hey guys i am learning maya with Student version and dreaming to have a Studio of my own in the future 
So i been following Digital tutors’ Tutorials on maya and when i going through those Courses The Tutor always Stays to quads and sometime uses triangles in some situations (like blending extruded faces back to the Surface) but he NEVER USES N-Gons (i hope the spells are correct ) i also knows that N-gons are a Pain
but he always mentioned about sticking to Quads all the time and never use N-gons.But when i am going through the Quick Start to Modeling in Maya VOL.4…he uses a Tool called bevel Plus… and it makes a awesome Shape out of the curves But in the front and back sides (not the up and the down sides) it makes a HUGE N-Gon ( i thought so cuz the edge flow doesen’t go through and i can see it has a Lot of vertix without edges) i uploaded the Pic
So the when we work in the Indrusty it is very hard to get the job done with N-gons (from what i heard) so in this one is it going to be problem (not the course i mean think that as a real life project if u can )
Please give In depth answer i am a noob and still learning hope ya will help me 
here is the images
http://prntscr.com/5e5g8j
QUADS QUADS?????? Why not in this one?
Quads are prefered because they produce predictable subdivision results. They also make it faster to select edgeloops and edge rings, faster selection when uv mapping, rigging and proper deformation when the mesh is animated.
That doesnt mean that tris and 5+ sided polygons are a bad thing. Sometimes they will work just as well, it all depends on the situation. When you are working with subdivision the tris and ngons will be turned to quads anyway, just try it and you will see, that’s what subd does. You could also have an all quad mesh containing poles that are not well placed and introduce artifacts.
Leaving tris and ngons on flat surfaces is usually the place to hide artifacts (on flat surfaces you could have a 60 sided polygon and have no visible artifacts provided you have support loops) In the example you’ve shown there is absolutely no quads on the sides, it’s one giant mess, if you were to throw on subdivision you would see a terrible mess because that flat surface doesn’t have proper support loops. The subd will have a hard time solving to quads. You could select the side ngons and simply create an “inset” and now you would have a row of quads that matches the top surface, the middle ngon will be perfectly flat and won’t show any mess.
I hope that made sense.
When you model something intended for subdivision, you need to keep it to quads, and in a pinch, tris. That’s the only way to make sure you can control the mesh flow after it’s subdivided.
If you’re modelling something that doesn’t need to be subdivided, in theory the whole quads and ngons thing can be thrown out of the window. You still want to generally keep it to quads for workflow purposes, but you can totally use large, flat ngons without worry.
Here’s an example:
The two models on the right are roughly identical looking, but the middle one is a subdivision of the model on the left. The right one is modelled directly using a cylinder and bevelled edges. Notice the big n-gon on top because I didn’t bother to fill it in manually. Doesn’t matter too much.
The subdivision approach is generally easier to edit, lighter to animate and work with and easier to unwrap. It’s usually considered the ‘correct’ way to model, and it’s an important skill to master and train yourself in.
Modeling directly is generally quicker and easier, and nobody will kill you if you use booleans or other ‘cheats’ to make shapes that are hard to do with subdivision. Good for mechanical stuff with lots of small parts you need to do quickly, or for mocking up your models before doing them in subdivision.
yeah i understand but what you mean by subdivided?? is it pressing 3 or smoothing it by adding geometry or is it convert the model to SubDs? thanks been a great help please answer too