Last week, I began to ponder the question, “Are the Royals are streakier team than should be expected?”Â I began to break down the streak data, leaving out any streaks shorter than four games, because 3-game streaks (and less) are highly dependent on the opponent (for example, your chances of having a three game losing streak are much higher when the Red Sox and Redbirds are in the other dugout).Â Then, as I was tediously sorting through season data on baseball-reference.com, I ran out of time before I had to load up the family for a trip to the K.Â Because, as any good scientist knows, first-hand observation is always the best kind, and the Royals were on a 4-game winning streak at the time.
Today (yes, I usually put things off till the last day) I finished the toil of recording teams’ winning and losing streaks of 4 or more.Â I think I’ve found an effective way of displaying the results, using charts copied from Excel.
If you are reading this on a mobile device, well, I feel for you.Â My eyes are barreling toward bifocalsville, so I can’t wait to try to make out the teams’ initials across the bottom.Â But on a 15″ screen, well, this thing is a beauty.Â This first chart is win streaks of 4 or more.Â I arranged the teams from lowest to highest season win percentage so you can compare teams with their “peers”.Â Teams that lose a lot would not be expected to have many long streaks, and conversely, winners should have more win streaks.Â If everything went as expected, we’d seeÂ the columns form aÂ sloping line upward to the right, with the winners have more, longer streaks.Â That is the general trend you see here.Â However, the outliers are the Indians and the Dodgers, as they had more wins in long streaks than the 6 teams ahead of them.Â We Royals fans are all too familiar with Cleveland’s streakiness… it made it very difficult to climb out of 3rd place in the AL Central.Â The Dodgers had 3 streaks of 6 games, plus a 10-gamer.Â Atlanta had the longest win streak of the season, 14 games, but they also had singular streaks of 10, 8, 6, 5, & 4.Â The two World Series opponents showed that you don’t necessarily need to rely on super-long streaks: they had more 4- and 5-game streaks than any other teams.
Our Boys in Blue see KCR on the chart), who felt really streaky to me, performed about as expected.Â Their 28 wins in long streaks puts them just above the mean and median (50th percentile) of teams, and since their winning percentage was right in that neighborhood, it’s hard to say that they were streakier than they should have been.Â The 9-game streak probably stands out in my mind, mainly because those don’t happen in Kansas City very often.Â Detroit and Cleveland also had long streaks at the same time, making the streak both more crucial at the time, and more memorable in history.
Here is the complement to the win chart: the schnide chart.Â The losers, well, they really did losing streaks well.Â Again, with the teams ordered by win %, you expect the columns to decrease left to right.Â This chart forms a slope a little better than the wins did.Â The bottom four of five teams skew the mean higher than the median, so the stinkers had an easier time stringing losses together.Â The mean and median are central around the 25.5 figure we saw among league wins in streaks.Â Note the streaky Clevelanders again, poking well above their expected number of losses in streaks.Â They had 3 five-game schnides.Â If I had to recklessly explain this chart, I’d say that it shows the importance of a “stopper” or two in your pitching rotation.Â The good teams kept the schnides from becoming rank by having someone shut the opponent down, but I have not backed that up with stats.Â Your World Champion BoSox did not have even one losing streak longer than three games.Â That means even the back half of the rotation had to come through sometimes, because even aces don’t win all their games.
On the losing streaks, the Royals were a little streakier than expected.Â Having an overall winning record, you would expect their total losses in streaks to be below the mean and even the median.Â If they could have picked up one win during the eight-game losing streak in May, say dividing it into a 4- and 3-gamer, they would have been right around the median (the 3 losses would drop off this chart entirely).Â The Yankees, with about the same winÂ % as the Royals,Â had no 8-gamers, but had 6 4-gamers and 5 5-gamers.Â Cincinnati, just slightly higher in win %, had 3 5-gamers.
To measure true streakiness, you have to put the good together with the bad.Â This combined chart does not distinguish between win streaks and loss streaks, they all count the same.Â The expectation here would be a U-shape formedÂ by the column heights, as mediocre teams probably do not have long streaks (their W-L outcome probability is about equal) and the winners and losers at either end have more proclivity for streaks.Â Not exactly how this came out.Â Â Cleveland and the Dodgers really put a lump in that expectation.Â We can crown the ‘Stros the streakiest, with Francona’sÂ Tribe a close second.Â Â The Mariners?Â Well, they fit the profile of Seattle weather… just kinda blah, with only an 8-game losing streak to break the monotony.Â Â Again, the Royals were just a little above average in streakiness, with 11 teams (out of 30) ahead of them.Â I maintain that it seems streakier to this long-time KC fan because we are accustomed to losing in bunches, with very few upticks in outlook over the past 20 years.
That reminds me, it was 20 years ago when the Royals had a 14-game streak to put them right in the playoff race, 2nd-longest in their history.Â What became of that team?Â We’ll never know what might have happened… the strike ended the season prematurely.Â How Royal.
Thanks for bearing with me here.Â You may get different results if you set the streak threshold higher or lower than 4.Â Maybe I’ll milk this thing for another week.Â Maybe I’ll analyze how many alternating streaks there were of winning and losing.Â If you have any suggestions, please comment.Â -Dave
About The Author
Leave a Comment
You must be logged in to post a comment.