Flight Performance
It has been widely publicised that the YF-22 was judged to be the more manoevurable of the 2, and indeed with thrust vectoring, that may well have been the case; but Northrop has kept the true performance of its aircraft very close to its chest, and at the Dem/Val phase, did not publicise performance nearly as much as Lockheed did for the YF-22. Lockheed was flamboyant, Northrop was matter-of-fact. Some did not understand the employment of a V-tail on the YF-23, they thought that it might be detrimental to high alpha controllability, but Northrop was adamant that this was not the case. It seems ridiculous to think that they would pursue a design that threatened to curtail manoevuring performance; every effort would have been made to present a design to the USAF that excelled in all areas. Too much was at stake to do otherwise. Although the allowable flight envelope at the time of the Dem/Val was understandably conservative, there were plans to open this right up, and there is reason to believe, despite the choice of tail configuration, that the Northrop aircraft had no major AoA restrictions. Regarding news that Lockheed had displayed the YF-22's ability to fly at an angle of attack of 60 degrees, Rooney said that Northrop decided that a 25 degree angle of attack was all that was needed in Dem/Val. Metz said the YF-23 could come through any angle of attack, "even backwards." He said the aircraft can regain control out of zero airspeed but it would have to fall to pick up the speed again. "No matter where it's at or oriented it will come out and start flying again," Metz said. "This airplane, as designed, has the best high angle of attack and spin characteristics of any airplane ever built by McDonnell Douglas or Northrop." Defense Daily 14 January 1991

There were, in my opinion, only 2 areas where these aircraft had distinctly different capabilities in terms of flight performance: close range ACM/post stall manoevuring, and straight line speed. Obviously, with thrust vectoring, the YF-22 dominated in the former. But take one look at the 23's shape, and you'll acknowledge that it left the YF-22 in the dust as far as speed was concerned. This plane was officially stated to have been taken to 50,000 ft and Mach 1.8, but this aircraft was capable of alot more. "It was really fast. It went... much faster... than the YF-22." (Paul Tackabury, Northrop Manager of Flight Test). One can only imagine what the YF-23/GE YF120 combo (PAV-2) must have been like.

Weapons Carriage Demonstration
Lockheed shot missiles for all to see, Northrop quietly demonstrated weapons launching on sims to the USAF. Jay Miller in his F-22 book stated that YF-23 PAV-1 carried an inert instrumented AIM-120 round for buffet tests of the weapons bay. This is unconfirmed. Rooney said his team was not concerned with the Lockheed team's decision to fire Sidewinder and AMRAAM missiles, which he says were not a requirement of dem/val. "We made a list of what we felt was important and we didn't share our list with Lockheed... and they didn't share their list with us," he said. "We didn't think that launching a very mature missile at seven-tenths Mach in level flight had any meaning whatsoever." Defense Daily 14 January 1991: Northrop Unconcerned About Missile Firings

Packaging
The total internal useable volume of the YF-23 appeared to be less than the YF-22. The YF-23's unprecedented emphasis on area ruling resulted in significant space voids in the design, whereas the YF-22 made efficient use of space bounded by the HAS requirements: it packed everything in a chunky monolithic box with no hour-glass contours. The only concession to voids are the boundary layer capture area and the subtle crease between the engine bays leading down towards the nozzles. The packaging of the YF-23 was such that it would be hard pressed to match the YF-22's weapons carrying capability without substantial changes to exterior dimensions or sacrificing internal fuel space. Contrary to popular belief, the YF-23 prototypes only ever had one bay, not two. The YF-23's bay had a maximum capacity of 5 missiles of the full-finned type: 3 AMRAAM's and 2 Sidewinders. The single weapons bay of the YF-23 actually had less total volume than the YF-22's 3 bays combined. There was not enough room to accomodate the full requirement. This was borne out in the recently released drawings of the F-23A EMD version, which show a complete redesign of the airframe. Area ruling had been relaxed significantly in exchange for greater useable internal volume, the main weapons bay was redesigned, with a second smaller bay added fwd for close range missiles. The YF-22/F-22 can accomodate 4 AMRAAM's and 2 Sidewinders, or 6 AMRAAM's and 2 Sidewinders if the AMRAAM's are of the clipped-fin variety. This capability was much closer to the USAF's requirement as stated in popular press at the time: that of 4 AMRAAM's and 4 Sidewinders.

In hindsight, now that comparisons can be made between the prototypes and EMD versions, it is apparent that Northrop's prototype was more of a 'bare-bones' hot rod, with a rudimentary weapons bay, and highly sinuous contours optimised for minimum drag, whereas the YF-22 was more representative of the final EMD F-22A. The 6 month delay that Lockheed requested from the AF SPO was put to good use making major revision of their concept, whereas Northrop made few major changes to their design, they were fundamentally happy with it. It has not been revealed how they ulitilized the extra six months.

Cockpit
Lockheed chose to install and test newly designed displays in the cockpit, helping to prove the viability of its EMD layout, whereas the Northrop prototypes used off-the-shelf F-15E type displays and instrumentation. However the HUD in the YF-22 was a very simple conventional unit, while Northrop used a wide-angle HUD, again, from an F-15E. Another difference was the way each team chose to execute the canopy, Northrop's having a cross brace where Lockheed's has none. The F-22 is the first jet in history to deploy to squadron service with a frameless canopy, building on the foundation laid down by the F-16, which was a significant step forward in pilot visibility.

Technical Risk
In terms of technical risk, Northrop went further out on a limb than Lockheed. The jet exhaust trough that Northrop designed for the YF-23 and its associated cooling system was a novel innovation, and the Air Force probably concluded that although the system was impressive, it represented greater technical risk than the thrust vectoring system used by Lockheed on the F-22. Thrust vectoring had already been developed and successfuly demonstrated on the F-15B S/MTD and was more of a known quantity. It offered a positive contribution to pitch authority that could not be matched by the Northrop approach.