The latest poll out of New Hampshire shows Donald Trump maintaining a big lead, sitting at 29 percent a full 14 to 19 percent over Rubio, Cruz, Bush, Kasich, and Christie.

But those are just the headline numbers of the poll. If you really want to understand the significance of a poll, you can't merely consume the headline numbers. You have to read a poll, and parse it.

Parsing a poll means putting it in context with other polls: What is the trend? What is the average? How reliable is this poll?

Parsing a poll also means looking beyond the headline numbers.

In this case, PPP's New Hampshire poll doesn't show a strong Trump in NH as much as a vulnerable one.

Here's how I read this one:

Sample: 515 likely Republican Voters; Margin of Error +/-4.3 percent. That's a good poll for a primary.

The Basics: A broad, but firming second tier

Compared to PPP's previous poll Trump is up 2 points (basically even), Bush is up 5 points (to 10%), Rubio is up 4 points (to 15%), Kasich is up 3 points (to 10%), Christie is up 1 pt to 11%, and Cruz is down 3 pts (to 10%). Put another way, these top 6 are up an aggregate of 12 points--all taken away from the third-tier candidates. That means mostly that supporters of Carson, Fiorina, Huckabee, and Santorum are gravitating towards the second-tier candidates, nearly evenly.

Favorable/Unfavorable: Bad for Bush and Trump

Cruz, Rubio, and Christie all have favorability ratings in the 50s in this poll, and unfavorables in the 20s. Trump, meanwhile, has the highest unfavorable, at 44 percent, while Jeb is just behind him with 42 percent.

What does this tell us?: Two things, I think.

1) If you're an establishment donor or voter, Jeb is at best your No. 3 pick. If you're looking for someone to beat national-poll leader Trump and Iowa frontrunner Cruz, you want a guy with a high upside potential. Christie and Rubio, this poll suggests, could be those guys. Jeb couldn't.

2) Trump will do worse as the field thins out. The difference between Trump's support (28%) and his favorability (49%) is 21 pts--call it the "Friend Zone." This is the smallest Friend Zone of any candidates. Rubio, for instance, has 40 percent of NH Republicans who like him but support another candidate. For Cruz, it's 44 percent. This means if candidates drop out before or after Iowa, Trump is the least likely candidate to benefit.

Second Choice and Head-to-Head: More bad news for Trump

Donald Trump is the second choice of only 7 percent of NH voters, putting him behind Cruz, Rubio, Bush, and Christie. Again, this limits his upside.

Most telling may be the head-to-head matchups on which PPP polled. Trump tied Jeb Bush, lost to Ted Cruz, and lost badly to Marco Rubio (52 to 40).

Both of these results, together with his league-leading unfavorability, suggest that as the field thins, Trump will sit steady (or fall), while others will rise.

Also, Rubio's far better performance head-to-head against Trump is another reason to pick him as the Establishment horse in New Hampshire.

Bottom Line: Trump would win

This poll, read close up, shows Trump's weaknesses in New Hampshire--especially if some of those other double-digit candidates drop out (or just fade away) before the primary.

But still, a 15 point lead is big. If these numbers were to hold for a month, Trump would probably win, even if Bush and Kasich dropped out. If these numbers held, one of the establishment candidates would need to emerge as the sole--or nearly sole--establishment options in order to beat Cruz and Trump in New Hampshire.

Timothy P. Carney, The Washington Examiner's senior political columnist, can be contacted at tcarney@washingtonexaminer.com. His column appears Tuesday and Thursday nights on washingtonexaminer.com.