Looking at several bike geometry details / fit guides, it would seem to be that one could really err on the side of choosing the right bike frame size without serious consequences. It seems that one bike size out of a set offered by xyz mfctr follows fit guide recommendations. If the surrounding frame sizes differ by 2cm, it would seem true that both a size above and a size below could be altered to fit this person just fine.
The one mistake that can't be made with the fit is that the seat tube must be allowed to get the .883 x inseam length recommendation or whatever allows that 20-30 angle at bottom dead center. This seems most easily made impossible by a true square geometry with a level top tube (for the shorter people, as the largest of 3 possible frame fits) or a compact frame (for the larger people, as the smallest of 3 possible frame fits). For example, as a 6'00 person with 34.5" inseam, my 2003 Litespeed Capella at full seat post extension puts me at the right leg angle. Unless I go with a 350mm post (currently 330), I cannot experiment with smaller angles unless buying a seat that gives a slightly higher rise. I probably have too small a frame since it's virtual top tube length is 56.3 cm. Calculations would suggest a 58.7cm frame; yet I fit the Litespeed fairly well with some mods. Thank the LBS for that idea and having the seat post 2in too low on their "custom fit" they charge $100 for.
However, frames that aren't quite as compact or square won't ever offer this screw up possibility (of not giving low upper-lower leg angle) since full seat post extension (assuming availability of a 330mm post isn't a big deal) would give possibility for more than full leg extension.
Now for top tube concerns. There are stems that range from 70mm to 140mm of varying angles. Assuming that the fork steerer tube hasn't been trimmed completely, and with the use of spacers above or below the steerer tube, a large combination of (x,y) coordinates of the handlerbar and stem junction is possible. One can see 7cm difference for a given angle, which if placing the ideal frame in the middle, gives an easy 2-3cm difference on either side for error. That accounts for the +/- one size frame difference in addition to the frame that hits the recommendation. There me be a strange angle rise resulting from putting a big person on a small bike, but the (x,y) points (z in the direction parallel to the top tube) are going to be the same between the "incorrect" and "correct" frame in the wireframe intersection points of the rider.
Yeah, cranks come in their recommended sizes differently but can be easily replaced and often cover two-three sizes. Light stems can easily be had for $10-30. I guess turning radius can be adversely affected on a unnecessarily small frame that fits within these guidelines, but I don't really see how the arguments for importance on bike fit gain credence.
One thing these ideas don't take into account is where the relative body inertial moment & center of gravity sits relative to the frame size and under different conditions. But heavier people probably don't have their weight calculated into their bike fit, and the same for light people, so those arguments could weigh against that in the same manner. Stronger riders less interested in comfortable frames might choose something smaller; low mileage riders interested in comfort might choose something large letting the frame take the abuse and suck up some energy. Preference isn't calculated into the "exacting measurement fit" either.
But as for where the relative points in space of elbows, shoulders, head, butt, knees, and feet, these points seem far more flexible than what the literature would suggest.
Changing a frame rarely seems like it would ever be an issue if one measures their own inseam and buys a frame based on the .67 formula. Erring on the + or - side would easily make the recommend points achievable.
How can one justify the $$$'s one would spend for an custom frame build? Or even for wanting the 1cm increments touted as so important by some companies.
The one mistake that can't be made with the fit is that the seat tube must be allowed to get the .883 x inseam length recommendation or whatever allows that 20-30 angle at bottom dead center. This seems most easily made impossible by a true square geometry with a level top tube (for the shorter people, as the largest of 3 possible frame fits) or a compact frame (for the larger people, as the smallest of 3 possible frame fits). For example, as a 6'00 person with 34.5" inseam, my 2003 Litespeed Capella at full seat post extension puts me at the right leg angle. Unless I go with a 350mm post (currently 330), I cannot experiment with smaller angles unless buying a seat that gives a slightly higher rise. I probably have too small a frame since it's virtual top tube length is 56.3 cm. Calculations would suggest a 58.7cm frame; yet I fit the Litespeed fairly well with some mods. Thank the LBS for that idea and having the seat post 2in too low on their "custom fit" they charge $100 for.
However, frames that aren't quite as compact or square won't ever offer this screw up possibility (of not giving low upper-lower leg angle) since full seat post extension (assuming availability of a 330mm post isn't a big deal) would give possibility for more than full leg extension.
Now for top tube concerns. There are stems that range from 70mm to 140mm of varying angles. Assuming that the fork steerer tube hasn't been trimmed completely, and with the use of spacers above or below the steerer tube, a large combination of (x,y) coordinates of the handlerbar and stem junction is possible. One can see 7cm difference for a given angle, which if placing the ideal frame in the middle, gives an easy 2-3cm difference on either side for error. That accounts for the +/- one size frame difference in addition to the frame that hits the recommendation. There me be a strange angle rise resulting from putting a big person on a small bike, but the (x,y) points (z in the direction parallel to the top tube) are going to be the same between the "incorrect" and "correct" frame in the wireframe intersection points of the rider.
Yeah, cranks come in their recommended sizes differently but can be easily replaced and often cover two-three sizes. Light stems can easily be had for $10-30. I guess turning radius can be adversely affected on a unnecessarily small frame that fits within these guidelines, but I don't really see how the arguments for importance on bike fit gain credence.
One thing these ideas don't take into account is where the relative body inertial moment & center of gravity sits relative to the frame size and under different conditions. But heavier people probably don't have their weight calculated into their bike fit, and the same for light people, so those arguments could weigh against that in the same manner. Stronger riders less interested in comfortable frames might choose something smaller; low mileage riders interested in comfort might choose something large letting the frame take the abuse and suck up some energy. Preference isn't calculated into the "exacting measurement fit" either.
But as for where the relative points in space of elbows, shoulders, head, butt, knees, and feet, these points seem far more flexible than what the literature would suggest.
Changing a frame rarely seems like it would ever be an issue if one measures their own inseam and buys a frame based on the .67 formula. Erring on the + or - side would easily make the recommend points achievable.
How can one justify the $$$'s one would spend for an custom frame build? Or even for wanting the 1cm increments touted as so important by some companies.