Yes, the Cleveland Indians have the longest winning streak in major-league history

The Cleveland Indians won 22 straight games this year; their amazing streak was finally snapped by the Kansas City Royals Friday night.

The streak was one game longer than the Cubs’ 21-game winning streak in 1935, and as far as I’m concerned, the Tribe now owns the record for the longest winning streak in major-league history.

But wait, you’re saying. You’ve been reading a lot about a 26-game streak the New York Giants had in 1916, and that the Indians came up four victories short of that streak, so the Giants presumably still hold the record.

I’m here to make the argument that the Giants’ streak should not be recognized as a “winning streak” and that Cleveland now holds the record.

The issue here is a tie game that happened during that streak after the Giants had reeled off 12 consecutive wins. After that game — which, under the rules of the day, was replayed the following day — the Giants won 14 more games in a row.

That’s 26 wins, to be sure, but it happened over a 27-game span, not a 26-game span. And one of those 27 games was not a “win.”

We’re arguing semantics here, I think. Grant Brisbee put this all together in an article last week, and here, I think, is the key paragraph (written before the 22nd straight win):

The issue is this: The Cleveland Indians have won 21 consecutive games. If they win one more game, they will have set a new record for the longest uninterrupted winning streak in baseball history. That’s not my verbiage; it belongs to SABR, and it’s pretty mindful of this stuff.

There was a tie game within the Giants’ 27-game span, which, as Grant notes in his article, did “interrupt” their streak. The game, a 1-1 tie called after eight innings, was replayed as part of a doubleheader the next day. We wouldn’t even be talking about this if the Pirates hadn’t knotted up that game with two out in the eighth. Honus Wagner hit a sacrifice fly for the second out in that inning to score the tying run for the Bucs. Rain then stopped the game shortly after the ninth inning began. If that run doesn’t score, the Giants would have won a rain-shortened game, and we might be talking about a 27-game winning streak.

But they didn’t, and so we are left, 101 years later, with the following facts. Then, as now, all statistics from a game of that nature that goes longer than five innings (an “official” game) count in a player’s record, as well as in the team’s overall record.

The latter is where the controversy comes, I think. Every player not only has his at-bats, hits, runs, fielding plays, etc. count — but they also get a “game played.” So does the team.

Look at the 1916 Giants’ game log and statistical record. They are shown with 155 games played. The 1916 N.L. standings show them with 86 wins, 66 losses, or 152 decisions. The other three games ended in ties, either because of darkness in that pre-lights era, rain, or perhaps both.

Well, where did those other three games go? They can’t simply be ruled to not exist. They were played. The statistics count. The starting pitchers, Burleigh Grimes and the charmingly-nicknamed Pol Perritt, get credit for not only games started, but for pitching complete games. (Incidentally, Grimes, who went on to a Hall of Fame career, was pitching in just his third major-league game that day.) All the position players from that game get credit for a game in their 1916 game log.

In my view, you can’t have that and not have a team game as well.

Here’s a national baseball writer who agrees with my position:

Read the whole Twitter thread and replies, they’re pretty interesting and a lively discussion of this topic. I’m 100 percent in agreement with Jeff Passan (though he wasn’t quite right about the length of the 1916 game; as noted above it went only briefly into the ninth inning).

The way games like this are handled today is different from how they were handled 101 years ago. In 2017, a game like that 1916 1-1 Giants/Pirates game, tied in the eighth inning and unable to be completed, would be suspended and completed before the next game between the two teams. The only exception to this is if it’s the final game of the year between the two clubs and has no bearing on postseason position. You need go back only as far as last year’s Cubs to find an example of this. The last game of 2016 between the Cubs and Pirates, September 29, 2016, was halted in the sixth inning tied 1-1 due to rain, and it rained all night in Pittsburgh and the teams couldn’t finish. The Cubs had long before clinched the N.L.’s best record and the Pirates were far out of contention, so the game was declared a tie, the Cubs’ first since 1992.

Here’s the recap I wrote for that game last year, which has some further details on tie games in general and for Cubs tie games in particular.

Note the use of the word “games” in that paragraph. I thus ask you this question: How many games did the Cubs play last year?

The answer isn’t 161, the sum of their 103 wins and 58 losses. It’s 162. The Cubs played 162 games in 2016. It says so here and here and here and the last link is from MLB.com, which presumably is the official record of Major League Baseball.

Bottom line, in my opinion: The 1916 New York Giants had a 27-game unbeaten streak, in which they played that many games without a loss: 12 wins, a tie, then 14 more wins. All of those must count as “games,” because otherwise you have the logical fallacy of statistics counting for players without a “game” being played, which doesn’t make sense.

The 2017 Cleveland Indians had a 22-game stretch during which they won all the games they played, a 22-game winning streak. That’s the longest winning streak in major-league history. It should be recognized as such. And the Cubs’ 21-game streak in 1935 should be recognized as the longest winning streak in National League history.