Sunday, January 12, 2014

Would You Rather Face a Streaking Team or a Slumping Team?

Fans, broadcasters and writers often talk about momentum within a game or over the course of several games.  A Tigers fan might get concerned, for example, if his favorite team is playing a game against an Indians team which had just won five games in a row.  The thought is that the momentum of a five game winning streak could carry into the next game giving the Indians an advantage.  On the other hand, the Tigers fan might be more confident if the Indians had just lost five games in a row.

A few months ago, I looked at the effect of a walk-off win/loss on the following game.  One might expect a walk-off win to create momentum and make a team more likely to win the next game.  Conversely, a demoralizing walk-off loss might cause a team to perform poorly in the next game.  Using 19 years of Retrosheet.org data, I found no correlation between walk off wins/losses and wins/losses in games after walk offs.  Teams simply performed at the same level in the game after the walk off as they did in ordinary games. 

I recently did a similar study of teams on winning and losing streaks.  Using retrosheet data from 1995-2013, I found 87,860 sets of five consecutive games followed by a sixth game.  Table 1 below shows that there were 3,400 instances of a team losing five consecutive games, 14,158 times when they went 1-4, etc.

Table 1: Record in Game 6 after a Five-game Set
Five-game Record
N
Game 6
win%
0-5
3,400
45.5
1-4
14,158
46.3
2-3
26,197
49.6
3-2
26,578
51.3
4-1
14,315
52.2
5-0
3,212
53.9
 


We can see that the better a team performed in the five-game set, the more likely they were to win game six.  For example, teams that went 0-5 won 45.5% of the time in the sixth game.  On the other hand teams which went 5-0 won 53.9% of game sixes.  So, at first glance, it looks as if winning streaks create momentum that carries into the next game.

We need to be careful here though.  You can't directly compare teams that lose five games in a row with teams that win five games in a row.  Teams that lose five games in succession tend to be inferior to those that win five in succession and thus are more likely to lose game six regardless of momentum.  The teams losing five in a row won 46.2% of their other games during the year which was similar to the their game six winning percentage of 45.5%.  Moreover, teams winning five in a row won 53.2% of their other games which again was close to their 53.9% winning percentage in game sixes.  The full results are shown in Table 2 below.

Table 2: Record in Game 6 Versus Overall Record
Five-game Record
Overall Win%
Game 6
win%
0-5
46.2
45.5
1-4
47.8
46.3
2-3
49.3
49.6
3-2
50.8
51.3
4-1
52.1
52.2
5-0
53.2
53.9
 


So, it appears that five-game losing and winning streaks did not cause teams to perform any differently than usual in the sixth game.  Instead, their performance in game six was similar to their overall record throughout the season.  To be more sure, I also looked at teams with three game losing/winning streaks and seven game losing/winning streaks and the results were similar to those in Table 2.

So, the next time the Tigers face a hot or cold team, it is probably better to look at the overall record of their opponent rather than worry about their short-term streaks.   

The information used here was obtained free of charge from and is copyrighted by
Retrosheet.  Interested parties may contact Retrosheet at "www.retrosheet.org"

10 comments:

  1. Well if our team can win with a consistent enough regularity, then all opponents by default are slumping in comparison. Who cares who we play if we are the best. We just need to get into the pole position and let everybody else ponder who they want to play while we are busy just taking the race regardless of who is jostling for second place.

    ReplyDelete
  2. Thanks for the info! This was a great article. Now I can't wait for the Tigers to take the field!!

    ReplyDelete
  3. Berdj Joseph RassamJanuary 15, 2014

    Can't wait for the Tigers to play ball in 2014!

    ReplyDelete
  4. I think you last 0-5 should be 5-0.

    ReplyDelete
  5. You're right David. Thanks.

    ReplyDelete
  6. One other element; at some point a "slump" or a "streak" is not just random data from insufficient samples but relevant to the actual quality of the team. When the '84 Tigers started 35-5, to some extent it was random, but at the same time the odds against any team that is not quite good going 35-5 are astronomical. So they were a "streaking" team that you certainly would not care to face. Similarly, the 2003 Tigers were pretty much always in a "slump", simply because they sucked, and a reasonably good team was very likely to beat them.

    ReplyDelete
  7. The signing of Tanaka doesn't help the fact that I would rather play anybody than the Yankees or the Dodgers just cause I don't enjoy watching teams with bloated payrolls play baseball. It's just boring and perverted to me and I don't care for it since we don't match up on the spending or make up for it by making shrewd decisions. To me it's just awkward and embarrassing to watch one team blow mammoth levels of money haphazardly playing against another team that blows a double-sized mammoth amount. It's sickening and disgusting, not competitively graceful. It just sucks being any team other than the Yankees or the Dodgers and having to watch them spent truckload after truckload while we stand idle twiddling our thumbs and not coming up with our own way to build a competitive advantage. If your only strategy to win at baseball is to write huge checks and overpay everybody, then that's not fun or respectable. We need to better as a team. A LOT BETTER!

    ReplyDelete
  8. Replies
    1. Appears to be working!

      Delete
  9. Good stuff as usual. It reminds me of the example that would have been used in an elementary stats class - if you toss Heads 5 times in a row, it doesn't mean that the next one will be Heads, nor does it mean that the next 5 will be Tails. Each event is independent, and in your research a bad team has a 47% chance of winning a game regardless of the results that have preceded it.

    Unless Ortiz hits a grand slam.

    ReplyDelete

Twitter

Blog Archive

Subscribe

My Sabermetrics Book

My Sabermetrics Book
One of Baseball America's top ten books of 2010

Other Sabermetrics Books

Stat Counter