Their reviews generally reflect average reviews

Double post, sorry

The difference between and 9 and 10 in this case could easily be just a case of personal opinion. They even complained about difficulty near the end of the game.

How so?  IGN gave Resistance 2 a 9.5.  It's average is an 87.  IGN gave Gears 2 a 9.5.  It's average is a 92.  Why would your rule apply to PSOM and not to IGN or some other publication?

@parasitic^^^I don't understand,,,,what does 'enemies are too smart' even mean?

didn't everyone say that GG has to improve the AI,,,now they are too smart?make up your mind people.

A good score. I expected nothing less. To be honest, I expected something more but in no way is a 9 a less then steller score. I cant wait to get this increadible gem in my hands day one. I am ready to surrender my life! :D

Doesn't anybody remember all the complaining that's happened over the years about the game's AI? Since E3 2007, people always complained about the AI. Heck, 1up (or was it EGM?) cited Halo 3 for having the smartest enemy AI and Killzone 2 for having the dumbest AI back in 2007.

As long as the AI is smart and not too cheap, I think I'll be okay.

Because PSOM have reviewed KZ2 and IGN haven't?

Your notion is not logical.  One score from a publication cannot be used to infer the likely average of a game.  PSOM NA gave MotorStorm: Pacific Rift a 6/10, a full 20 points lower than the game's average on GameRankings.  They gave Killzone 2 a 10.  Can we infer from that that most sites will give Killzone 2 20 points higher than PSOM's score?

Well they are different magazines, you are talking about the US mag, this it the UK mag, but generally I think they are very accurate, and apart from Mirror's Edge, they are usually +- 10% at most, which is good considering they do whole numbers only. Admittedly, this does not mean it will end up with an aggregate of 88%, but it means it will probably range from 95 to 85