The only time flagstick in helps is one dead center hits that would have been 5+ feet past the hole. Leaving the flagstick in hurts more than helps.

As Prof. Mase’s data and study establish, it is much more likely that a putt would be striking the flagstick off-center as opposed to dead on. According to our probability calculations above, putts that contact the flagstick off-center with a glancing blow occur almost three times more often (72 percent vs. 28 percent). So does the flagstick help in these situations? Once again, the answer is an overwhelming “No.” In Mase’s testing, he chose a pace where the ball was rolling 4 1⁄2 feet beyond the hole. (Remember that at 2 1⁄2 feet by the hole, 100 percent of the putts were made whether the flagstick was in or out.) Here are the results of the off-center testing performed with the Perfect Putter at Cal Poly.

[...]

For a putt traveling 4-1/2 feet past the hole. The total numbers are as follows:

Fiberglass pin: 55/90

Pin out: 81/90

Taper pin: 32/90

Dual-diameter pin: 34/90

In total, the make percentage with the flagstick out was 90 percent. The average with a flagstick in was 45 percent, as high as 61 percent with the fiberglass pin and as low as 36 percent with the dual-diameter pin. In general then, a putt that would have struck the flagstick off-center is twice as likely to go in with the flagstick out as it is with the flagstick in.

(http://www.autoadmit.com/thread.php?thread_id=5520724&forum_id=2#47606095)

notice tour players don't do this shit

(http://www.autoadmit.com/thread.php?thread_id=5520724&forum_id=2#47604622)