There's a way to virtually guarantee that the next T-shirt you buy online will fit you perfectly
Sure, you already know how a T-shirt should fit. You just haven't found the elusive "perfect-fitting T-shirt" yet.
Since there's really no such thing as "made to measure" T-shirts - though some companies are trying - the best method of finding a favorite is trial and error.
However, if you have a favorite t-shirt brand that already fits you at least mostly well, congratulations - you're in luck.
A new website called Threadbase.com has done all the hard comparison work for you. They've already measured all the most popular T-shirt brands in each sizes - 800 in total - to figure out exactly how each brand fits. Just because two sizes are both labeled "small" does not mean they'll fit anywhere close to the same. In fact, Threadbase found that a Zara "XL" is actually the same size as a "medium" T-shirt at J. Crew.
Here's how the database works:
You enter in a T-shirt that you already own from the drop-down menu, and specify that you would like a shirt that fits a little differently. For example: say you want it a little slimmer or longer in the hem. The database already knows the measurements of that exact shirt you entered, and will use that to compare to similar shirts in its database, giving you some other brands and sizes to try that fit that criteria.
One of these new suggestions might be your new favorite T-shirt, found without wasting time and money on trying other brands. You can also measure yourself and use those numbers to search the database, as well, if you don't have a shirt you'd like to use for comparison.
There is a little wrinkle to this by the name of manufacture variance. When a clothing garment like a shirt is manufactured in a large factory, they leave a little room for slight errors in measurements. Some brands are better than than others with this, but every company suffers a little bit from it.
To account for this, Threadbase measures at least 20 of the same shirts in the same size, and takes the average of the numbers in their final measurements.