How to Win at Betting Baseball

I have received many inquiries about a piece I wrote in March 2008 on Covers about betting baseball and I feel this piece is my most valuable contribution at Covers. Many have enjoyed reading it and I wanted to post it again before the baseball season begins. For those who have read it previously, I am sorry for the redundancy although I have added text and made numerous revisions. Best of luck to everyone this baseball season.

Written in March 2008 (revised in March 2009 and March 2010).

WHY BETTING UNDERDOGS IN BASEBALL IS DIFFERENT FROM ANY OTHER SPORT.
 
There are several different approaches and strategies to handicapping baseball, but the most fundamental and obvious starting point is one that is taken completely for granted. Baseball is quite unique when it comes to betting because the gaps between the best and worst teams are far closer than in any other sport. Let's look at the other major betting sports.
 
In the NFL this season, the Indianapolis Colts were 14-2 winning 87.5% of their games. The worst team in the NFL this past season was the St. Louis Rams at 1-15, winning just 6.25% of the time.

In the NBA this season, Cleveland is currently 47-14 winning 77% of their games and Lakers are 45-15 winning 75% of their games. The worst team in the NBA is the New Jersey Nets at 6-53 winning 10.2% of the time.   
 
In NCAA Basketball this season, Kansas, Syracuse and Kentucky are all 27-2 with a 93.1% winning percentage. The “Board” team with the worst record in the country is Marist at 1-28, winning 3.4% of the time. 
 
In NCAA Football, Alabama and Boise State both finished undefeated at 14-0 (100%) while Eastern Michigan and Western Kentucky both finished 0-12 (0%).
 
Meanwhile, the 2009 baseball season was not that much different than any other baseball season.
The New York Yankees had the best record in the majors last season at 103-59 (63.6%) and the Los Angeles Angels had the second best record at 97-65 (59.9%). The two worst teams in the majors were the Washington Nationals at 59-103 (36.4%) and the Pittsburgh Pirates at 62-99 (38.5%). These records are nowhere close to the best and worst records of any other major betting sport.
 

So there you have it:
 
Sport                Best             Worst
NFL                 87.5%            6.25%
NBA                77.0%          10.20%
NCAAB          93.1%            3.40%
NCAAF           100%             0.00%
MLB                63.6%          36.40%

While it seems like an obvious point, it is often overlooked that the worst team in baseball still wins at a percentage well in excess of at least triple the worst team in any sport. Taking it a step further, based on last season's numbers, the worst team in baseball won slightly less than 2 out of every 5 games, while the best team in baseball won slightly more than 3 out of every 5 games. Let's face it, in NCAAB, if you put Kansas up against Marist 100 times this season, Kansas would probably win 99% of the time if not every time. In the NFL, if the Colts played the Rams 10 times, they'd easily win 9. In the NBA, if the Cavs played the Nets 100 times, they’d probably win close to 90% of the time. In college football, if Alabama faced Eastern Michigan or Western Kentucky 100 times, they would win at least 95% of the time if not every time. It's nothing like that in baseball.
 
In 2009, the worst team in the majors, the Washington Nationals, walked into Yankee Stadium and beat the best team in the majors, the New York Yankees 2 out of 3 times in June and showed a huge profit for the series (more than +5 units). The Houston Astros finished 74-88 (45.7%) but beat the National League Champion Philadelphia Phillies 6 out of 8 times and the team in the NL with the best record, the Los Angeles Dodgers (95-67,  58.6%) 4 out of 7 times. The Pittsburgh Pirates at 38.5% beat the Florida Marlins at (87-75, 53.7%) four out of six times and had a better record against the East Division (18-16) than did the Division Champ St. Louis Cardinals who were 17-16. The Mets were 25 games worse than the Dodgers but had a better record against the Central Division at 23-17 while the Dodgers were 22-19. The Phillies had a 6-12 Interleague record while the Pirates were 8-7. Anyone can beat anyone in baseball!

The point can also be further highlighted by looking at starting pitchers. Let say we’re back in March 2009 and I gave you the choice of Groups A or B below and told you that you had to bet one unit on each pitcher in each start for the 2009 season in the group you choose.

GROUP A:

Tim Lincecum

Zach Greinke

Roy Halladay

Johan Santana

Clayton Kershaw

Chad Billingsley

Ryan Dempster

James Shields

GROUP B:

JD Martin

Braden Looper

Kevin Correia

David Huff

Matt Palmer

Clayton Richard

Barry Zito

Sean West

Most people before last season started would’ve chosen Group A for sure. In fact, if I asked many people today (on March 3, 2010) to choose Group A or B as the better investment last season (after the fact), I still think most people would choose Group A.

Let’s look at the results:

GROUP A

                                    Team Record               Profitability 1 unit/gm

Tim Lincecum               19-13                           -1.62 units

Zach Greinke                17-16                           -5.07 units

Roy Halladay                17-15                           -6.73 units

Johan Santana              13-12                           -1.71 units

Clayton Kershaw          14-16                           -10.34 units

Chad Billingsley            17-15                           -5.51 units

Ryan Dempster 13-18                           -9.05 units

James Shields               15-18                           -13.35 units

Total                         125-123 (50.4%)            -44.33 units

GROUP B

                                    Team Record               Profitability 1 unit/gm

JD Martin                     10-5                             +10.94 units

Braden Looper 20-14                           +8.06 units

Kevin Correia               19-14                           +9.75 units

David Huff                    14-9                             +7.16 units

Matt Palmer                 11-2                             +13.58 units

Clayton Richard            9-3                              +8.93 units

Barry Zito                     18-15                           +7.39 units

Sean West                    12-8                             +6.74 units

Total                          113-70 (61.7%)             +72.55 units

To put things into perspective, try and think about the potential biggest one game baseball mismatch for this upcoming season. How about the Nationals with their 5^th starter at the Phillies with Roy Halladay on the mound?  I make the line between -300 to -400 but for arguments sake, I will go with -400. How big of a mismatch is it really? We see some really big mismatches in other sports. In the first round of the NCAA tournament, you can expect some 30-40 point favorites with money lines of -10,000 in the #1 vs. #16 games. In CFB, Alabama would be close to a 40 point favorite over Western Kentucky which would easily be -5,000 to -10,000 on the moneyline if you could find such a line. The Colts would be approximately17 point favorites or more at home against the Rams which would give a moneyline around -2000. So is the biggest mismatch of the season in baseball really that much of a mismatch when you compare it to other sports? Last night in CBB (March 3, 2010), Tennessee State was a 9 point underdog in the first round of the Ohio Valley Tournament at Austin Peay. The money line was Austin Peay -450. Tennessee State won outright. That victory last night by Tennessee State was a greater upset (and really wasn’t that shocking) than any upset you will see in the major leagues this season.
 
To sum it up, too many baseball bettors shy away from dogs like they can't win when it is clear that any given underdog can beat any favorite on any given day at a rate unlike every other sport. Until people get comfortable in baseball playing underdogs on a regular basis and seeking out value, there is no chance to win betting baseball in the long run.  

WINNING PERCENTAGE IS NOT SO IMPORTANT IN BASEBALL HANDICAPPING.
 
Too many people get caught up in what a handicapper's winning percentage is in baseball and to me that just further explains to me that people do not understand betting baseball, as successful baseball handicapping is not a function of winning percentage. I would go so far as to say that if any average baseball handicapper had a gun to his head and was asked to hit 60% over the course of the season, he probably could. Simply, if a baseball bettor took every favorite of -200 or greater, he would probably hit 60%. The problem is even if you played all favorites with the average bet being a favorite of -200, at 60%, over the course of 100 plays at $100 each, you would lose $2000. At an average of -200, a handicapper would need to hit 67% to even show a negligible profit. Don't forget, the best team in baseball last season only hit at a 63.6% clip.
 
Taken a step further, a handicapper playing an average of -150 could hit 60%, but at 60% (a better % than the best MLB team last season) he would only breakeven.
 
Conversely, a handicapper playing an average of +110 would only need to hit at a rate of 47.7% to breakeven. So I ask you, who is the better baseball handicapper, the 60% capper playing an average of -150 or the 48% capper playing at an average of +110? The answer is the 48% capper as he shows a profit while the 60% capper above breakeven. The point I am driving home is winning percentage in baseball is less significant than other factors.
 
For those of you who followed me at Covers, you are aware I play more games than almost any other handicapper. Lord knows there were countless haters who would state that I played too many games and couldn't be successful. I respect they are entitled to their opinion, but based upon my baseball model, in 2007, I won +0.12 units per game played, so my overall profit increased by an average of 0.12 units with each game I played. That is not to say I should bet every game, but based upon my handicapping techniques, the more games that fit my personal criteria, the more profitable I was. 
 
I played 1,172 games in 2007 and finished with a record of 626-546 (53.4%) and made a profit of +140.86 units. Let's say this season I play 1000 games at an average of +110 (slightly higher than last year). Here is what I can expect if you assume that I flat bet every game at the same amount:
 
Win %                Unit Profit
47.7%                     Even
48%                      +8 units
49%                      +29 units
50%                      +50 units
51%                      +71 units
52%                      +92 units
53%                      +113 units
54%                      +134 units
 
I won't use anything above 54% because it just won't happen. In 2007, I hit over 53.4% and as such, had an outstanding season. Even if I don't reach that standard, at 50%, I will still make a handsome profit.

REDUCED JUICE BOOKS ARE IMPORTANT TO USE.
 
It is scary to think that some baseball bettors actually play through books that offer 20 cent lines. Simply put, that is suicide. In fact, I am guilty too, as in 2007, I played at 10 cent shops and did not use Matchbook or Pinnacle, which are two of the major reduced juice shops. For those of you not familiar with places like Matchbook, you can basically play with 1% juice. When a game is -120, take back +110 at dime shops, at Matchbook, it is probably -116/+115. Here is a comparison of what I would expect if I averaged +110 for the season at a reduced shop like Matchbook, compared to a 10 cent line which would cost me five cents per play overall (meaning I would average +105, thus the +105 column below), compared to a 20 cent line which would cost me 10 cents per play on average (meaning I would average +100, thus the +100 column below):
 
 

Win %         +110             +105        +100 
47.7             Even            -22.15u      -46u
48%            +8 units        -16u           -40u
49%            +29 units      +4.5u         -20u
50%            +50 units      +25u          EVEN
51%            +71 units      +45.5u        +20u
52%            +92 units      +66u           +40u
53%            +113 units    +86.5u        +60u
54%            +134 units    +107u         +80u
 
By using a reduced shop like Matchbook compared to 10 cent and 20 cent shops, if I hit 50%, I earn 50 units on the season in my example, but at the 10 cent store, I would earn 25 units and at the 20 cent store, I would breakeven. These are HUGE differences. I reviewed my numbers from 2007 and had I used a reduced juice book like Matchbook and averaged a savings of 5 cents per game (that is conservative as it is closer to 8% if you are a good shopper) then I would've earned an additional 39 units last season. YOU MUST USE A REDUCED JUICE BOOK.  

HANDICAPPING GAMES.
 
The key to handicapping baseball is searching for value in relation to the line. It’s very similar to a horse race. Let say I give you a racing form for a 10 horse race today at Belmont Park and the only thing I delete from the racing form is the odds of each horse. After evaluating all ten horses, you will have a perception in your mind of which horse you think has the best chance to win and you will probably be able to rate the horses in your mind from most likely to least likely to win the race. Let’s say you like the #1 horse the best, the #2 horse 2^nd best and the #3 horse third best. Does this mean you should bet the #1 horse? ABSOLUTELY NOT!  You need to evaluate your perceptions in relation to the betting line. Now let’s fill in the final factor in my fictitious horse race. The odds on the #1 horse are even money, the odds on the #2 horse are 2-1 and the odds on the #3 horse are 20-1. Now there’s no doubt that the #1 horse is the most likely to win the race, however in this analysis, you would be a fool to not bet the #3 horse. Based upon these odds, the #1 horse would have to win this race one out of every two times for you to breakeven. The #2 horse would have to win the race one out of every three times for you to breakeven while the #3 horse would only need to win the race one out of 21 times for you to breakeven. In my example, if you believe the #3 horse is the 3^rd best horse in the race, then the horse you should bet on in this example is the #3 at 20-1 odds even though you feel the #1 is the most likely to win the race.

Similarly, a bet in baseball should not necessarily be a bet on who you think will win, but rather who you think will outperform the line and odds and provides the best value. The most common error I see in many baseball bettors is they first choose who they think will win rather than evaluating the line. How often have we all heard someone say something along the lines of, “I love the Tigers tomorrow with Verlander and I am going to make a huge bet on them” before a line is even posted? What I am trying to state here is that there is no way to love any team until you know what the line is.

Take for example opening day with the Yankees hosting the Red Sox. It will probably be CC Sabathia against Jon Lester or Josh Beckett. Depending on the line, I could invest in the Yankees in certain circumstances or the Red Sox in others so it is clear my wager is not based upon who I think will necessarily win the game, but rather who based upon my handicapping, provides the best value. Said in a different way, I am shooting to determine which team I think linemakers and the general public are undervaluing based upon my techniques of handicapping. If the Yankees open up at EVEN in that game, I very well could play the Yankees. If the Yankees open up at -160, I could very well invest in the Red Sox.
 
I like to think of each game I handicap as based upon my research and understanding of a particular line, if the two teams in question played 100 times under the same fact pattern, how many time (what %) do I think in my mind each team would win, particularly, how many time will the underdog win. Here are the win percentages that an underdog would need to win in order to breakeven at a particular underdog line:

%Win         Breakeven Line
50%                 Even
48%                +109
46%                +118
44%                +127
42%                +138
40%                +150
38%                +163
36%                +178
34%                +194
32%                +212

I highlighted the 40% and 38% marks, because that is the line where a team in this situation would have to perform at the level of the worst team in major leagues over the course of the season. When I break down each game, if I feel a team will perform better than the line, then it is worth an investment. For example, if on opening day, I believe the Red Sox have a 46% chance of winning in my mind, if I can get reasonable value better than +118 (say, maybe 123 to 125+), then I am buying. Set in non-arithmetic terms, if I think there are factors that linemakers and typical gamblers are not factoring into the line of a particular game that I think are relevant to one side, it will often mean I believe that team will outperform the line and I will invest in them.

I have read a lot of gibberish that write-ups are tripe and just ways a poster tries to convince a reader on why to play a certain side. I view it differently. When I write games up, the write-ups were basically the relevant statistics which I saw as reasons why the side I was supporting was undervalued with respect to the line. It's not necessarily the reasons why one team will win the game, but analytical reasons why I believed the team I was betting provided better value than was seen by linemakers and the general public.  

I have no set formula for my handicapping and each day I am not sure where my research and thoughts will lead me. I like looking deep into the numbers and for angles that I think give me a better advantage than the current line. Take for example my write-up last season on the Cubs starting pitching. I have certain mental notes that I want to look closely at as the season progresses. This is just one example of the kind of things I look at:

Ted Lilly - A noted fly-ball pitcher. Vulnerable to the long ball. Look to fade against power hitting teams and especially when the wind blows out at Wrigley. Remember not to give him an edge against teams with strong left handed hitters, but an edge against predominantly right handed hitting lineups as ironically he is tougher against righties.

Carlos Zambrano - Tougher on lefties. Look to fade in day games as the last three seasons he has allowed more than 1.5 earned runs per start during the day. Streaky and has faded in August the past two years. Don't fade him when he is hot and fade him when he begins to show two consecutive bad outings.

Rich Harden - With injury concerns, Piniella will baby him. Look to fade Harden, at high prices in situations where the Cubs' bullpen is coming off of heavy recent use, as Harden will typically go 5-6 innings at best and will need to rely on the bullpen for 3-4 innings to close things out.

Aaron Heilman - Struggles with lefties. Look to fade him against teams stacked with strong lefties like the Phillies but play him against teams with heavy right handed lineups.

Sean Marshall - Road ERA is 1.25 runs better on than at home. He actually struggles against left handed hitters so look closely at opposing teams lineups.

Another thing I look to do is analyze the home plate umpires after the first game of a series (in the first game of the series they aren't announced until about 30 minutes before the game but after the first game, just look who the first base umpire was the day before, as he will be behind the plate the next day) and figure out their style and trends in relation to the starting pitchers trends. There are all kinds of umpire statistics readily available and many umpires will show generous strike zones or stingy strike zones. A stingy umpire matched with a pitcher with control issues or a pitcher that likes to nibble the corners can be a recipe for disaster while a generous umpire can be a recipe for success.

I incorporate countless other factors into my handicapping which would take hours and hours to try and explain and it might not even make total sense. Regardless of the factors you utilize as a handicapper, look for factors that are relevant to you which make you feel a team is being undervalued by the general public and linemakers. In your mind, never say, I love the Tigers tomorrow without knowing the line. If you love the Tigers that day, love them because you feel there are factors important to you that provide value in your investment in relation to the line and not just because you think the Tigers will win. When you look for value instead of just who will win, you will become a much better baseball handicapper.

You are missing the fact that the linesmakers know this and this kind of stuff is all factored into the line.  Rarely do you see teams that are +300 or higher in mlb, usually the highest on most days is around +200 or so.   While in the nba, you'll have dogs that are +1500 or higher on any given day.  If anything, mlb bettors take TOO MANY dogs because they are afraid to lay the juice even if it warrants being laid. 

If you're betting on the worst teams every day, there is no guarantee you will profit on them just like any other sport.  You need to find value in the underdogs you take, and likewise, find the value in the favorites.  Too many people in here stick to taking dogs only, when you really should be taking almost an even mix of favorites/dogs.   Just my two cents, gl this year game.

You could also pick out plenty of great starters that were profitable and plenty of soso pitchers that had losing years.  This test is not a good one as it is results-oriented.  You should take all the "good" pitchers as a whole, when doing a test such as this.

Felix Hernandez, Jon Lester, Josh Beckett, Justin Verlander, Josh Johnson, Adam Wainwright, Matt Cain, etc. were all big winners this year.  That does NOT mean they will be this year.  Some of those starters such as palmer and martin did not have many starts, so their numbers are very flukey.  Alot of times unknown pitchers DO have value however, just because people expect the worst out of these guys, so in those situations I would agree with game.  Also, the top guys really don't have much value as a whole unless they struggle early in the year (which definitely happens because of bad luck and what not)