It was late on a Monday night on an overnight casino stay, and the table games pit was mostly empty. I was at a $10 blackjack table with three other players - a man and a woman in their late 20s or early 30s who seemed to be a couple, and a fellow with a shaved head and gray beard who appeared to be in his 50s or early 60s.

Mostly, it was an easy-going, jovial atmosphere. We all had our ups and downs, but one streak of six dealer busts in a row put everyone a little better than even.

We toasted our wins - I with my usual bottle of water - and ignored each other's strategy quirks. The only times anyone raised any strategy objections were when I split pairs of 8s against the dealer's 10-value.

The first time, I had a $10 bet on the table, then pushed out $10 more to split the pair. I busted one hand by drawing a 6 followed by a King, but drew a 5 and then a 6 on the other for 19. The dealer had a 4 down, then drew a 3 for 17. I lost the bust but won on the 19 to break even for the hand.

"You got lucky there," the middle-aged gent said.

I smiled and said, "I'll take it."

The enhanced chance to break even is one reason basic strategy calls for us to split 16 against a 10. When that 16 consists of a pair of 8s, the lose-least situation is splitting the pair. If I just stood on my 16, my average loss would be 53.7 cents per $1 wagered, of $5.37 for my $10 bet. If I hit, meaning I'd bust anytime I draw a 6 through 10, my average loss would be 53.5 cents per $1, making hitting a marginally better play than standing, just as hitting any other 16 against a 10 is a slightly better play than standing.

But by splitting the pair and starting each hand with 8 instead of 16, I reduced the average loss to a little over 23.7 cents per dollar on each of the new hands. Since my bet was now $20 instead of $10, that meant my overall average loss would be $4.75, a better deal than the $5.37 loss for standing or $5.35 for hitting.

A little later in the session, I split 8s again, this time against a Jack. This time I drew a 10 on one 8, pulled a 3 on the other for an 11-total. I doubled down on that one, and drew a 9. That gave me 20.

When the dealer turned up a Queen for a two-card 20, my 18 lost and my double down pushed. Though I had $30 on the table, I lost the equivalent of my original $10 wager.

"Man, you really like to live dangerously," the bearded fellow said. "I don't think my stack of chips is big enough for that."

I laughed and said nothing, but what I was thinking is that my stack of chips wasn't big enough not to make the loss-reducing plays. Not that I didn't understand his point. Sometimes splitting the pair means you're going to absorb a double loss.

But overall, the good, bad and so-so outcomes favor splitting 8s versus a 10. You can't always play to win, but when the best option is to try to lose less, that's the option to take.

Gambling author and columnist John Grochowski's weekly newspaper column began at the Chicago Sun-Times and is now syndicated nationally. He also regularly makes TV and radio appearances about gambling. His column appears weekly.