Pretty near every leader on earth, in business, government, or any other realm, has something riding on who will become the world’s most important leader, the U.S. president. Now that the debates are over, let’s review the ways we might foretell the outcome. They reveal not only what’s likely to happen, but also something about how we got here.

Nationwide preference polls: These get all the attention, but of course they mean little. We elect presidents by electoral votes, not by a majority of votes cast. So while the Real Clear Politics average of polls shows Hillary Clinton leading Donald Trump by a wide margin, 45.2% to 39.2%, let’s acknowledge it and then forget it. We have better tools available.

Electoral maps based on state polls: These give us a more precise understanding. The latest RCP map shows Clinton with 260 electoral votes vs. 170 for Trump – a big lead for Clinton, but neither candidate has the 270 needed to win. That’s because RCP says nine states are toss-ups, including three of this election’s battlegrounds – Florida, North Carolina, and Ohio. The fourth battleground state, Pennsylvania, is classified only as “leans Clinton,” not “likely Clinton.” When RCP assigns each of those states to Clinton or Trump, even by a tiny margin, then Clinton is a big winner, 333 to 205.

Prediction markets: People betting real money by trading contracts that pay off according to an election’s winner have a strong record of accurate forecasting. The Predictwise site maintained by Microsoft Research economist David Rothschild amalgamates activity in these markets. It’s favoring Clinton 91% to 9%, and the most recent debate did not noticeably affect the numbers. On the contrary, this indicator reveals that the most significant turning point in the whole campaign was the first debate, on September 26. Before that event, traders gave Clinton only a 68% chance, but her numbers jumped sharply as soon as the debate ended and have climbed steadily since, plateauing around 90% a week ago.

The peso’s price: Trump’s vocal hostility toward immigrants from Mexico and trade with Mexico has made the dollar-peso exchange rate a sensitive barometer of his chances. When he looks strong, traders flee the peso; when he looks weak, the peso appreciates. The peso’s value plunged after the party conventions last summer and was hitting new all-time lows until – guess when – September 26. It’s value jumped after that debate and has continued to climb.

Sign up for daily insights, updates, and opinion on leadership and leaders in the news at the Power Sheet.

So here’s the bottom line: Ignore the national polls. Based on the electoral map, Clinton indeed looks overwhelmingly likely to win; Trump would win only if he could take virtually all the toss-up states, most of which are trending against him. Probably the most trustworthy indicators are the odds derived from real people betting real money on the outcome, which show Clinton extremely likely to win. And a message to all pundits: Please stop parsing the transcripts of each debate; the bigger story is clear. Before the debates, Trump at least had a chance; the tide turned after the first debate, and it has never turned back.