I thought it'd be a worthwhile thing to take a snapshot once again of the state of the presidential polling, right before tonight's debate happens. This is the first Electoral Math column in roughly a month, so there's been some movement, although it is debatable whether much of it is all that meaningful. Many of the battleground states are balanced so closely that minor movements back and forth aren't really all that significant, at this point. Overall, the picture seems to have improved slightly for Kamala Harris, but it is definitely going to be a very close race, if the polling can be believed.

Over the past month, the Democratic National Convention was held and we're now beyond Labor Day, which is traditionally the kickoff to the end phase of the general election campaign. Over that time, Kamala Harris improved her standings in the national polls, but her momentum seems to have reached its limits (at least for now). The polling bump she got really happened before the convention even began, but for now the numbers have settled in to a very tight race. Websites that track national polls have Harris up anywhere from one percentage point to a little over three points at the moment, which has been holding fairly steady for the past few weeks. But of course national polling is largely meaningless, since we use the Electoral College to actually elect presidents.

So let's dive in to the state-level numbers, to see a better picture of the state of the race. As always, we use the fantastic Electoral-Vote.com site for all our data. We urge you to check it out to see individual graphs of every state's polling over time, to get a clearer picture of how tight this race truly is.

We begin with our breakdown of Electoral Votes (EV). Red represents Trump's share (measured from the top), Blue represents Harris's share (measured from the bottom), and whichever crosses the midpoint has the advantage. To put it another way, if all the polls are correct and blue is above the green line, Harris will win. If red is below the green line, then Trump will win.

[Click on any of theses graphs to see larger-scale versions.]

As I said, it's been a month since we checked in, so there has been some movement since then (the previous Electoral Math column happened on August 14th). Since then, things have moved back and forth marginally, with Harris retaining her edge but failing to gain much of any ground.

Trump started with 235 Electoral Votes, moved up to 260 EV at one point (only 10 shy of the number necessary to win), but then fell back down to 230 EV. Multiple polls were posted today, which raised him back up to 246 EV.

Harris started with 287 EV -- 17 more than necessary to win -- and then improved briefly to hit 297 EV. She then dipped below the winning margin, down to 262 EV, improved once again back up to 292 EV, fell below the line again to 257 EV, and finished up with today's movement back up to 292 EV.

The white areas of the graph -- states where the polling is tied -- fluctuated from zero (no tied states at all) up to 51 EV (with Georgia, North Carolina, and Pennsylvania all tied). It dropped from this high back down to zero again today.

Most of this movement happened due to the razor-thin margins from differing polls in five of the battleground states: Arizona, Georgia, Nevada, North Carolina, and Pennsylvania. Two others, Michigan and Wisconsin, also had multiple polls with very thin margins, but Harris seems to be holding the edge in both (at least for now).

Due to today not having any states perfectly tied, both candidates slightly improved their percentages of the total number of Electoral Votes. Trump improve his standing by two percentage points to wind up with 46 percent, while Harris improved by a single point to end up with 54 percent.

Let's take a closer look at each candidates' relative strengths and weaknesses. As always, in these next two charts, polls which show a 10 point lead or better are considered "Strong," while a gap of between 5 and 10 points is "Weak" and a lead of under 5 points is only "Barely" in their column.

We begin with Donald Trump's chart:

Trump's graph shows a lot of movement between the categories, but overall not much changed in his final total. This was due to two large states essentially balancing on the cusp between Weak and Barely -- Florida (with 30 EV) and Texas (40 EV).

Florida was the most active of the two, flipping from Weak Trump to Barely Trump and then back again multiple times. Since August 14th (our previous column), Florida moved to Barely, back to Weak, back to Barely, back to Weak, and today changed back again to just Barely Trump. That 30 EV bounce is mostly what created the rollercoaster-ride nature of the Weak numbers in Trump's graph. At the very end of the period, Texas also went from Weak to Barely Trump for three days, but then returned to Weak with today's polls. And Trump also got some bad news from Ohio, which moved from Strong Trump down to Weak Trump with today's polls.

The rest of the movement in Trump's graph came from very small shifts in the battleground states. Arizona moved from Barely Harris to Barely Trump, which was his only positive news in these marginal shifts. Three states flipped but then flopped back to where they started: Nevada shifted from Barely Harris to Barely Trump and then back again. Georgia shifted from Barely Trump upwards to Weak Trump but then flopped to Barely Harris and then Tied, only to wind up right back at Barely Trump. And Pennsylvania won the prize for flippiest state of all, shifting from Barely Harris to Barely Trump, then to Tied, Barely Trump, Barely Harris, Tied, Barely Harris, Tied (whew!)... before winding up right back where it began, at Barely Harris. Pennsylvania might be the key to the entire Electoral College, so it's pretty easy to see why so many polls have been released here (and how close the race truly seems to be). Trump lost ground in North Carolina, but it's been almost as flippy as Pennsylvania (Tied to Barely Harris to Tied, Barely Trump, Tied, Barely Trump, Tied, then winding up for now at Barely Harris).

Throughout all of this, Trump has remained below the 270 EV needed to win. However, he remains in much better shape than he was doing in the 2020 race, just to put it in perspective.

Trump's Strong category has been remarkably stable for the entire race so far. He began the period with 117 EV, which only fell to 100 EV today, with the loss of Ohio (down to Weak Trump). Trump's Weak numbers have bounced around a lot more, beginning at 91 EV, dropping briefly down to 61 (with the loss of Florida to Barely), but then recovering back up to his high for the period of 107 EV. From there, Trump hit 77 EV, 61 EV, and 91 EV in quick succession. Four days ago Trump lost Texas (down to Barely) and dropped all the way to 51 EV, but today recovered a bit by flipping Texas back to Weak while losing Florida down to Barely, and adding in Georgia again for a total of 78 EV. All of that adds up to a loss of 17 EV in Strong and a loss of 13 EV in Weak for the period. Trump's Barely numbers bounced all over the place, from a low of 17 EV up to a high of 76 EV, before ending up at 68 EV.

What all of this bouncing around added up to was Trump moving from a total of 235 EV up to 260 EV, down to 230 EV, before winding up at 246 EV -- a difference of only 11 EV for the whole period.

Of course, I've always thought the best measurement of a candidate's strength is to look at "Strong plus Weak," and in this Trump has lost some major ground -- but probably just for now. Trump's total in the Strong plus Weak measurement went from 208 EV briefly down to 178 EV before recovering and improving all the way up to 224 EV -- which is the highest he's been since his initial lead (built up against Joe Biden) collapsed, after Kamala Harris took over. But then Trump's numbers slumped back to 178 EV, recovered to 208 EV, slumped even further to 168 EV before finally settling on 178 EV again today. But this isn't really as significant as it might seem, since virtually all that movement was due to two incredibly large states (Florida and Texas) hovering on the cusp between Weak and Barely. Trump regained Texas today only to lose Florida, but if Florida strengthens back up for him he'll be back at 208 EV again -- which is exactly where he started.

Now let's take a look at Kamala Harris's chart to see how she's been doing:

As you can see, Harris has overall been pretty stable while the swing states do their swingy thing. She did get three pieces of good news during the period, however, although one of them was not that significant (merely filling in a large gap in the polling). Colorado hadn't been polled since the spring, so it had been in the Weak category all along, but just today a new poll was released showing Harris way up in the state, so it has now moved into the Strong column for her. But the better news for Harris was seeing Maine move from Weak to Strong early in the period, and then today in addition to the new poll from Colorado there was a new poll from Virginia which moved the state from Barely all the way to Strong for Harris.

This had the effect of moving Harris's Strong numbers from 162 EV up to 166 EV in the middle of the period, and then a big jump upwards to 189 EV today. These gains are probably permanent, although Virginia could balance on the cusp between Strong and Weak in future polling (it's not all that blue a state, after all). Colorado and Maine, however, seem fairly safe. During the same period, Harris's Weak numbers fell, from 51 EV to 37 EV, but this was good news since it represented Colorado and Maine getting stronger. Overall, in the important Strong plus Weak measurement, Harris improved by 13 EV, moving from 213 EV to 226 EV. This still leaves her 44 EV short of the final goal, but in better shape than Trump.

As with Trump, Harris's Barely numbers ping-ponged around as the battleground states moved to and fro between the two candidates. She started the period with 74 EV in her Barely column, and it fluctuated between a high of 84 EV and a low of 44 EV to wind up with a fairly healthy 66 EV.

Also similar to Trump, Harris's overall numbers moved around a bit with the battleground swings, but ended up almost where they had started. She began with a total of 287 EV, rose to 297 EV before falling back to 262 EV, rose again to 292 EV, fell again down to 257 EV, before finishing up today with 292 EV -- five more than when the period began, and 22 EV more than she needs to win the race.

My Picks

And now we get to the part of the program where my gut feelings influence things. After all, raw data can be misleading or just incredibly outdated, so this is where I attempt to correct for such things (case in point: Indiana is listed as "Barely Trump," which I find unbelievable -- the last poll was in April and it's almost certainly a lot stronger for Trump than that...).

Which leads me to one technical note -- down at the bottom of this column I have the full lists of every category here by state, complete with their Electoral Vote numbers. At the end of the data section is a list of all the states whose polls are so old they polled for Biden, not Harris. This time around, we had a large number (13) of states which added recent polls, so check that list out if you'd like to see what's been recently updated and which states are still woefully out of date.

In this segment, I divide all my picks into: "Safe," "Probable," and "Leaning" for each candidate, with a final: "Too Close To Call" category at the end. So let's take a look at where things stand.

Likely States -- Harris

Safe Harris (18 states, 217 EV)
Due to new polling in both states, I am optimistically moving both Maine and Virginia up from "Probable Harris" this time around. I had already included Colorado here last time (since I knew the poll showing weaker support was wildly out of date). Perhaps I am being too optimistic about both Maine and Virginia, but for now I'm pretty confident they'll both go for Harris in November.

Probable Harris (2 states, 9 EV)
This category shrank by the same two states (Maine, Virginia) moving upwards. Otherwise, no changes. I considered moving both New Hampshire and New Mexico upwards, but I would really like to see some more polling from both before making that move. It was leaked that Team Trump is essentially pulling all their efforts out of New Hampshire, which is a good sign they're giving up on it, but we'll have to see....

Likely States -- Trump

Safe Trump (20 states, 129 EV)
The only change here was moving Iowa up from Probable Trump. I'm not sure why I initially pegged it as only Probable Trump and although there hasn't been any polling since June (when Biden was still on the ticket), I think the days of Iowa being a swing state are long gone, at this point.

Probable Trump (3 states, 60 EV)
As noted, Iowa moved up to Safe Trump. Of the three states which remain, Ohio seems the best fit for the category as even though they've got favorite-son JD Vance on the ticket with Trump, the polling has been surprisingly weak here for them. Alaska will likely move up to Safe Trump eventually, but for some reason I'm not ready to do that yet. And Texas is probably pretty safe territory for Trump as well, but the polling hasn't been as strong as it really should be, so I'm leaving it here for now.

Tossup States

Lean Harris (2 states, 25 EV)
In both Michigan and Wisconsin, Harris has held the edge throughout multiple polls, so I left both states in the Lean Harris category. She's doing so well in Wisconsin that I even considered moving it up to Probable Harris, but I'm going to wait until at least next time to do so (adding favorite-son Tim Walz to the ticket gives me more confidence in saying this, I should mention). The big move here was the loss of Pennsylvania down to Too Close To Call, as it's really anyone's guess at this point which way it will go.

Lean Trump (1 state, 30 EV)
No change here. Florida is still leaning Trump, but the polls there have been weaker than I would have expected for him (it is, after all, his new "home state"). Eventually I will probably move Florida up to Probable Trump, but it doesn't quite seem justified yet.

Too Close To Call (5 states, 68 EV)
Four states stayed the same here this time around (Arizona, Georgia, Nevada, and North Carolina), while Pennsylvania moved down from Lean Harris to join them. I did consider moving Arizona up to Lean Trump (he's had a very slight edge in the recent polling), but more numbers are needed before I can honestly do so. Georgia, Nevada, and North Carolina are all neck-and-neck in the polling, and I had to reluctantly add Pennsylvania to the list as well.

Final Tally

This is going to be a very close race, folks. I know -- this comes as no surprise, right? But it's even more true than it was in 2016 and 2020, both of which were nail-biters in their own way.

Kamala Harris didn't get much of a bump from her convention, since she had already rocketed upwards when Biden pulled out and she replaced him on the ticket. Of course, tonight's debate could be crucial in changing the flavor of the race (in either directly, really), which is why I wanted to put out this snapshot right before it happened.

The changes in category this time around were fairly minimal, which leaves the numbers pretty much exactly as they were last time. Harris has the same 226 EV in the same 20 states in Safe and Probable, while Donald Trump maintained his numbers as well, at 189 EV from 23 states.

The only real change this time is that neither candidate reaches the magic number of 270 EV when you add in their Lean states. Harris moves up to 251 EV, while Trump winds up at 219 EV. This obviously leaves Harris in a better position, needing only 19 EV (the exact size of Pennsylvania), while Trump needs a bigger 51 EV to cross the finish line.

Neither candidate is dominant right now. The race truly could go either way. It's certainly something to consider, when watching Kamala Harris debate Donald Trump tonight, that's for sure.

[Full Data:]
(State electoral votes are in parenthesis following each state's name -- and please note a lot of these have changed since 2020, due to the reallocation of House seats after the Census. Washington D.C. is counted as a state, for a total of 51.)

Kamala Harris Likely Easy Wins -- 20 States -- 226 Electoral Votes:

Safe States -- 18 States -- 217 Electoral Votes
California (54), Colorado (10), Connecticut (7), Delaware (3), Hawai'i (4), Illinois (19), Maine (4), Maryland (10), Massachusetts (11), Minnesota (10), New Jersey (14), New York (28), Oregon (8), Rhode Island (4), Vermont (3), Virginia (13), Washington D.C. (3), Washington (12)

Probable States -- 2 States -- 9 Electoral Votes
New Hampshire (4), New Mexico (5)

Donald Trump Likely Easy Wins -- 23 States -- 189 Electoral Votes:

Safe States -- 20 States -- 129 Electoral Votes
Alabama (9), Arkansas (6), Idaho (4), Iowa (6), Indiana (11), Kansas (6), Kentucky (8), Louisiana (8), Mississippi (6), Missouri (10), Montana (4), Nebraska (5), North Dakota (3), Oklahoma (7), South Carolina (9), South Dakota (3), Tennessee (11), Utah (6), West Virginia (4), Wyoming (3)

Probable States -- 3 States -- 60 Electoral Votes
Alaska (3), Ohio (17), Texas (40)

Tossup States -- 8 States -- 123 Electoral Votes:

Tossup States Leaning Harris -- 2 States -- 25 Electoral Votes

Michigan (15), Wisconsin (10)

Tossup States Leaning Trump -- 1 State -- 30 Electoral Votes
Florida (30)

Too Close To Call -- 5 States -- 68 Electoral Votes
Arizona (11), Georgia (16), Nevada (6), North Carolina (16), Pennsylvania (19)

Polling data gaps:

[Note: during the primaries, Zogby conducted a poll in every state on April 21, with the sole exception of Washington D.C., which has had no polling at all this cycle. But I thought it'd be more useful to keep track here of which states have not been polled since Joe Biden dropped out of the race.]

New polling data with Harris -- 13 States
(States polled with Harris for the first time since our previous column, with the dates of the new poll in parenthesis.)

California (9/5), Colorado (9/8), Illinois (8/28), Maryland (9/8), Missouri (8/16), Nebraska (8/27), New Mexico (8/22), Oklahoma (8/30), Tennessee (8/29), Utah (8/9), Vermont (8/19), Virginia (9/8), West Virginia (8/27)

No polling data with Harris -- 22 States
(States which have not been polled since July 21st, with the dates of their last poll in parenthesis.)

Alabama (4/21), Alaska (4/21), Arkansas (4/21), Connecticut (4/21), Delaware (4/21), Hawai'i (4/21), Idaho (4/21), Indiana (4/21), Iowa (6/14), Kansas (4/21), Kentucky (4/21), Louisiana (4/26), Massachusetts (7/18), Mississippi (4/21), New Jersey (4/21), North Dakota (4/21), Oregon (4/21), Rhode Island (6/14), South Carolina (4/21), South Dakota (5/13), Washington D.C. (--), Wyoming (4/21)

-- Chris Weigant

Follow Chris on Twitter: @ChrisWeigant