Kansas City Has Won Seven Straight. They Are a .435 Team.

The Royals and the Phillies have the identical record over their last ten games. Stretch the window to fifty and one is a contender and the other is 57–74. Hot is not a fact you look up; it is a choice about how far back you count.

The Sports Page · The Professor · how to spot a real hot streak

The question: who is the hottest team in baseball right now?

The answer depends entirely on how many games you count. Fix the window and it is Milwaukee; they have not been below .600 over any span you care to measure.

Here is the surprising part. You would think hot is something you look up. The typical club can be made to look .200 better or worse just by changing where you start counting. Over a full season that is 32 wins, conjured out of nothing.

What you get from it: two questions that take ten seconds and settle the argument.

Two teams, one record

Kansas City has won seven in a row. Over their last ten they are 8–2.

Philadelphia has won eight in a row. Over their last ten they are also 8–2.

Through ten games these are the same team. Now watch what happens as we pull the window back. At twenty the Phillies are 15–5 and the Royals are 11–9. At fifty the Phillies are .560 and the Royals are .460. On the season it is 72–58 against 57–74.

Nothing changed except how far back we looked. Ten games is a small enough number that almost anything can happen in iti.

The window is a dial, and anyone can turn it

So we measured every club eight ways on the same morning. All thirty are below.

Look at the left edge. Over five games, seven of the thirty clubs are playing .800 ball or better, and another seven are at .200 or worse. Half the league is either magnificent or dreadful.

Now look at the right edge. Over fifty games, the number of clubs playing .800 ball is zero.

None of that heat was real. It was five games. The typical club’s best window and worst window sit .200 apart; Kansas City’s sit .540 apart, the widest in baseball. Give a fan the choice of window and they can prove almost anything about almost anybody.

The second question: did they outscore anyone?

Wins can come cheap. Runs cannot. So once we have fixed the window, we check whether the runs agree with the record.

The Royals over twenty games have outscored their opponents by exactly one run. One. That is a team winning coin flips, and coin flips stop.

The Phillies over the same twenty are +42. The Padres over thirty are +42. Those are teams beating people up.

A record tells you what happened; runs tell you whether it was earned.

The third question: who did they beat?

Here is the objection that should have occurred to us sooner. A +42 might be six games against the Rockies.

A fix already exists. Take a club’s run differential, then add the average rating of every opponent it faced; solve until the ratings agree with each other. Sports-Reference calls it the Simple Rating System. Bill James’s Pythagorean formula, for all its fame, does not do this.

Over a full season it barely matters. Everybody’s schedule evens out, and the hardest and easiest slates in baseball differ by about a tenth of a run a game.

Over twenty games it matters a great deal. There the range runs from −0.57 to +0.48 runs a game — a 21-run swing across the window, half of Philadelphia’s whole differential.

So we adjusted, and the honest result is that it changed very little this time. Detroit had the softest slate of any contender. Philadelphia’s +42 came against dead-average opposition and survived intact.

And none of it tells you who wins in October

We took every postseason series since 1998 — 223 of them — and asked which measure picks the winner.

The standings pick it 50.9% of the time. That is a coin flip. A club’s won-lost record, the number everyone quotes, tells you nothing at all about a short series.

Run differential does better at 57.8%. The schedule-adjusted last twenty gets 57.4% — twenty games carrying what a hundred and sixty-two carry.

Now hold that against the standard. We treat 75% as the point where a difference becomes noticeable. The best measure in baseball reaches 58%, under a third of the way from guessing to that line.

So who does nobody want to face?

Milwaukee, if you mean the team with no weak spot. Last 5, 10, 15, 20, 25, 30, 40, 50 — they are at .600 or better in every single one. There is no soft patch to wait out. Over their last thirty they have outscored people by 27, and their record matches that almost perfectly. No luck in it at all.

San Diego, if you mean playing the best baseball today. They are 7–3 over ten, 14–6 over twenty and 21–9 over thirty. That is .700 at all three, which almost never happens, and a +42 run differential backs it.

And if you want the October answer, it is the Yankees, who have allowed 3.13 runs a game over their last thirty. Nobody wants that in a short series no matter what the standings say.

One to shout at

Detroit has outscored its opponents by 46 runs over the last thirty games and gone 15–15.

The runs say that should have been about 19–11. They are winning big and losing close, which is the most maddening way to be mediocre. If you want a grievance to nurse this September, that is a good one.

What to take home: two questions, ten seconds

When somebody tells you a team is hot, ask three things: how many games, what the run differential says, and who they beat. That is the whole method, and it takes about a minute.

The first stops anyone choosing the window after they know the answer. The second asks whether the wins were earned or borrowed. The third asks who they were borrowed from.

You will use this outside sport more than inside it. Think of a fund quoting its best three years, a school quoting its best cohort, a politician quoting the quarter that suits. Each has picked a window and hopes you will not ask why that one. Ask why that one.

And be fair to Kansas City. Seven straight is seven straight, and it is a lovely thing to watch. It just is not evidence of anything yet.