John Connolly and Marty Walsh emerged as the two candidates remaining to replace incumbent Mayor Tom Menino. Walsh — despite never appearing as the favorite prior to yesterday — received 18.5 percent of the vote, topping would-be favorite John Connolly, who took home 17.2 percent.

That’s close. Very close.

Overall, 112,804 votes were cast during yesterday’s preliminary. That means the 1.3 percent difference between the Walsh and Connolly translates to a difference of 1,448 votes — 20,838 people voted for Walsh; 19,420 people voted for Connolly.

Assuming that the Nov. 5 election draws, at least, the same amount of voters as yesterday’s preliminary, Walsh and Connolly will be battling for a share of the 72,546 votes cast for other candidates. In order to make up his margin of defeat, Connolly will have to receive approximately 2 percent more votes than Walsh from Bostonians who supported other candidates.

Just how realistic is this?

During the 1993 preliminary election, Menino received 30,060 votes, or 26.89 percent, of the 111,781 votes cast. James Brett finished second with 25,052 votes (22.41 percent).

Despite four fewer preliminary candidates in 1993, only 1,092 more voters turned out in 2013 — which featured a more diverse field. Therefore, 1993 preliminary results have the potential to provide interesting insight into the battle now underway between Connolly and Walsh.

So, 20 years ago, Menino beat Brett by 4.8 percent in the the preliminary — 3.5 percent more than the difference separating Walsh and Connolly. When it came time for the 1993 general election, Brett needed to make up 5,008 votes — or 8.8 percent of the 56,669 votes cast for the six other candidates.

We all know what happened — Menino won.

During the 1993 general election, Menino received 64.45 percent (74,448) of the vote while Brett garnered only 35.54 percent (41,052); a nearly 29 percent difference. That means Menino received 44,388 more votes than he did in the preliminary. Brett only saw an increase of 16,000.

If that 3.2 percent increase from the 1993 preliminary to general election is applied to the 2013 results, 116,413 total votes would be the expected total turnout come Nov. 5.

Hypothetically, if 116,413 voters turn out on Nov.5, Connolly would have to receive about 1,513 more votes than Walsh, to make up his 1.3 percent deficit following yesterday’s preliminary.

Remember, however, Menino increased his margin of victory from 4.8 percent to about 29 percent in 1993 when 3,619 voters were added for the general election to win with 64.45 percent of the vote. Statistically, Menino received 2,332 of the 3,619 additional votes.

If the same math is applied to Walsh’s 1.3 percent margin of victory — Menino’s margin of victory was six times more in the general election than the primary (4.8 to 29) — he would defeat Connolly by about 7.8 percent, or 9,080 votes — roughly 62,746 to 53,666.

These figures are hypothetical, based on statistics from the 1993 general election and can’t be viewed as an actual prediction of a Walsh victory. If anything, these statistics show that Connolly, while only losing by 1,448 votes in yesterday’s primary, might need to make up more ground than just 1.3 percent come Nov. 5 — possibly as much as 6.5 percent.

This is a subtle reminder to cast your vote a little over a month from now. This is the first time in 20 years Boston is facing new leadership. There could be only 9,080 votes separating the candidates. It’s cliche, but every vote does count.

Click here to see why Connolly could emerge as the winner in November.