At recent Passover Seders, Jews quoted an ancient rabbi who exclaimed, “Here I was a man of 70 years and did not understand why Passover should be celebrated at night until Ben Zoma [another sage] explained it.”

That is just how I felt at an event sponsored by California Democratic Rep. Jane Harman’s SecureUs and the Third Way, listening to Kenneth Pollack discuss alternative security strategies for Iraq. I have spent years criticizing George Bush on Iraq but did not understand the alternatives until Pollack explained them.

Criticizing the president for misleading us into war and for not having a plan to win the peace is right and easy. What’s wrong with Bush’s current policy though? Pollack’s answer: plenty. His critique is central to understanding why generals are demanding Defense Secretary Donald Rumsfeld’s resignation, arguing that he ignored years of Pentagon planning.

Pollack outlines two basic strategies for the current battle. Bush and Rumsfeld chose a traditional counterinsurgency. Tactically, that means offensive operations to kill or capture Iraqi insurgents. It has not worked. Everything we thought we learned about guerrilla war predicted that it would not work.

Such strategies usually do more harm than good. The insurgents are most prevalent in the Sunni triangle, where support for them is greatest. What happens when U.S. forces chase fighters into those areas? Like good guerrillas from Francis Marion to Mao, they melt into the local population, fighting mostly in circumstances of their choosing.

American forces, properly concerned about protecting our troops, break down doors, demand women and children lie on the ground while their homes are ransacked and shoot at any car that comes too close on the road. While such actions protect American soldiers, they do not exactly help win Iraqi hearts and minds.

Thus the very tactics necessary to prosecute the strategy do violence to the ultimate goal of generating opposition to the insurgency and support for the Iraqi government. Contradictions of this kind are a clear sign of poor strategy.

Moreover, the Bush-Rumsfeld strategy does not properly match our goals to our resources. Military planners have long known that success in a traditional counterinsurgency requires 20 security personnel for every 1,000 locals. The simple math, undone or unacknowledged by the administration, suggests it would require about 400,000-500,000 security personnel to mount that kind of war. When Gen. Eric Shinseki voiced this accumulated military wisdom, Rumsfeld and his deputy Paul Wolfowitz slapped Shinseki down.

With about 200,000 coalition troops, we lack the resources on the ground to prosecute the strategy laid out by the president and his secretary of defense. Moreover, because our troops are in the Sunni triangle they are not in other population centers, creating a security vacuum filled by the militias now engendering civil war.

Pollack does suggest an alternative. Its premise: Security for the Iraqi population is a prerequisite for reconstruction and for building an effective government that can engender popular support.

Therefore, instead of hunting down insurgents in the Sunni triangle, Pollack recommends we focus on providing security  in other parts of the country where support for reconstruction is greater. Secure the areas that most want to be secured, enabling residents to return to a semblance of normal life. That semblance of normalcy will deprive the insurgents of support, increasing stability. Over time, those islands of stability will spread out like an “oil stain.”

Pollack’s approach, too, has problems. But it is certainly better than the strategy Bush and Rumsfeld are employing — a strategy I for one can now critique with understanding.

Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982, including Sen. John Kerry (D-Mass.) in 2004.