On Monday, President Trump announced that representatives from Washington and Tehran will meet this Saturday to discuss a potential deal over Iran’s nuclear capabilities and that Iran would be in “great danger” if a deal isn’t struck.

Yet Trump’s previous threats to bomb Iran have hindered the administration’s chances of getting the deal it wants. At first glance, the concept behind Trump’s threats toward Tehran seems simple enough: If a country doesn’t do a given action, the alternative will be much more painful. Unfortunately for the Trump administration, this logic is equivalent to the strategies of gaming aficionados in Risk or Civilization VI.

The subject of war and peace is a complicated one. Dealing with Iran is no exception.

An enemy’s perception of threats is important for deterrence to work. Several factors reduce the credibility of Trump’s threats to bomb Iran, and Tehran is likely aware of them. Ultimately, the administration is almost certain to fail to stop Iran from getting the bomb.

Trump seems to be relying on the concept of “madman theory” to coerce Iran from developing a nuclear weapon. This theory calls for the threat issuer to make himself seem irrational and unpredictable, with the goal being to instill fear into the minds of the opposing side, making it concede to the threatening party’s demands.

At first glance, the madman theory seems to follow the logic of coercion, with its use of threats to manipulate an enemy’s behavior. But on further inspection, using madman behavior is a poor bargaining approach in international politics. The madman’s outrageous threats lose credibility over time. If war is akin to a game of cards, one side will eventually call a bluff if the other player bluffs on every hand. Once this happens, the bluffer has only two options  —  play their hand or fold.

Furthermore, the target of the madman’s threats views the demands as insatiable. Why would the target country concede to the demands if he will just demand more and more?

In fact, modern history presents multiple examples of the madman theory failing to coerce the other side. President Richard Nixon’s nuclear threats failed to stop the North Vietnamese, and Saddam Hussein’s cultivated madman personas led to his downfall.

American military limitations further erode Trump’s credibility to stop Iran from building a bomb. Though the U.S. is likely to locate and destroy a good number of Iran’s nuclear capabilities, it is nearly impossible for the U.S. to destroy all of them. Iran has taken countermeasures to hide and disperse its nuclear facilities and capabilities.

Even if the U.S. destroyed all of Iran’s nuclear capabilities — an unlikely outcome — Tehran would still possess the knowledge to build nuclear weapons, as getting rid of nuclear weapons does not erase a country’s knowledge of how to build them. Taking out Iran’s nuclear capabilities might set back its nuclear program, but the underlying problem of Iran becoming a nuclear power would still remain.

The credibility odds also don’t favor the U.S. on the grand strategic level. Bombing Iran would reduce America’s resources for balancing against China and dealing with domestic issues. The logistical and financial costs of bombing Iran wouldn’t be cheap. It would take large weapons to take out Iran’s hardened nuclear capabilities. To do this, the U.S. would likely use stealth bombers, probably the B-2, and the GBU-57 Massive Ordnance Penetrator as bombs. Though the exact operational costs are unknown, the B-2 cost roughly $170,000 per hour to fly in 2021, and the MOP bombs cost a total of roughly $8.5 million for the Air Force to procure for FY 2025; only a small number of these have been delivered since 2015.

Should the U.S. drop these bunker-buster munitions on the Iranians, what would be the Trump administration’s next step? Should the military stop escalating if it fails to take out all of Iran’s nuclear capabilities? What would the administration do if Iran retaliates against the U.S. or Israel? Understanding the end goal of armed conflict is something that leaders and policymakers fail to do time and time again. The result is forever wars.

Ironically, Trump’s threats of preemptive strikes embolden Tehran to build the bomb. Iran’s strategic environment is not looking good. Israel has severely weakened Hezbollah and its ally in Syria fell. Iran’s weakened allies, coupled with Israel’s previous attacks on Iran, make it more likely for Iran to race to getting the bomb in an attempt to increase its bargaining power. A nuclear weapon is the great equalizer in international politics.

A return to a realistic foreign policy — one that requires compromise — is the best way for the Trump administration to salvage any nuclear deal with Tehran. This would mean accepting that Iran has nuclear energy capabilities, easing sanctions and stopping the threat of massive retaliation against the country. Over the long term, the administration should disentangle the U.S. from the Middle East, a region where America has no core strategic interests.

War is humanity’s most unpredictable undertaking. Those who say otherwise, or claim to know how each phase of a conflict will unfold, are either misinformed or purposefully misleading their constituents.

Trump should think long and hard about going through with these strikes. His administration and the Pentagon should keep the following quote from the late defense expert Fred Iklé in mind: “Those with power to start a war frequently come to discover that they lack the power to stop it.”

Benjamin D. Giltner is a Washington D.C.-based defense and foreign policy analyst.