Book review: Trading Bases

A new baseball season means it’s a new baseball book season, and the new tomes are hitting the store shelves and Kindle electronic pages.

One particular work sparked some curiosity from me, Trading Bases: A Story About Wall Street, Gambling, and Baseball (not necessarily in that order) by Joe Peta.

As you might gather from the title, Peta, who is a lifelong baseball fan, has spent pretty much his entire life on Wall Street. When a car hit him while he was crossing the street and he was downsized out of his job in early 2011, he decided to spend his recovery time devising a system for betting on baseball games, which he then tried out during the 2011 season. The book is about his experiences.

Before going any further, I should note a few things. I’m not a gambler. I’ve never done any online wagering on any Vegas websites. I’ve never laid any bets down on a baseball game in my life. Aside from some family football pools and a few loose change poker games in college, it’s passed me by. Perhaps that’s why the book got my attention, since it combines something that I know about (baseball) with something I don’t (gambling).

Peta’s writing style might be the book’s biggest advantage. While you wouldn’t expect a stockbroker to be much of a writer, he has a nice conversational style that flows very well. Thus, no matter the subject, Peta keeps the reader engaged.

That’s nice because, while the book is primarily about baseball, Peta will bounce around and write various chapters focusing on his old life in the stock market, trying to show how the lessons he’s learned prepared him for his baseball betting, and demonstrating how Wall Street can learn quite a lot from sabermetrics. Fortunately, even if don’t know your Dow from your Jones and care not to, these chapters are still interesting. It must be nice to be that good a writer.

Those Wall Street chapters fit in well in this book because, in general, Peta adopts an episodic approach here. While there is the main plot of the development and implementation of his gambling system, some chapters are little essays and vignettes on the side that are designed just to showcase part of his interest in baseball and further flesh out Peta himself. For example, you get chapters on his favorite sports bar and on his family and baseball.

But the main focus of the book remains his betting on baseball. Peta is an advocate of sabermetrics, not surprising since he’s based his entire adult life on having hard data. What’s interesting about his betting system is how the basics of what he does are really quite familiar to anyone involved in the sabermetric community.

In many ways, this book is the story of the triumph of the nerd—the sabermetric nerd. He takes some well-worn concepts and bases a gambling system on it that (spoiler!) works pretty well for him. (It sure wouldn’t be much of a book if his system bombed out completely, now would it?)

Peta begins the year looking at a few basic nuggets of sabermetric wisdom to figure out how teams will do and which ones are over/undervalued. Pythagoras projections, WAR, DIPS, third-order wins—these things have been around for years, and they serve as the basis of Peta’s approach.

(Actually, though he seems well versed in sabermetric theory, Peta doesn’t use the term “third-order wins” and seems unaware of it. First discussed by Dan Fox in the 2006 Hardball Times Baseball Annual and then following him over to Baseball Prospectus, it takes Pythagoras a step further. It notes that some teams score or allow more/fewer runs than expected based on clutch hitting, luck, or other reasons not likely to be repeated. Adjust for these factors and Pythagoras, and you should have an idea how much a team over- or underachieved. Peta uses his pet phrase “clusterluck” to describe this phenomenon, which is a nice phrase, but he doesn’t even acknowledge the existence of third-order wins).

The foundation of his system if fairly simple, but Peta really gets involved in its implementation. He uses his system to rate what he thinks the odds are that a team will win each game, compares those odds to the Vegas line, and—based on the difference—decides whether, and how much, to bet.

This is where Peta goes far beyond what a normal reader of sabermetrics would. He has a pretty detailed system for viewing the overall quality of the team, the quality of the starting pitchers, etc., and he’ll do this for every game. As the year goes on, Peta gradually mixes real-season results with his preseason predictions until, by the end of the year, his system is entirely based on what has actually happened rather than on what he expected would happen.

One lesson Peta took from Wall Street with regard to his baseball-betting enterprise is never to risk the principal investment itself. Peta used to work for Lehman Brothers, which went belly-up because it didn’t do that. Peta describes how he designed his system to ensure he doesn’t make the same mistake. He decides a set percentage of his overall principal based on the edge his system gives him on the Vegas line. The largest bet he’ll make is two percent of his principal, and that only happens once or twice a week.

By and large, Peta does well. He ends up betting on a large number of games, and he ends up making money with his system. He’s 1,087-1,008 on the games he bet on during the season, which is only a 52-percent success rate, but that’s all he needs to turn a profit. He does especially well in his bigger bets, which makes sense because those are the ones he thought he should have the biggest edge. Because he does so well in his bigger bets, his return is around 30 percent profit on the year.

What’s more, while writing the book in 2012, the publisher gave him some of the publicity money for this book to bet with, so he went to Vegas and did well for a second straight year.

A Hardball Times Update
Goodbye for now.

Peta didn’t turn a profit all the time—he got killed all midseason long in 2011—but he was looking long term, and that paid off.

Purely from a gambling perspective, there are some interesting nuggets. Most notably, Peta does the math to show how betting on baseball by and large has a better return than sports with a point spread like football or basketball, because the house’s take of the action isn’t as much in baseball.

Also, during his summer in Vegas in 2012, he makes some interesting criticisms of the casinos and how they do a terrible job treating their regular customers. That’s one place Peta feels sports betting could learn from Wall Street.

While this book often reads like a triumph of sabermetrics—guy uses it to make money betting games all year round—I don’t know if that means the average sabermetrically inclined fan should try to replicate it. The key thing for Peta is less about having a general understanding of sabermetrics and more about applying it to every team for every game and then setting his bets accordingly.

Surely what he did can be replicated—heck, he pretty much gives all his info away in the book—but what Peta did takes a good chunk of work every day to stay on track. After all, if he was just a little less effective, instead of betting correctly on 52 percent of his games, he’d go under 50 percent and end up the year in the red. The devil is in the details, and that helped Peta win.

In all, this was a very enjoyable book and shows how a guy can use some common sabermetric knowledge to make some money for himself by regularly edging the Vegas lines. If that sounds interesting to you, I bet you’ll really like this book.

Newest Most Voted
Inline Feedbacks
View all comments
Phil Birnbaum
11 years ago

I didn’t know the expression “third-order luck” either, until I read it here.  I’m a bit behind in the state of the art, I guess.

Chris J.
11 years ago

Third order wins, Phil.  Third order wins.

Huh.  I figured it was better known than that.  I’d heard of it and I don’t do a very good job following up on this stuff, so I assumed it was out there.

phil birnbaum
11 years ago

Right, third-order wins.  Oops.  The phrase I wrote doesn’t even make sense.

11 years ago

I know “third-order wins” only because they show up on BP’s playoff odds page, which I check from time-to-time.

Joe Peta
11 years ago


I want to thank you for the very thoughtful review you gave my book.  As a reader of many comment sections, I’m a firm believer that the subject of a review should never rebut a review; the author had his say in the book.

So I’m only here to add to the discussion of 3rd order wins.  You suspected correctly that when I wrote the book, I was not aware of them, nor BP’s use. Like another comment mentioned, I too learned of them looking at the Playoff Odds report near the end of the 2011 season.  Further, in the 2012 Bill James handbook, he led a small discussion on hitting efficiency by team.  My “cluster luck” discussion is nothing more than a slight repackaging of these two topics.  Further, the sabermetric community has been aware of the importance of sequencing for quite some time.

So while it doesn’t break any new ground (and I tried to take pains to credit everyone whose work I used—including frequent THT contributor and personal favorite Matt Swartz) I would like to see this topic discussed more often.  We all know “10 runs = a win” “10% of fly balls are home runs” etc. but I never hear it said, “2 hits equals a run.”  But it’s true across MLB and very stable from year to year.

(Best example ever:  Seattle scored the franchise’s 25,000th run last year in the same game they recorded their 50,000th hit.)

Examining why a team is on the happy or sad side of 2 hits per run (hitting or pitching) through the regression I use in the book often reveals whether that state is skill-based (and therefore repeatable) or the result of fortunate sequencing (and therefore lucky and not likely to be repeated.)

Thanks again Chris for taking the time to both read and then write about Trading Bases.  I’m immensely grateful to anyone in the sabermetric community who picks up the book.

Chris J.
11 years ago


Gotcha.  Thanks for replying.

Nice book.

11 years ago

I am about 1/3 of the way through the book. It is well-written and Peta’s understanding and use of sabermetrics to create his models are solid.

There are a few (insignificant) things that he gets wrong, but mostly, so far, it is accurate, well-written, and an interesting read.

He’s probably figured this out already after 2 years of betting, but he bets way too many games, most of them likely having no edge at all. There is a lot of noise in the difference between his “line” and the sports book “line” such that a small difference is really no difference at all, once you include the (albeit small) “juice” (typically between 1.5 and 2%)

In other words, if the difference between his odds and the implied odds of the sportsbook line is say 4 percent, that really means that the true difference is something like 1 percent (as Phil B. found, which is correct, that difference has to be heavily regressed toward zero). Once you subtract the juice (I am assuming that the implied odds are not including the juice), you are left with no edge!

That is why only his larger bets, with a larger estimated edge, are the only bets which yield a positive return. And of course, the larger the perceived edge, the fewer bets you can identify.

IOW, in only a small subset of his large number of bets, at least in 2011 (I have not gotten to 2012 in the book yet, if he even discussed his results for that season) does he have any edge at all.

Plus, while his “Kelly” betting system is fine (it balances minimizing your risk of going broke, which is important in a long-term endeavor, unless you can replace your bankroll at will, with maximizing your compound return – increasing your bet size as your bankroll grows), making small bets even with a true small edge is silly.

Betting less than 1% or so of your bankroll is generally a waste of time. Peta, at least in 2011 makes bets of as small as .1% of his bankroll, which is ridiculous.

Here’s why: Say, your bankroll is $100,000. If you make a .1% wager, that is a 100 dollar bet. That doesn’t sound like too little, but the reason he is making a .1% wager is that his presumed edge on that bet is only 1% or so. That means that his expected return on that $100 wager is 1 slim dollar! It is not even worth the time to place that wager (unless it is completely automated, I guess, and your “overhead” is close to zero).

More importantly, when it comes to handicapping and estimating your edge, it is impossible to have that kind of precision such that you can distinguish between a presumed .5% edge and a 1.5% edge. That would be like if I told you that player A was a better fielder than player B because the former had a 3 year UZR of 2.50 and the latter, 2.51.

All that is moot anyway, since, as I already indicated, it is likely that a large portion of Peta’s wagers, at the low end of the spectrum, have no edge at all. That is simply because in only a small percentage of all lined games, in this day and age (when the lines are very good, especially in baseball) can any handicapper get an edge, if there is any significant juice at all (say, at least 1%). That is kind of a “law of sports wagering.”

Finally, a couple of minor mistakes in Chris’ review above:

“I’ve never done any online wagering on any Vegas websites.”

That would be true because there are no “Vegas websites” where you can wager online. Unless he means “Vegas style” online sportsbooks, which are “located” in various countries offshore, like Jamaica, Costa Rica, Antigua, etc. And of course it is likely illegal (in the U.S.) for them to do any business with persons located in the U.S., although it is not clear that the bettor himself is breaking any laws via the wagering itself.

There are some casino sportsbooks in Vegas where you can do “electronic” wagering, but you have to be located in the state of Nevada.

“He’s 1,087-1,008 on the games he bet on during the season, which is only a 52-percent success rate, but that’s all he needs to turn a profit.”

It is true that it is very difficult to gain an overall small edge (something like 5% would be brilliant – although it depends on the percentage of games you are wagering on) in a relatively small number of games, but in a sport, like baseball, where you bet on a “money line” (e.g., you might bet on an underdog where you wager $100 to win $150 or even $250, depending on the “odds” that the sportsbook sets) rather than a point spread, your record, like 1100-1000, is irrelevant to your profit/loss since these are not even money bets (where you would have to win almost 53 percent to break even to a “20 cent” line, your typical line for most sports, or a little more than 51% for a “dime” line, which many sportsbooks have for baseball). IOW, whether your w/l record on bets reflect an overall profit or loss completely depends on the average money line you are betting.

11 years ago

Thanks MGL for covering what I would have written.

A 52% win rate is junk, you need about an 80% win rate to survive.  The first year (loving it in Nevada) I was betting baseball, I was mainly learning the ropes and reading between the lines – and I also came out even, which like Peta had a lot more to do with conservative money management than picks (the Nationals were awesome at end of season).  It was the third year that really kicked in, much like getting to the show from the minors.  I don’t think Wall Street has an edge here, the two systems are very different – gambling doesn’t have stop orders, and the upside is fixed.  Wall Street has another game every millisecond and the season never ends.

Your best returns are certainly not from betting every game, but keeping Kelly in mind doesn’t hurt.

11 years ago

Kindled it.  Thanks for the review.  Can’t wait to see if it works.  Don’t even need to use money, can just work it like fantasy baseball.  Me vs. me.

11 years ago

“A 52% win rate is junk, you need about an 80% win rate to survive.”

I don’t know what that means. As I explained in excruciating detail (for those that are not familiar with sports betting), your win percentage is meaningless in money line sports like baseball.

If you bet mostly dogs (which most successful sports bettors do, at least in the past when the public tended to bias the lines toward the favorites) then a successful win percentage might be 50% or even 45% (again, depends on the odds/line of your average wager).

As far as 80%? If you are betting nothing but heavy favorites, then yes, you might need an 80% win rate, but it is extremely unlikely to get an edge by wagering on large favorites, at least more than a handful per season.

The success of a sports bettor is measured by his overall edge and the number of wagers per season, as a percentage of the number of available wagers. The overall edge is calculated by the dollars won (or lost, in which case the answer is a negative number) divided by the dollars wagered (dollars wagered could be risked or wagered, which will yield slightly different results).

In any finite period (one season or 10 seasons) your results are obviously a combination of your “true edge” (which no one knows) and random fluctuation. If a sports bettor can get a 5% edge or more and still wager on a significant percentage of games (say 25%), as I said, that is considered a very good result. Again, I am taking about true edge. If you have a 5% edge and wager on 500 games, heck, 16% of the time (one tail of a bell curve with a SD of .04 in your p) you will have a result of minus 3% or more! (A 5% edge means that your p is .525, not including the juice. So one SD less, .04, for 500 games, is a p of .485, which is a losing result.) This is why it is so difficult to estimate your true edge from your results. It is in addition to the uncertainty, a Bayesian problem. For example, if a typical sports handicapper (of many thousands) wins 5% in a season of 500 bets, frequent statistics tells us that the 5% edge is an unbiased estimate of his true edge, but Bayes tells us that there is a still a large chance that he has no true edge at all and just got lucky, for all kinds of reasons related to the “a priori probabilities.”

If you can get a 5% edge on 500 bets, that is a win of 25 bets of course. If your average bet is 2% of your BR (bankroll), which is not unreasonable, that is an ROI of 50%, which is not too shabby!

For a basketball or football bettor, where you usually make even money wagers (you can bet a money line too in these sports, although it is much less common), and the juice is typically higher than in baseball, around 4.55%, the break even point is of course 52.275%, so to have a 5% edge, you need a win rate of 54.78%, which is again, considered to be quite a feat, although many handicappers have boasted of win rates a lot higher than that.

While your win rate depends on the number of games you bet (If you want to have a higher win rate, only bet the small number of games where you think you have the highest edge, although that will cost you a lot of money, since you are ignoring the many more games in which you have a smaller edge).

For even money bets, a win rate of 80% over many wagers per season? Not even close to possible. There is no magic point at which the possible becomes the impossible in terms of true edge, but a win rate of 80% is impossible. 70% is impossible. 60% (edge of over 15%)? Unlikely. Possible I suppose. Certainly not for many games.

Since the goal of the sports bettor is to bet as many games as possible with an edge, he is never going to have a really high win rate unless he intentionally or accidentally (by setting the threshold too high – the opposite of what Peta did) does not wager on the many games with a lower edge.

11 years ago

I am 44% through the book. Too much rambling about the financial markets and things unrelated to the central thesis of the book – using sabermetrics to beat the Vegas baseball line, in my opinion. I realize that wagering on the stock market is similar to sports wagering (although I thought that it is impossible to “beat” the market), but a few pages here and there would suffice, for my tastes. Certainly someone else might have a different opinion on Mr. Peta’s stories of his stock trading experiences, family history, and music opinions and knowledge (he spends a few pages talking about Bruce Springsteen songs). Mr. Peta reminds me a little of Keith Law in this regard. I never much cared for Keith’s opinions about music, food and the cinema either.

One thing that really confused me is this. Maybe I am missing something and someone can clear it up:

He talks a lot about his childhood passion for baseball and about how much he loved going to games. He was brought up in Pennsylvania and was a Phillies fan. His father used to take him to games at the old Veteran’s Stadium, where the Phillies played from 1971 to 2003.

In the book, Mr. Peta writes:

“Ever try to define a color? …it is very hard…Emerald is the color of the sun-drenched grass you spot from the bowels of a baseball stadium…

…as a kid you equate grass with playing…I’d seen plenty of Phillies games on television by the time I was seven, and yet I couldn’t believe how bright green – yes emerald – the field looked when I spied it through the many ramps we walked up to get to our seats. We were in the infamous 700-level of…Veteran’s Stadium.”

Get where I’m going with this?

As far as I know, Veteran’s Stadium never had any grass! Am I missing something? And the artificial turf was known as being quite drab – almost grey or blue looking, wasn’t it?

So what’s going on here? Is the author making this story up? If yes, what else is he making up in the book? (I have a pathological hatred for people that make things up when they are supposedly telling a true tale. Not the people themselves, just the behavior…)

11 years ago

MGL, whoa!  If what you quoted is word for word and you didn’t leave out a significant detail with the dots you inserted, then as a nit pick, I don’t think he says that Veteran’s Stadium had grass.  He says he describes grass as being emerald and as a child he equated grass with playing.  Then he says his first glimpse of the field (note, no grass here) he couldn’t believe how bright green, yes emerald the field (still not grass) looked.  Still a nit pick.

Can’t wait to read the book and find this passage.  Might even give me an idea why it’s in there in the first place, but I guess as Chris said in the review, he does talk about family, etc.

11 years ago

I hope the book is shorter than MGL’s review(s).  As for MGL’s comment/rant, “That is simply because in only a small percentage of all lined games, in this day and age (when the lines are very good, especially in baseball) can any handicapper get an edge.” 

Evidence for this [that no baseball handicappers can get an edge in anything more than a small pct. of games]?  Apparently you don’t know many professional baseball gamblers.  Which is fine.  But don’t speculate on something you don’t know and then proclaim it as fact.  And how exactly would you know if the lines are “very good” vs., say, “mediocre”?

11 years ago

“Then he says his first glimpse of the field (note, no grass here) he couldn’t believe how bright green, yes emerald the field (still not grass) looked.”

Sounds like a stretch to me, especially since the turf at Veteran’s Stadium was not bright or particularly green. He didn’t specifically say that he was looking at grass, but he certainly strongly implied it.

I finished the book last night. I skipped over a lot of the Wall Street stuff (there was a lot in the latter half), as it bored me to tears (I have a particular distaste for anything to do with “Wall Street trading,” as it is a completely phony industry).

On the whole, as I said, it was well-written, interesting, and the guy is very bright and knows his stuff with regard to Wall Street trading and to some extent, sabermetrics. A bit too pretentious for my taste, although, I would recommend it.

11 years ago

MGL knows his stuff.

11 years ago

MGL, I’m only guessing here, but by “A 52% win rate is junk, you need about an 80% win rate to survive”, it sounds like the BR being used to bet with is the same as the ‘real-life’ BR/savings account. So I guess asym is incorporating costs of living.

Again, I’m only guessing…

Dave S
11 years ago


I’m born and raised in Philly… saw the very first game at the Vet with my dad, and the very last game with a friend.  And yes, while the Vet did get run down for a while, it still was striking to walk in and see that big green field… like it is in every ballpark!

Cut the guy some slack for romanticizing his old baseball baseball memories.  Nothing wrong with that.

11 years ago

Like I said, I have this pathological disdain for making things up (romanticizing) in a work of non-fiction. That’s just me. If you read the book, he focuses on the whole grass thing, because when he brings his daughter to a game, she says something like, “Daddy’s there’s grass!” Just seemed weird to me, and unnecessary.

The reason I think it is wrong to “romanticize” things in a non-fiction book is because once you are found out, the reader wonders what else in the book is true and what else is “romanticized,” and the credibility of the entire book comes into question. I believe it is considered a no-no by publishers as well. What was that book that was discredited, “A Million Little Pieces?”

Marc Schneider
11 years ago

This is a common complaint in reviews about non-fiction books-if the author makes a small mistake, how can you be sure he isn’t wrong on the big stuff? To a certain extent, I agree, but I don’t see how “romanticizing” his baseball experiences-if that’s what he did-calls into question the rest of the book. (And I’m not even sure he romanticized it.  He simply reported his own feelings about the stadium; whether they were objectively true or not really isn’t important.) It’s more of an issue when the author gets a specific fact wrong that he should know and that calls into question his knowledge of his subject.  In this case, I don’t see how mistaking the color of the field at the Vet-assuming he did-really calls into question his knowledge of his own gambling system.

11 years ago

Agreed … romanticizing the green field doesn’t bother me.  Maybe the author didn’t realize it was turf … wouldn’t change the validity of the story, really.

One thing I wondered about … in a footnote, he talks about Pete Rose being smuggled into the stadium to see his son play.  But Rose wasn’t barred from being a spectator if he bought a ticket … so what’s that all about?  Just an apocryphal story the author repeats?  Or did he get smuggled into the clubhouse or something?

11 years ago

” In this case, I don’t see how mistaking the color of the field at the Vet-assuming he did-really calls into question his knowledge of his own gambling system.”

I would agree. I over-reacted. Also, I don’t doubt that the field was green (I just looked at some old pictures, and the turf was indeed a bright green), my beef was that he implied that it was grass.

“Maybe the author didn’t realize it was turf … wouldn’t change the validity of the story, really.”

Clearly he knew and knows the difference between turf and grass. He is a self-professed baseball nut!

When an author in a non-fiction work makes things up (and I am NOT saying that the author deliberately made up this “story” or the notion that the Vet had grass) that don’t change the validity of the thesis in the book, it can still be a major faux pas. In this case, not so much. But in other cases, yes. Let’s say I am writing a book about baseball, or trains, or whatever. And let’s say that I am talking about my background, childhood, family, etc. I am not “allowed” to make things up about them even if they have nothing to do with the thesis of the book. If it comes out that I am misrepresenting or lying about anything in the book, for whatever reasons, it calls into question the veracity of everything in the book.

Again, I am not referring to this book at all. I think it is an honest book. I am referring to non-fiction books in general.

BTW, that is one reason why I don’t like some of Michael Lewis’ books. He clearly makes up and exaggerates so much. I should say that I find them very interesting and sometimes informative, but I never know what to accept as the truth and that bothers me.