The Effect of Weather on Team Performance – MLB 2016

jackb4

May 1, 2020

My dataset that I will be working with is from kaggle at this link.

This dataset includes data from every single Major League Baseball game played during the 2016 season. It includes categories such as home team, away team, venue, attendance, and even weather.

For this project, I will be exploring the question:

How does weather affect team performance?

Team Records – 2016

Team	W	L	G	Win %
Arizona Diamondbacks	69	93	162	0.425926
Atlanta Braves	68	93	161	0.422360
Baltimore Orioles	89	74	163	0.546012
Boston Red Sox	93	72	165	0.563636
Chicago Cubs	114	64	178	0.640449
Chicago White Sox	78	84	162	0.481481
Cincinnati Reds	68	94	162	0.419753
Cleveland Indians	104	72	176	0.590909
Colorado Rockies	75	87	162	0.462963
Detroit Tigers	86	75	161	0.534161
Houston Astros	84	78	162	0.518519
Kansas City Royals	81	81	162	0.500000
Los Angeles Angels of Anaheim	74	88	162	0.456790
Los Angeles Dodgers	96	77	173	0.554913
Miami Marlins	79	82	161	0.490683
Milwaukee Brewers	73	89	162	0.450617
Minnesota Twins	59	103	162	0.364198
New York Mets	87	76	163	0.533742
New York Yankees	84	78	162	0.518519
Oakland Athletics	69	93	162	0.425926
Philadelphia Phillies	71	91	162	0.438272
Pittsburgh Pirates	78	83	161	0.484472
San Diego Padres	68	94	162	0.419753
San Francisco Giants	89	78	167	0.532934
Seattle Mariners	86	76	162	0.530864
St. Louis Cardinals	86	76	162	0.530864
Tampa Bay Rays	68	94	162	0.419753
Texas Rangers	95	70	165	0.575758
Toronto Blue Jays	94	77	171	0.549708
Washington Nationals	97	70	167	0.580838

This table shows each team with their respective amount of wins and losses. It seems a bit weird that most teams have 162 games, yet some teams have more, like the Chicago Cubs with 178. This is because the Cubs played in the playoffs while other teams didn’t quite make it so far. Since playoff games were still games in 2016, I decided to keep them in the dataset.

I also added win percentage, which is the great equalizer between teams that played only 162 times and teams that played much more. This will help out a lot more later when it comes to comparing different teams and different weather.

I haven’t touched on the weather yet, but being able to create a table like this is essential to the analysis, as one will be created for each weather type.

This dataset has columns for a few different aspects to weather, but I’m only going to choose one. Instead of choosing sky conditions or wind speed I will choose temperature. This is mainly because for sky conditions there isn’t enough data for rainy games, and a lot of games are listed as unknown.

For temperature, I am going to split the data up into four categories, mostly based off what I feel is a fair description of weather. (Baseball is played April through October, so while 60 isn’t really cold, it’s colder in comparison to summer)

Cold : 60 and below
Pleasant : 61-72 degrees
Warm : 73-81 degrees
Hot : 82 and above

While these ranges aren’t really the nicest looking, they provide a slightly more even spread.

This cell here shows the game distribution for each of the four weather types. As you can see they aren’t the most even, but the sample size should be large enough to pull meaningful data for each.

Wins By Weather

Cold

Team	W	L	G	Win %
Arizona Diamondbacks	1	1	2	0.500
Atlanta Braves	6	7	13	0.462
Baltimore Orioles	16	8	24	0.667
Boston Red Sox	16	13	29	0.552
Chicago Cubs	20	11	31	0.645
Chicago White Sox	9	9	18	0.500
Cincinnati Reds	9	12	21	0.429
Cleveland Indians	19	17	36	0.528
Colorado Rockies	13	10	23	0.565
Detroit Tigers	12	9	21	0.571
Houston Astros	4	5	9	0.444
Kansas City Royals	3	7	10	0.300
Los Angeles Angels of Anaheim	3	1	4	0.750
Los Angeles Dodgers	1	4	5	0.200
Miami Marlins	3	5	8	0.375
Milwaukee Brewers	2	6	8	0.250
Minnesota Twins	7	16	23	0.304
New York Mets	10	14	24	0.417
New York Yankees	10	14	24	0.417
Oakland Athletics	8	8	16	0.500
Philadelphia Phillies	11	9	20	0.550
Pittsburgh Pirates	9	8	17	0.529
San Diego Padres	6	12	18	0.333
San Francisco Giants	24	16	40	0.600
Seattle Mariners	11	11	22	0.500
St. Louis Cardinals	5	5	10	0.500
Tampa Bay Rays	2	4	6	0.333
Texas Rangers	2	4	6	0.333
Toronto Blue Jays	7	4	11	0.636
Washington Nationals	7	6	13	0.538

Pleasant

Team	W	L	G	Win %
Arizona Diamondbacks	13	18	31	0.419
Atlanta Braves	8	21	29	0.276
Baltimore Orioles	31	25	56	0.554
Boston Red Sox	31	24	55	0.564
Chicago Cubs	31	19	50	0.620
Chicago White Sox	25	18	43	0.581
Cincinnati Reds	12	24	36	0.333
Cleveland Indians	25	16	41	0.610
Colorado Rockies	15	27	42	0.357
Detroit Tigers	24	26	50	0.480
Houston Astros	13	16	29	0.448
Kansas City Royals	24	21	45	0.533
Los Angeles Angels of Anaheim	30	32	62	0.484
Los Angeles Dodgers	40	35	75	0.533
Miami Marlins	17	14	31	0.548
Milwaukee Brewers	26	27	53	0.491
Minnesota Twins	18	32	50	0.360
New York Mets	24	12	36	0.667
New York Yankees	20	22	42	0.476
Oakland Athletics	38	55	93	0.409
Philadelphia Phillies	17	16	33	0.515
Pittsburgh Pirates	19	19	38	0.500
San Diego Padres	28	35	63	0.444
San Francisco Giants	38	33	71	0.535
Seattle Mariners	46	28	74	0.622
St. Louis Cardinals	24	18	42	0.571
Tampa Bay Rays	52	61	113	0.460
Texas Rangers	20	22	42	0.476
Toronto Blue Jays	42	42	84	0.500
Washington Nationals	25	18	43	0.581

Warm

Team	W	L	G	Win %
Arizona Diamondbacks	32	53	85	0.376
Atlanta Braves	24	26	50	0.480
Baltimore Orioles	22	23	45	0.489
Boston Red Sox	29	22	51	0.569
Chicago Cubs	45	22	67	0.672
Chicago White Sox	25	34	59	0.424
Cincinnati Reds	29	37	66	0.439
Cleveland Indians	39	18	57	0.684
Colorado Rockies	14	22	36	0.389
Detroit Tigers	30	25	55	0.545
Houston Astros	52	47	99	0.525
Kansas City Royals	32	29	61	0.525
Los Angeles Angels of Anaheim	22	37	59	0.373
Los Angeles Dodgers	36	22	58	0.621
Miami Marlins	46	51	97	0.474
Milwaukee Brewers	29	39	68	0.426
Minnesota Twins	25	41	66	0.379
New York Mets	31	27	58	0.534
New York Yankees	35	18	53	0.660
Oakland Athletics	11	22	33	0.333
Philadelphia Phillies	24	31	55	0.436
Pittsburgh Pirates	31	35	66	0.470
San Diego Padres	25	23	48	0.521
San Francisco Giants	19	21	40	0.475
Seattle Mariners	20	24	44	0.455
St. Louis Cardinals	27	27	54	0.500
Tampa Bay Rays	10	19	29	0.345
Texas Rangers	30	19	49	0.612
Toronto Blue Jays	29	19	48	0.604
Washington Nationals	33	23	56	0.589

Hot

Team	W	L	G	Win %
Arizona Diamondbacks	23	21	44	0.523
Atlanta Braves	30	39	69	0.435
Baltimore Orioles	20	18	38	0.526
Boston Red Sox	17	13	30	0.567
Chicago Cubs	18	12	30	0.600
Chicago White Sox	19	23	42	0.452
Cincinnati Reds	18	21	39	0.462
Cleveland Indians	21	21	42	0.500
Colorado Rockies	33	28	61	0.541
Detroit Tigers	20	15	35	0.571
Houston Astros	15	10	25	0.600
Kansas City Royals	22	24	46	0.478
Los Angeles Angels of Anaheim	19	18	37	0.514
Los Angeles Dodgers	19	16	35	0.543
Miami Marlins	13	12	25	0.520
Milwaukee Brewers	16	17	33	0.485
Minnesota Twins	9	14	23	0.391
New York Mets	22	23	45	0.489
New York Yankees	19	24	43	0.442
Oakland Athletics	12	8	20	0.600
Philadelphia Phillies	19	35	54	0.352
Pittsburgh Pirates	19	21	40	0.475
San Diego Padres	9	24	33	0.273
San Francisco Giants	8	8	16	0.500
Seattle Mariners	9	13	22	0.409
St. Louis Cardinals	30	26	56	0.536
Tampa Bay Rays	4	10	14	0.286
Texas Rangers	43	25	68	0.632
Toronto Blue Jays	16	12	28	0.571
Washington Nationals	32	23	55	0.582

Now let’s combine those tables into bar graphs that show how each team performs in different weather.

A few conclusions we can draw are:

Atlanta Braves played much worse in pleasant weather than other conditions.
Texas Rangers played better as the temperature got higher
Kansas City Royals and Texas Rangers played worse in cold weather than other conditions, but not at all bad as the Los Angeles Dodgers did

However, these graphs aren’t the very best ways of comparing from team to team. For that, I’m going to compare ratios. Now I will do bar graphs, but for how well a team performed in one certain weather compared to their overall win percentage.

These graphs do a better job of comparing from team to team. Since I used ratios, these graphs aren’t skewed by teams having a higher or lower total winning percentage; it is all relative. Based off these graphs, we can conclude:

The Los Angeles Dodgers played terribly in the cold while their LA counterpart, the Los Angeles Angels of Anaheim played amazing. These ratios might be more extreme due to too small of a sample size, as the Dodgers only played five games in the cold while the Angels only played four. The Philadelphia Phillies might be a safer bet as the best team in the cold, as they played 20 games. Sample size is only an issue for cold games, as overall there were less cold games than other weather types
The Atlanta Braves did not play well at all in pleasant weather, while the New York Mets and Chicago White Sox played great.
The Oakland Athletics played the worst in warm weather, while the New York Yankees and San Diego Padres played the best.
This one is interesting. The San Diego Padres, who played the best in warm weather, played poorly in hot weather. Also playing poorly was the dome team Tampa Bay Rays, who may not have been so used to the sun. The Oakland Athletics, who didn’t play well in warm games, played the best in hot games. What a difference a few degrees could make.

We should also note that any team with a ratio greater than one played better in that weather than average, while ratios less than one played worse. From now on I’ll refer to these as “good ratios” and “bad ratios”.

For cold weather, the northeast coast teams from Philadelphia and Baltimore are towards the very top. Other teams from colder climates with good ratios are Toronto and San Francisco (San Francisco may not be generally thought of as a cold climate but they played 32 home games in cold conditions, more than any other team (second was Chicago Cubs with 21, which included many playoff games)). The worst team is the Los Angeles Dodgers, who play in warm and sunny Los Angeles. Other teams from warmer climates with bad ratios are Miami and Texas. Indoor teams with bad ratios are Tampa Bay and Milwaukee.

For pleasant weather, again, northeast coast teams New York Mets and Philadelphia are really high on the list, while teams from hot climates, Texas (Arlington) and Houston, have bad ratios. However, Miami has a hot climate and they have a good ratio.

For warm weather, warm climate teams like San Diego, Los Angeles Dodgers, and Atlanta have good ratios. However the Los Angeles Angels didn’t perform so well. The problem with warm and pleasant weather is that these conditions aren’t really extreme and during the spring and summer every city experiences many days with these conditions. Hot weather, however, isn’t quite the same way.

For hot weather, the hot climate teams like Arizona, Houston, and Texas all have good ratios. (I’ve been to a Texas Rangers game in Arlington, TX, and I can confirm it was indeed, very, very hot and sunny. I can’t say the same about the Chicago games I’ve been to.) The worst three teams are all from not-hot climates. San Diego is very far south, but it rarely gets up to temperatures that I listed as ‘hot’ (Stay classy, San Diego). Seattle is not hot, being in the rainy pacific northwest. Tampa Bay has hot weather, but their team plays in a temperature regulated dome where none of the games are listed as ‘hot’.

Overall, this was a fun little project on how different teams performed in different temperatures. A possible cause for why teams played better is based off of where they play their home games. For each condition, we notice that the best teams typically have home ballparks with a similar climate.