The New York Yankees are 24-10, which is a fairly surprising to any time travelers who came from a few weeks ago and saw that the team went 9-9 to start the year. They have since won 15 of their last 16 games, two of the last three in dramatic fashion. The Bombers sit a game out of first place as I type this.
There’s a concept in statistics called signature significance, and it has already been applied to some projection systems in research. One series of events can demonstrably move the true talent needle if beyond the norm by enough of a margin. Even if it didn’t, there is still a more rote sabermetric argument as to why this is a great thing.
If the team were, say, a true talent .550 win percentage team and they still won 24 games instead of the median 18, those “extra” six wins get banked. That would mean an additional six wins on top of the original projection. Even if the Yankees regress, and they are likely to, this stretch has been so good that you can’t merely erase it.
There’s another metric that stands out. According to FiveThirtyEight’s ELO rating, which uses a numerical rating based on how well teams do against other teams’ talent level, the Yankees are at 1574. That’s the best mark in baseball. The last time they held the top spot was for a brief period on June 11th of last year, and then in May of 2013.
ELO, while flawed in many ways, is an interesting way of showing teams vis-a-vis each other. I like to use it to show how teams have nudged their talent level in one direction or another. ELO also has historical records, so it was actually a good exercise to look at the following: which teams reached a mark of at least 1574 before May 6th, and what was their season result? Here is each team’s pre-May 6th peak:
- 1586 on May 6, 2010: Lost ALCS
- 1579 on April 23, 2003: Lost World Series
- 1586 on April 14, 1999: Won World Series
- 1582 on May 6, 1951: Won World Series
- 1578 on May 6, 1943: Won World Series
- 1586 on April 19, 1940: Third place, 88 wins
- 1578 on May 6, 1939: Won World Series
- 1579 on May 5, 1937: Won World Series
- 1589 on April 27, 1933: Second place, 91 wins
- 1579 on April 30, 1932: Won World Series
- 1588 on April 19, 1928: Won World Series
This is what I call a fun fact. This accounts for seven of the Yankees’ 27 World Championships, and the caveats are obvious. Teams change over the course of a season, and even teams like 1927 and 1998 slowly crept their way to greatness. The other caveat is that this includes teams pre-wild card era, so if you were that good early on, you likely created a league lead so comfortable it nearly guaranteed a pennant by July.
Yet in every fun fact, there’s always an element of truth. The truth of that is that if a team is this good in the beginning of the season, it sets a floor so high that unless there is a full-scale collapse — which can always happen in this sport — you’re almost guaranteed to win 90 wins and a playoff spot.
That’s no satisfaction to a team like the Indians, who showed the folly of trusting in a streak or what could be considered signature significance. Even though they won 22 straight games, losing three in a row to the Yankees ended their season.
It should be encouraging, though, that even in a stretch as good as this one, the team will likely play at a high, competitive level from now until October. I’m not going to say it means there’s a 63% chance of a World Series win, but playing this well is exciting enough for now.