. Required fields are marked *

Work from anywhere with secure, real-time collaboration, thanks to BIMcloud!   Find out more