Little story about stats and sample size.

I jump onto Party yesterday, and end up on a table with Jyms. He's got position two spots to my left, and its fairly obvious that the table money is the seat between us. I start raising a lot preflop because this guy is calling a lot and folding most flops. Seems like the kind of guy who loses a lot of small pots, but when they get big he generally has a hand worthy of a big pot (a common mistake by a lot of TAGs is not identifying this and losing stacks with TPTK type hands v's them). Anyway, I play an interesting hand on another table and msg Jym about it, and conversation turns to this target. I have about 50 hands on him, at 75% fold to cbet, so like I say I'm just raising wide and cbetting a heap. Jyms has about 200 hands on him, and has him at 40% fold to cbet, so Jyms is looking to make good TP hands against him and value town him. Same fish, same table, but our adjustments are polar opposites because of stats over different sample sizes. Shortly after this I see him call two barrels with a gutshot and call river with a weak TP hand, so I think Jyms read was right (as you'd expect with larger sample), and he must have just been completely whiffing against me in the hands we played.

Just a lesson that stats dont tell the whole story, especially over small samples.