The sharpest fact from Saturday is not that my board went 4-2. It is that a market closed South Florida a 68% favorite at -215 and watched the Bulls lose 13-17. That is NHL line movement today's loudest lesson even though it came out of a football stadium: the close is an opinion, not a verdict, and the loudest opinions on the board are the ones that get bid past where they deserve to be.

Here is the whole day, nothing deleted, nothing cherry-picked.

The receipts from Saturday, October 3

PickTookClosedResultFinal
Tampa Bay Lightning ML-150-195Win3-1
Buffalo Sabres ML-213-225Win4-3
Edmonton Oilers ML-186-197Win3-1
South Florida Bulls ML-200-215Loss13-17
Dallas Stars ML-127-136Loss1-2
Colorado Avalanche ML-257-285Win6-1

Four wins, two losses, plus 0.06 units flat. Six picks carried a close and I beat the close on all six. A perfect CLV day that netted a rounding error is the most honest sentence I can write about this business.

Where I beat the close

Tampa was the day's best piece of work and it had nothing to do with the 3-1 final. I got the Lightning at -150 and the market dragged them to -195 by puck drop. Forty-five cents of movement on a home side is the market agreeing with me out loud, and it agreed too late to help anybody who waited. Edmonton did the same thing in miniature, -186 to -197, and the Oilers handled Seattle 3-1 the way the number said they would.

Colorado at -257 to -285 was the opposite kind of satisfying. You hold your nose paying better than two and a half to one, and then the Avalanche hang a 6-1 on St. Louis and you wonder why you were nervous. Buffalo was the sweaty one. 4-3 over Chicago, a one-goal game at -213, and if that puck goes the other way late this column opens very differently.

Where the close beat me

Dallas is the loss I will defend. Took the Stars at -127 in Nashville, watched the number climb to -136, lost 1-2. Right side, wrong night. One goal, a road building, no mechanism in the result that says my read was broken.

South Florida is the loss I will not defend as cleanly. I took -200 on the one non-hockey ticket on the board, the market pushed to -215, and Temple won outright 13-17. When a 68% favorite loses, somebody's model was mispriced and mine was holding the same side as the mistake. Bad read, not bad luck. The lesson is one I keep relearning: beating the close and being right are two different jobs, and only one of them pays on the night. Checked the almanac on the general principle and it has never once been repealed.

One line of color and then I will leave it alone: the biggest money anywhere Saturday was better than three million in Polymarket volume on Florida at Missouri, which tells you where the public attention lived while the hockey board quietly did the work.

What NHL line movement today teaches about tonight

Two things. First, the early hockey number is where the value lives right now. Five of my six tickets closed worse than I bought them, and the Tampa move was half a goal of market opinion appearing out of nowhere between my fill and the close. Early season, thin opinions, soft lines. That is a tell.

Second, heavy chalk is still chalk. I went 3-1 on moneylines priced -186 or worse and cleared six hundredths of a unit for the trouble. Small edges over and over, no heroics, and no pretending that 4-2 is a hot streak when the ledger says it is basically break-even. This is entertainment with variance attached, 21+ where legal, and the day you need it to pay the bill is the day to walk away. 1-800-GAMBLER is there for a reason.

What I am watching tonight: whether the early hockey prices keep drifting toward my side after I buy them, the same way Tampa and Colorado did. Six for six is a small sample. Twenty for twenty is a process.