I have heard a few times on GCN that 25 mm tires roll faster than 23s and 28 mm tires roll faster than 25s. How is this possible? I would think that a larger contact patch would produce more friction thereby slowing things down a bit. I must be missing something. Can someone explain this?

Where the rubber meets the road: What makes cycling tires fast? - VeloNews.com

Scroll down to "Wider tires are faster" section, near the bottom of the page.

A large contact patch doesn't increase friction for two reasons:
1-Ideally, the rubber of the contact patch isn't actually sliding against the ground, it's sitting statically on the ground.
2-Even if it was sliding on the ground, friction tends to increase with larger total forces, not larger contact regions.

Wider tires deform in a shorter, rounder, smoother region at the ground contact; as a result, less energy is wasted deforming the tire walls as the tire rolls and as it deforms from bumps.

Also, as you get toward bumpier and bumpier surfaces, lower PSI is important for low rolling resistance. Much better to absorb the bumps with tire deformation than let them transmit higher up, jostling the bike and rider around! And wide tires are able to handle lower PSI.

This.

Often repeated but this part is not correct. With decent tires, differences in Hysteresis losses due to contact patch shape are not nearly so significant as people think, but this is not my point.

Even running equal pressure, the wider tire has almost the same length contact patch. But you run a wider tire at lower pressure (or else you loose most benefit), in which case the length of the contact patch is significantly longer. The tire casing flexes more, but real world testing indicates this is not a significant factor.

Smooth steel drum tests with contemporary performance tires say there is slightly more resistance with wider tires, but textured drums (tread plate) and real world tests say wider tires are clearly faster in terms of complete system rolling resistance.

Ah, so it's almost entirely in the transmission of dynamic factors. Interesting.

True, but I'd also point out that high friction between the tire and the road is a good thing and results in less energy loss. Think about having a tire with greatly reduced friction. Then when you push hard on the pedals for your sprint the rear wheel would spin and slide against the road surface and instead of your energy going into accelerating your bike it would go into the scrubbing action of the slipping tire tread on the road surface. And if you took a turn at speed with that tire it would slide to the side - at best again losing some energy but probably also resulting in a crash with even more loss of energy (and skin).

Energy, or 'work', is equal to the applied force times the distance over which it acts. In the case of a tire on the road the energy lost to friction is equal to the frictional force times the movement of the tread in contact with the road relative to the road surface. We want the static friction to be high enough so there is essentially no relative motion of the tread vs. the road; i.e. no slipping of the tire and therefore essentially no energy loss.

Overinflated tires are problematic but I find no difference between 23 vs 25 on flats or rolling ground. There is a slight weight penalty on long climbs, which I can feel after a few hours but nothing that drags me down much...timing is a bit slower for me. I don't use 28s so can't comment on the diff with those. Fast downhills are more confident on 25s and 25s can be inflated slightly lower, which makes them more comfortable. I'm going to use 25s on the Roubaix and 23s on the Tarmac. I currently have the same 25s on both.

I refuse to ride narrower than 28. Not racing though.

There isn't a huge night and day difference on smooth pavement. Rough surfaces require larger tires, larger tires require less pressure, less thought about route selection and are less prone to pinch flats. 23mm tires will force you to be more careful about the line you take, possibly making for a faster ride.

Yeah, those long climbs that go on for a few hours can really benefit from 23s.

I have no idea about the rolling dynamics of a 25 or even 28 over a 23, but my hands and butt tell me the ride is smoother and therefore I can go faster and longer. A quality tire of 28mm inflated to 90 lbs is perfect for me.

Here's the deal...

The [same tire] 28 inflated to the same pressure as a 25 has lower rolling resistance. The same thing with a 25 vs a 23. The thing is, however, most people might run a 28 at 90 PSI and a 23 at 110 PSI.

Also, the 28 will have more aerodynamic drag than a 23...obviously.

Just ride what you're comfortable with in the end. 23-28 is the range, I think 25 is the happy medium, and it just so happens that's what's on my road bike.

How much do you weigh, and what are the roads like?

They would benefit so much more with maybe a fat bike.

Bad. Occasional pothole when returning home on a rainy night could ruin my day. Not worth the risk - especially since I'm not racing. 28mm quality tire (gp 4000, or 4 seasons) is very quick and light, not much penalty, just the comfort/safety bonuses.

If you tell me details about you like your weight, height and your preference.Then i will help you.Thanks in advance.