Playoff teams and their losing streaks

There's an adage passed on by old-school English soccer coaches that "Good teams don't lose three in a row." I have no clue if this is true by math, or not, but it belies a sentiment that is important; at a certain point, a talented team stops a skid. Be it the "luck" of the dice throw for every new game, an immeasurable clubhouse dynamic, a coaching staff talented at making adjustments and players who follow said adjustments, or something else, long losing streaks are more rare for talented teams. From my experience playing, even past teams I was on with comparable talent, some teams just seemed to have a certain little piece that kept them from long droughts of failure. Baseball doesn't happen in a vacuum, and neither does loss sequencing. Yet, I'm not quite sure how to explain it.

The table and charts presented here represent a first glance. A shot across the bow regarding losing streaks and their weight on a playoff season. By no means is the following conclusive and finished. What it is however, is a look at all teams, playoff and non-playoff teams in MLB from 2012 until the past season. Since the creation of the second wildcard team, I wanted to know just how many losing streaks (arbitrarily set at four games and longer) have occurred, by playoff teams and by non playoff teams. Here is some mostly raw data:

You'll notice a few things here. Let's start with the far left column. This is the amount of games in the losing streak, 4 representing a four-game losing streak and so on. The rest of the columns are rather self explanatory, the second column represents playoff teams since 2012 and the number of losing streaks. Since 2012, there have been 69 four-game losing streaks by playoff teams. The way the counting was done for this data was that a five game streak would not also count as a four game streak. All losing streaks are distinct periods of losing. The next column represents the same thing but for the MLB teams that would not make the playoffs. The final two columns show the total number of losing streaks and finally, the percentage of that game streak length performed by a playoff team. Ultimately, there have been 620 distinct losing streaks of four games or more in the MLB since 2012.

An interesting thing to consider, before we dive a little deeper here, is that of the 620 losing streaks of four games or longer in the MLB since 2012, only 115 of them belong to playoff teams. That's 19% of the streaks belonging to the playoff squads. One could imagine, with 1/3 of teams making the postseason every year, that this number would stray closer to 33%, if all was even. This is one of the many "why's" I want to understand. Let's move forward with a chart.

Here we have a pretty little thing. The first thing to consider is that, in case you missed it above, no playoff team in the second wildcard era has had a losing streak longer than nine games. It happened once, to the AL West Champion A's in 2012. What is striking here is the R-squared of this line. It goes without saying that good teams win more games, however, why do good teams not sequence losses more often? The difference here is only 5 games, yet the trend shows a distinct pattern that playoff teams are exceedingly less-likely to lose 7 games in a row than 4. Adding the non-playoff teams to this chart looks something like this:

Here we have the same chart as above but with the added data of the losing streaks performed by non-playoff squads. You'll noticed that the trendline actually fits this data even better, and the slope of the line is much steeper than the "playoff" trendline. It's difficult for me to find a cause for the discrepancy in steepness, but maybe it has to do with the ability for playoff teams to limit four-game losing streaks. It may also simply be that there are more data points in the "non-playoff" set, allowing the line to fit better with increased sampling. Keep in mind, in all of MLB since 2012, only 10 losing streaks of nine games have occurred. The sample-size starts to dwindle.

This is, as mentioned above, by no means conclusive. My experience playing this sport for a long time tells me that there is something here, that avoiding losing streaks is a quality of a good team beyond mathematical odds, yet I could be entirely wrong. Here I am simply presenting the beginnings of what will likely be a much deeper dive into the subject of losing streaks and playoff odds. Ultimately, a model showing a losing streak length to playoff odds correlation would be ideal. If we could get predictive with this data, say, to be able to say that once a team has a seven-game losing streak, they have an 8% chance of making the playoffs, that would be something. I have no clue if we can get there, however,

There are a lot of questions here, and, hell, maybe this data is completely meaningless. I want to believe it isn't.

The good news? There were eighteen four-game losing streaks by playoff teams last year. Every single team had at least one. Kansas City, the Dodgers, and Pittsburgh all had three. Only Kansas City and St. Louis avoided five game losing streaks and made the playoffs last year. That means at different points of the 2015 season, the Pirates and the Dodgers had three distinct four-game losing streaks and one five game losing streak! The 2015 Mets had two losing streaks of six games and made the playoffs. So did the 2014 Giants.

It's early.

goms