Best and worst MLB pitchers to bet on
Chad Millman
ESPN Insider
Name the best pitchers in baseball.
You may be compelled to give me an opinion: Roy Halladay, Tim Lincecum, Felix Hernandez. But there are many truths to that answer. Dillon Gee has the best winning percentage; Cliff Lee has the most strikeouts; Jair Jurrjens has the lowest ERA.
The beauty of stats is that they offer us an alternative answer to any argument that is centered on the premise of who is No. 1. And oftentimes the right answer, as opposed to the one you feel in your gut, will surprise you.
Unless you're a bettor, that is. <OFFER>If you wager for a living, you make it your business to know exactly who the best pitcher is in baseball every day of the season. You know who the best pitcher is on every staff. You know who the best closer is, who the best setup man is and who pitches best with men on base when they are wearing plain white undershirts, has a mom whose middle name begins with the letter Z and can make really great rhubarb pie.
As I've mentioned in this blog, column or whatever you want to call it, baseball is so hard for squares to handicap because it requires a level of devotion that is unattainable for anyone hoping to maintain gainful employment. It's not just about matchups, but about the value in those matchups. Too often you are asked to pay a premium on favorites, especially when a star pitcher is on the mound. For example, up until recently Lee hadn't been having that special of a year. But in 10 starts, he has yet to be an underdog. In fact, only once has Philly been offered as less than a minus-140 favorite. That's why so many wiseguys I know scope out those opportunities and look to play the dogs.
The other day I was chatting with my Insider editor Ben Fawkes (the truth is I told him I was approaching burnout and needed help coming up with an idea). He suggested finding out who the best pitchers to bet on at this point in the season have been. I thought that was brilliant. And because I am not just a sponge who steals ideas, but a trained journalist who understands how to build on them, I expanded his premise: I wanted to find out the worst pitchers to bet on, too!
I loved this premise because it highlighted another theme I am constantly harping on in this space: public perception versus betting reality. Amateurs bet names; sharps bet numbers. And it's rare that the two are the same. Bookmakers know this and they adjust their prices accordingly -- which is why consistently betting with the public will consistently make you a loser.
Step 2 in my process, after holding up Fawkes for an idea, was getting someone to do the work for me. For that I called Tom Federico, who runs TeamRankings.com. Federico and his buddy Mike Greenfield were two Stanford geeks (one studied math, the other engineering) who happened to be bigger sports geeks. They turned that passion and genius into something we can all benefit from: sports betting analysis. I profiled their philosophy and algorithm during the NCAA tournament.
My question for Tom was pretty simple: Do you track pitchers? And if you do, who are best ones to bet on or against?
His response: "It's in our database."
The next morning I had a comprehensive list of not just the best and worst pitchers to bet on or against, but of every MLB pitcher's performance in a variety of betting categories. "For each pitcher we pulled all of his starts, whether he got a decision or not didn't matter," says Federico.
Below is a breakdown of best and worst, based on a variety of different categories. All the data is through June 8 and all of the explanations are from Federico.
"Here are the pitchers who would have landed you the most profit if you bet $100 on the over every time …"
<!-- begin inline 1 --><TABLE><THEAD><TR><TH>Pitcher</TH><TH>Over/Under</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Jason Marquis</TD><TD>OVER</TD><TD>10-2-0</TD><TD>+$709</TD><TD>59%</TD></TR><TR class=last><TD>Jonathon Niese</TD><TD>OVER</TD><TD>9-3-0</TD><TD>+$518</TD><TD>43%</TD></TR><TR class=last><TD>Zack Greinke</TD><TD>OVER</TD><TD>5-0-2</TD><TD>+$455</TD><TD>65%</TD></TR><TR class=last><TD>Barry Enright</TD><TD>OVER</TD><TD>5-0-1</TD><TD>+$455</TD><TD>76%</TD></TR><TR class=last><TD>Nelson Figueroa</TD><TD>OVER</TD><TD>5-0-0</TD><TD>+$455</TD><TD>91%</TD></TR></TBODY></TABLE>
<!-- end inline 1 -->
"And the most profitable pitchers to take the under on …"
<!-- begin inline 2 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>Over/Under</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Jered Weaver</TD><TD>UNDER</TD><TD>10-2-1</TD><TD>+$709</TD><TD>55%</TD></TR><TR class=last><TD>Josh Beckett</TD><TD>UNDER</TD><TD>9-2-1</TD><TD>+$618</TD><TD>52%</TD></TR><TR class=last><TD>Philip Humber</TD><TD>UNDER</TD><TD>9-2-0</TD><TD>+$618</TD><TD>56%</TD></TR><TR class=last><TD>Paul Maholm</TD><TD>UNDER</TD><TD>9-2-1</TD><TD>+$618</TD><TD>52%</TD></TR><TR class=last><TD>Ryan Vogelsong</TD><TD>UNDER</TD><TD>7-1-0</TD><TD>+$536</TD><TD>67%</TD></TR></TBODY></TABLE>
<!-- end inline 2 -->
"The money line profits are a bit more complex. I pulled together the profits you would make based on several different bet-sizing methods. Here are the explanations."
1. Constant Bet Size: "Here, the same amount is wagered on every game, regardless of the odds. This ends up making your underdog bets more important (because the difference between winning and losing is larger). The five most profitable pitchers so far using this method (betting $100 on every game), have been …"
<!-- begin inline 3 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Jason Marquis</TD><TD>9-3</TD><TD>+$936</TD><TD>78%</TD></TR><TR class=last><TD>Kevin Correia</TD><TD>9-4</TD><TD>+$850</TD><TD>65%</TD></TR><TR class=last><TD>Dillon Gee</TD><TD>8-0</TD><TD>+$805</TD><TD>101%</TD></TR><TR class=last><TD>Aaron Harang</TD><TD>9-3</TD><TD>+$666</TD><TD>56%</TD></TR><TR class=last><TD>Kyle Drabek</TD><TD>9-4</TD><TD>+$620</TD><TD>48%</TD></TR></TBODY></TABLE>
<!-- end inline 3 -->
"The pitchers who would have lost you the most money …"
<!-- begin inline 4 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Clayton Richard</TD><TD>3-10</TD><TD>-$666</TD><TD>-51%</TD></TR><TR class=last><TD>Livan Hernandez</TD><TD>3-10</TD><TD>-$675</TD><TD>-52%</TD></TR><TR class=last><TD>Ubaldo Jimenez</TD><TD>2-9</TD><TD>-$757</TD><TD>-69% </TD></TR><TR class=last><TD>Paul Maholm</TD><TD>2-10</TD><TD>-$819</TD><TD>-68%</TD></TR><TR class=last><TD>John Danks</TD><TD>2-10</TD><TD>-$836</TD><TD>-70%</TD></TR></TBODY></TABLE>
<!-- end inline 4 -->
2. Constant Profit: "Here, you wager enough that your 'profit' on each bet is the same. For example, wager $50 on a +200 game, and $200 on a -200 game, so that the profit on each is $100. This ends up making your bets on favorites more important. The top pitchers in terms of profit for this bet sizing method (assuming your target profit for each method is $100) are …"
<!-- begin inline 5 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Dillon Gee</TD><TD>8-0</TD><TD>+$800</TD><TD>94%</TD></TR><TR class=last><TD>Jason Marquis</TD><TD>9-3</TD><TD>+$695</TD><TD>77%</TD></TR><TR class=last><TD>Roy Halladay</TD><TD>10-3</TD><TD>+$587</TD><TD>25% </TD></TR><TR class=last><TD>Kyle McClellan</TD><TD>8-2</TD><TD>+$561</TD><TD>45% </TD></TR><TR class=last><TD>Aaron Harang</TD><TD>9-3</TD><TD>+$560</TD><TD>44% </TD></TR></TBODY></TABLE>
<!-- end inline 5 -->
"And the worst …"
<!-- begin inline 6 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Paul Maholm</TD><TD>2-10</TD><TD>-$656</TD><TD>-61% </TD></TR><TR class=last><TD>Madison Bumgarner</TD><TD>3-9</TD><TD>-$680</TD><TD>-53% </TD></TR><TR class=last><TD>John Danks</TD><TD>2-10</TD><TD>-$891</TD><TD>-67% </TD></TR><TR class=last><TD>Chris Carpenter</TD><TD>4-9</TD><TD>-$911</TD><TD>-50% </TD></TR><TR class=last><TD>Ubaldo Jimenez</TD><TD>2-9</TD><TD>-$1096</TD><TD>-69% </TD></TR></TBODY></TABLE>
<!-- end inline 6 -->
3. Constant Wager + Profit: "Here, your bets are sized so that the sum of your wager and profit are equal for each game. This means that every game is equally important to your bottom line -- it doesn't matter which games you win, only how many. The top profits by this method come from the following pitchers (assuming wager+profit=$100) …"
<!-- begin inline 7 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>Over/Under</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Dillon Gee</TD><TD>8-0</TD><TD>+$394</TD><TD>97%</TD></TR><TR class=last><TD>Jason Marquis</TD><TD>9-3</TD><TD>+$393</TD><TD>78% </TD></TR><TR class=last><TD>Kevin Correia</TD><TD>9-4</TD><TD>+$321</TD><TD>55% </TD></TR><TR class=last><TD>Aaron Harang</TD><TD>9-3</TD><TD>+$297</TD><TD>49% </TD></TR><TR class=last><TD>Kyle Drabek</TD><TD>9-4</TD><TD>+$285</TD><TD>46% </TD></TR></TBODY></TABLE>
<!-- end inline 7 -->
"And the worst …"
<!-- begin inline 8 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Madison Bumgarner</TD><TD>3-9</TD><TD>-$313</TD><TD>-51% </TD></TR><TR class=last><TD>Chris Carpenter</TD><TD>4-9</TD><TD>-$351</TD><TD>-47% </TD></TR><TR class=last><TD>Paul Maholm</TD><TD>2-10</TD><TD>-$361</TD><TD>-64% </TD></TR><TR class=last><TD>John Danks</TD><TD>2-10</TD><TD>-$423</TD><TD>-68% </TD></TR><TR class=last><TD>Ubaldo Jimenez</TD><TD>2-9</TD><TD>-$437</TD><TD>-69% </TD></TR></TBODY></TABLE>
<!-- end inline 8 -->Of course, this is the thing about stats, especially when it comes to betting: Past performance is no guarantee of future results. In fact, Jason Marquis, who is dominating the "best pitcher to bet on" stats, is more likely to show a regression to the mean than he is to continue making anyone money.
"Also the sample sizes are so small," says Federico. "We don't consider it predictive. It's mostly interesting. If we were to expand the analysis over the number of years we would get [a bigger] number of starts and see something that could be meaningful, but the flip side is that the support crew and makeup of teams might change."
In other words, even with the stats, you can't take your eye off the ball.
Chad Millman
ESPN Insider
Name the best pitchers in baseball.
You may be compelled to give me an opinion: Roy Halladay, Tim Lincecum, Felix Hernandez. But there are many truths to that answer. Dillon Gee has the best winning percentage; Cliff Lee has the most strikeouts; Jair Jurrjens has the lowest ERA.
The beauty of stats is that they offer us an alternative answer to any argument that is centered on the premise of who is No. 1. And oftentimes the right answer, as opposed to the one you feel in your gut, will surprise you.
Unless you're a bettor, that is. <OFFER>If you wager for a living, you make it your business to know exactly who the best pitcher is in baseball every day of the season. You know who the best pitcher is on every staff. You know who the best closer is, who the best setup man is and who pitches best with men on base when they are wearing plain white undershirts, has a mom whose middle name begins with the letter Z and can make really great rhubarb pie.
As I've mentioned in this blog, column or whatever you want to call it, baseball is so hard for squares to handicap because it requires a level of devotion that is unattainable for anyone hoping to maintain gainful employment. It's not just about matchups, but about the value in those matchups. Too often you are asked to pay a premium on favorites, especially when a star pitcher is on the mound. For example, up until recently Lee hadn't been having that special of a year. But in 10 starts, he has yet to be an underdog. In fact, only once has Philly been offered as less than a minus-140 favorite. That's why so many wiseguys I know scope out those opportunities and look to play the dogs.
The other day I was chatting with my Insider editor Ben Fawkes (the truth is I told him I was approaching burnout and needed help coming up with an idea). He suggested finding out who the best pitchers to bet on at this point in the season have been. I thought that was brilliant. And because I am not just a sponge who steals ideas, but a trained journalist who understands how to build on them, I expanded his premise: I wanted to find out the worst pitchers to bet on, too!
I loved this premise because it highlighted another theme I am constantly harping on in this space: public perception versus betting reality. Amateurs bet names; sharps bet numbers. And it's rare that the two are the same. Bookmakers know this and they adjust their prices accordingly -- which is why consistently betting with the public will consistently make you a loser.
Step 2 in my process, after holding up Fawkes for an idea, was getting someone to do the work for me. For that I called Tom Federico, who runs TeamRankings.com. Federico and his buddy Mike Greenfield were two Stanford geeks (one studied math, the other engineering) who happened to be bigger sports geeks. They turned that passion and genius into something we can all benefit from: sports betting analysis. I profiled their philosophy and algorithm during the NCAA tournament.
My question for Tom was pretty simple: Do you track pitchers? And if you do, who are best ones to bet on or against?
His response: "It's in our database."
The next morning I had a comprehensive list of not just the best and worst pitchers to bet on or against, but of every MLB pitcher's performance in a variety of betting categories. "For each pitcher we pulled all of his starts, whether he got a decision or not didn't matter," says Federico.
Below is a breakdown of best and worst, based on a variety of different categories. All the data is through June 8 and all of the explanations are from Federico.
"Here are the pitchers who would have landed you the most profit if you bet $100 on the over every time …"
<!-- begin inline 1 --><TABLE><THEAD><TR><TH>Pitcher</TH><TH>Over/Under</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Jason Marquis</TD><TD>OVER</TD><TD>10-2-0</TD><TD>+$709</TD><TD>59%</TD></TR><TR class=last><TD>Jonathon Niese</TD><TD>OVER</TD><TD>9-3-0</TD><TD>+$518</TD><TD>43%</TD></TR><TR class=last><TD>Zack Greinke</TD><TD>OVER</TD><TD>5-0-2</TD><TD>+$455</TD><TD>65%</TD></TR><TR class=last><TD>Barry Enright</TD><TD>OVER</TD><TD>5-0-1</TD><TD>+$455</TD><TD>76%</TD></TR><TR class=last><TD>Nelson Figueroa</TD><TD>OVER</TD><TD>5-0-0</TD><TD>+$455</TD><TD>91%</TD></TR></TBODY></TABLE>
<!-- end inline 1 -->
"And the most profitable pitchers to take the under on …"
<!-- begin inline 2 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>Over/Under</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Jered Weaver</TD><TD>UNDER</TD><TD>10-2-1</TD><TD>+$709</TD><TD>55%</TD></TR><TR class=last><TD>Josh Beckett</TD><TD>UNDER</TD><TD>9-2-1</TD><TD>+$618</TD><TD>52%</TD></TR><TR class=last><TD>Philip Humber</TD><TD>UNDER</TD><TD>9-2-0</TD><TD>+$618</TD><TD>56%</TD></TR><TR class=last><TD>Paul Maholm</TD><TD>UNDER</TD><TD>9-2-1</TD><TD>+$618</TD><TD>52%</TD></TR><TR class=last><TD>Ryan Vogelsong</TD><TD>UNDER</TD><TD>7-1-0</TD><TD>+$536</TD><TD>67%</TD></TR></TBODY></TABLE>
<!-- end inline 2 -->
"The money line profits are a bit more complex. I pulled together the profits you would make based on several different bet-sizing methods. Here are the explanations."
1. Constant Bet Size: "Here, the same amount is wagered on every game, regardless of the odds. This ends up making your underdog bets more important (because the difference between winning and losing is larger). The five most profitable pitchers so far using this method (betting $100 on every game), have been …"
<!-- begin inline 3 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Jason Marquis</TD><TD>9-3</TD><TD>+$936</TD><TD>78%</TD></TR><TR class=last><TD>Kevin Correia</TD><TD>9-4</TD><TD>+$850</TD><TD>65%</TD></TR><TR class=last><TD>Dillon Gee</TD><TD>8-0</TD><TD>+$805</TD><TD>101%</TD></TR><TR class=last><TD>Aaron Harang</TD><TD>9-3</TD><TD>+$666</TD><TD>56%</TD></TR><TR class=last><TD>Kyle Drabek</TD><TD>9-4</TD><TD>+$620</TD><TD>48%</TD></TR></TBODY></TABLE>
<!-- end inline 3 -->
"The pitchers who would have lost you the most money …"
<!-- begin inline 4 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Clayton Richard</TD><TD>3-10</TD><TD>-$666</TD><TD>-51%</TD></TR><TR class=last><TD>Livan Hernandez</TD><TD>3-10</TD><TD>-$675</TD><TD>-52%</TD></TR><TR class=last><TD>Ubaldo Jimenez</TD><TD>2-9</TD><TD>-$757</TD><TD>-69% </TD></TR><TR class=last><TD>Paul Maholm</TD><TD>2-10</TD><TD>-$819</TD><TD>-68%</TD></TR><TR class=last><TD>John Danks</TD><TD>2-10</TD><TD>-$836</TD><TD>-70%</TD></TR></TBODY></TABLE>
<!-- end inline 4 -->
2. Constant Profit: "Here, you wager enough that your 'profit' on each bet is the same. For example, wager $50 on a +200 game, and $200 on a -200 game, so that the profit on each is $100. This ends up making your bets on favorites more important. The top pitchers in terms of profit for this bet sizing method (assuming your target profit for each method is $100) are …"
<!-- begin inline 5 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Dillon Gee</TD><TD>8-0</TD><TD>+$800</TD><TD>94%</TD></TR><TR class=last><TD>Jason Marquis</TD><TD>9-3</TD><TD>+$695</TD><TD>77%</TD></TR><TR class=last><TD>Roy Halladay</TD><TD>10-3</TD><TD>+$587</TD><TD>25% </TD></TR><TR class=last><TD>Kyle McClellan</TD><TD>8-2</TD><TD>+$561</TD><TD>45% </TD></TR><TR class=last><TD>Aaron Harang</TD><TD>9-3</TD><TD>+$560</TD><TD>44% </TD></TR></TBODY></TABLE>
<!-- end inline 5 -->
"And the worst …"
<!-- begin inline 6 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Paul Maholm</TD><TD>2-10</TD><TD>-$656</TD><TD>-61% </TD></TR><TR class=last><TD>Madison Bumgarner</TD><TD>3-9</TD><TD>-$680</TD><TD>-53% </TD></TR><TR class=last><TD>John Danks</TD><TD>2-10</TD><TD>-$891</TD><TD>-67% </TD></TR><TR class=last><TD>Chris Carpenter</TD><TD>4-9</TD><TD>-$911</TD><TD>-50% </TD></TR><TR class=last><TD>Ubaldo Jimenez</TD><TD>2-9</TD><TD>-$1096</TD><TD>-69% </TD></TR></TBODY></TABLE>
<!-- end inline 6 -->
3. Constant Wager + Profit: "Here, your bets are sized so that the sum of your wager and profit are equal for each game. This means that every game is equally important to your bottom line -- it doesn't matter which games you win, only how many. The top profits by this method come from the following pitchers (assuming wager+profit=$100) …"
<!-- begin inline 7 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>Over/Under</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Dillon Gee</TD><TD>8-0</TD><TD>+$394</TD><TD>97%</TD></TR><TR class=last><TD>Jason Marquis</TD><TD>9-3</TD><TD>+$393</TD><TD>78% </TD></TR><TR class=last><TD>Kevin Correia</TD><TD>9-4</TD><TD>+$321</TD><TD>55% </TD></TR><TR class=last><TD>Aaron Harang</TD><TD>9-3</TD><TD>+$297</TD><TD>49% </TD></TR><TR class=last><TD>Kyle Drabek</TD><TD>9-4</TD><TD>+$285</TD><TD>46% </TD></TR></TBODY></TABLE>
<!-- end inline 7 -->
"And the worst …"
<!-- begin inline 8 -->
<TABLE><THEAD><TR><TH>Pitcher</TH><TH>ATS record</TH><TH>Profit</TH><TH>ROI</TH></TR></THEAD><TBODY><TR class=last><TD>Madison Bumgarner</TD><TD>3-9</TD><TD>-$313</TD><TD>-51% </TD></TR><TR class=last><TD>Chris Carpenter</TD><TD>4-9</TD><TD>-$351</TD><TD>-47% </TD></TR><TR class=last><TD>Paul Maholm</TD><TD>2-10</TD><TD>-$361</TD><TD>-64% </TD></TR><TR class=last><TD>John Danks</TD><TD>2-10</TD><TD>-$423</TD><TD>-68% </TD></TR><TR class=last><TD>Ubaldo Jimenez</TD><TD>2-9</TD><TD>-$437</TD><TD>-69% </TD></TR></TBODY></TABLE>
<!-- end inline 8 -->Of course, this is the thing about stats, especially when it comes to betting: Past performance is no guarantee of future results. In fact, Jason Marquis, who is dominating the "best pitcher to bet on" stats, is more likely to show a regression to the mean than he is to continue making anyone money.
"Also the sample sizes are so small," says Federico. "We don't consider it predictive. It's mostly interesting. If we were to expand the analysis over the number of years we would get [a bigger] number of starts and see something that could be meaningful, but the flip side is that the support crew and makeup of teams might change."
In other words, even with the stats, you can't take your eye off the ball.
