The Knicks lost by 50, then won two nights later. Ignore the preseason scoreboard

On Tuesday night in Toronto, the Raptors beat the New York Knicks 148-98. They led by 53 late in the game. Two nights later, at Madison Square Garden, the same Knicks beat the same Raptors 122-120.

If you took the first score seriously, you spent Wednesday worrying about New York. If you took the second one seriously, you spent Friday feeling better. I think both reactions were wrong, and the reason is worth getting straight before the season starts, because the same thing will happen again next October.

What actually changed

The margin moved 52 points in 48 hours. Nothing about either team's talent changed in that time. What changed was the list of players.

In Toronto, Jalen Brunson, Karl-Anthony Towns, OG Anunoby, Mikal Bridges and Josh Hart did not appear in the box score. New York's leading scorer was James Wiseman, with 18 points. The Knicks made 7 of 41 threes. Toronto played its regulars, and Kawhi Leonard scored 21 in 17 minutes.

At the Garden, Brunson, Towns, Anunoby and Bridges all started. Brunson scored 20 with seven assists. Towns had 13 points, 15 rebounds and nine assists. Leonard played 23 minutes and scored six. The game was decided by Tyler Kolek, who hit a 28-foot three with 53.2 seconds left to break a 117-117 tie.

So the first game was a Raptors team close to full strength against a Knicks team without its five best-known players. The second was much closer to a fair fight. Neither one tells you who will be better in April.

The scoreboard is the least useful number

A preseason game is a practice with a crowd. Coaches decide in advance who plays and for how long, and they keep to it whatever the score. A team can be down 30 and still sit its stars, because winning is not the point.

That makes the final score close to useless. It measures which coach chose to play his regulars longer, more than it measures anything else. The Knicks are 1-4 in the preseason. I would not change my view of them by a single win because of it.

It happened elsewhere in the same week. On Tuesday, Oklahoma City came back from 20 down to beat Indiana in overtime. On Thursday, the Thunder lost by 27 in Houston, and they are 1-4 too. I do not believe Oklahoma City is a 1-4 team, and neither should you.

What is worth watching

None of this means you should switch preseason off. You just have to watch something other than the score.

Who closes. When a game is close late, the coach's choice of who stays on the floor says something. Kolek was on the court in the last minute at the Garden, and he took and made the biggest shot of the night. That is a small sign about how much his coach trusts him, and it will be worth checking whether it continues when the games count.

Minutes for young players. On Tuesday, Collin Murray-Boyles scored 19 for Toronto on 7-of-7 shooting from the field. On Thursday he played 22 minutes off the bench and scored 12. A player who keeps getting real minutes in October is a player his coach wants to learn about.

Who does not play. When a team's starters sit, that is the story, not the score. Ask whether it is planned rest or an injury, and wait for the team to say which.

The objection

Someone will say a 50-point loss has to mean something, even in October. It meant something about the Knicks' depth that night, because the players who did appear were mostly reserves. That is worth knowing if you care about the end of the bench. It is not worth knowing if you want to know whether New York can win the East.

The fair test is simple. When both teams play their regulars for similar minutes, a preseason result is a little more useful. When one team does not, throw the score away and keep the box score.

The rule I use

Before reacting to a preseason score, look at the minutes column first. If the stars played on one side and not the other, the result tells you about the coaches' plans, not about the teams. The Knicks and Raptors showed that twice in one week. Soon the games start to count, and then the scoreboard is worth reading again.

Update: the same lens on the Spurs, where lineup minutes say more than the score.