
The Economics of Moneyball with Skip Sauer 10/23/2006
What this covers
Russ Roberts and Skip Sauer work through whether Michael Lewis's account of the Oakland A's survives statistical testing, drawing on Sauer's research with John Hakes. The first question is whether on-base percentage was cheap relative to slugging percentage, with sluggers like Mike Schmidt underrated because of low batting averages. The second is whether the A's actually exploited that gap. Sauer says the edge was real but not constant. It faded over the study period and was dissipated by competition, which he attributes to the statistic becoming widely known rather than to Lewis exposing a secret. He also takes up Steve Levitt's objection that the A's success came from pitching, conceding the fact but not the conclusion. Playoff results, he adds, are too random to judge a strategy by.
The middle of the conversation is about measurement. Sauer describes using betting market odds to estimate how much a starting pitcher shifts win probability, then checking that against Bill James's rough split of the game into offense, pitching and fielding. Voros McCracken's claim that pitchers control little on balls in play gets a qualified endorsement as a first approximation.
The last stretch asks why inefficiencies persist. Sauer offers order statistic chasing, entertainment value to fans, and weak owner incentives under baseball's quasi-monopoly, with Al Davis and slow American League integration as illustrations. Roberts adds digressions on seen versus unseen costs, Sam Peltzman's risk compensation, FDA approval, electrician licensing, and Iannaccone's view of religious competition.
Sauer argues the Moneyball edge in on-base percentage was a real, testable mispricing in the baseball labor market that closed once competitors adopted it, and that persistent inefficiencies in baseball are better explained by owners' weak incentives to win and quasi-monopoly protections than by managerial stupidity.
- Regression-style tests show walks and on-base percentage were underpriced relative to their contribution to winning in the early Moneyball years, and the anomaly disappeared by the end of the study period
- Stupidity explanations are rejected in favor of rational or entertainment-driven explanations that remain untested
- Weak ownership incentives and restricted entry into the league let teams underperform without going out of business
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Baseball's quasi-monopoly with restricted entry lets teams whose owners pursue non-winning goals, such as Al Davis using the Raiders for lawsuits or owners seeking stadium subsidies, persist, unlike in leagues with promotion and relegation where poor performers are replaced.
“we have a quasi monopoly system set up in in the States and that that can perpetuate you know I you know errors in management”
Racial discrimination in the American League likely contributed to its weaker competitive performance: many American League teams for a long time did not integrate, while National League teams integrated faster, and the National League won the All-Star game for years during the 1960s.
“you can be a racist which most teams in the American League were for a long time and as a result they lost the all-star game”
Occupational licensing for electricians reduced the number of electricians, raised their prices, led people to do more dangerous repairs themselves, and caused more accidental deaths, so licensing intended to protect the public can increase harm.
“they found that when licensing was more restrictive for electricians there were fewer electricians but more accidental deaths”
A bad-looking but seen event misleads: a shortstop who makes spectacular plays may be less valuable than one who makes routine plays and gets to more balls, so evaluating by highlight plays is misleading, just as trade debates focus on visible lost jobs and miss created jobs.
“what is seen misleads as to what is fully going on in international trade right you see the jobs that are lost you don't see the jobs that are created”
Peltzman-style risk compensation: when government makes cars safer through airbags, anti-lock brakes, or seatbelt mandates, drivers respond by driving more recklessly, so accidents may rise and pedestrians can be harmed, although the net effect on mortality is unclear empirically.
“that in turn encourages people to drive more recklessly and can lead to more accidents and potentially more fatalities”
On-base percentage was undervalued in the baseball labor market relative to its contribution to winning games, while slugging (extra-base hits and home runs) was overvalued, and Sauer and Hakes found the Moneyball claim held up to statistical testing.
“on-base percentage is relatively undervalued in by relatively we mean relative to its contribution to winning games whereas the ability to slug ... was relatively overvalued”
The undervaluation of on-base percentage was not constant: the A's were at or near the top of the league in drawing walks every year in the early period, but by the last year of the study period the anomaly disappeared and offensive skills were priced in line with their contribution to winning, as expected in an efficient labor market.
“the last year of the period that we study that anomaly goes away the offensive skills are valued in the marketplace”
FDA drug approval raises the cost of innovation, so drugs have fewer side effects but fewer drugs are produced and patients wait longer for access, and economists studying this find the net effect is deadly even though the lives lost to over-regulation are unseen.
“the net effect is unfortunately and tragically very deadly we don't see the lives that are lost due to over-regulation of the pharmaceutical industry”
Laurence Iannaccone's argument is that the Islamic world, despite a seeming state monopoly on religion, is more competitive than it appears: anyone can set up as a teacher, sects compete for adherents, and that competition supports high religiosity, unlike the monolithic state churches of Northern Europe.
“what we think of as a monolithic monopoly in a place like Saudi Arabia is in fact more competitive that it might appear from the outside”
The postseason, especially the initial five-game series, is close to a crapshoot with substantial randomness, so a team's roster-building strategy cannot be judged solely by its playoff success.
“the postseason particularly in the the initial five-game series stage is a crapshoot yes a lot of randomness”
Sauer measured pitcher quality using betting market odds: the impact of a starting pitcher on a team's win probability, holding the team constant, can be read from how the odds change when a different starter pitches, and the best starters shift win probability by roughly 0.15 to 0.225.
“your best pitchers have an impact of about 0.15 to 0.0225 or so”
Teams may be irrationally chasing the top of the talent distribution, but this need not be irrational: teams may be maximizing revenue rather than winning percentage, which Sauer says is consistent with baseball criticism of the A's for not trying to build champions.
“teams may not be so interested in winning right the whole thing they might simply be interested in maximizing their revenue”
Steve Levitt's main critique of Moneyball was that the A's offense looked average while their pitching was excellent, so their success came from pitching rather than the on-base strategy; Sauer concedes the pitching fact but argues it does not refute Lewis's on-base claim.
“his criticism wasn't directed so much at our paper but at the thesis in the book itself”
Lots of sluggers like Mike Schmidt were undervalued because they walked a lot and hit for power but had low batting averages; Schmidt's on-base percentage was around .350 to .400, so fans and pundits underrated him relative to his real contribution.
“someone like Mike Schmidt who is one of the greatest players of all time I think was underappreciated”
The betting-market-based measure of pitching productivity and the measure built from actual play-by-play outcomes weighted by James's contributions line up almost perfectly, which is an indirect test of James's weighting scheme that is only informative if the betting market is efficient.
“the two measures the probability based measure from the betting market and the probability based measure from performance in the actual plays that take place line up perfectly”
Bill James's split of baseball productivity, roughly half offense, a third pitching and a sixth fielding, implies fielding accounts for only a little over 15 percent of the game, and Sauer and Hakes found that the weighting lines up with the data when pitching is tested.
“so that leaves you with about a little over 15% of the game comes from variation in fielding skill”
Owner approval is not the main barrier to a bad owner being replaced, since other owners would likely welcome a more aggrieved owner's replacement if the league's quality were not harmed; the restricted entry system, not the approval process, is the binding constraint.
“if Al Davis does a horrible job yeah and embarrasses the league they'd love to still force him out yeah but let's say just us sort of like a really mediocre job”
The A's success came from getting on base cheaply: the on-base skill they were concentrated on was acquired inexpensively, which is the other half of the Moneyball hypothesis.
“they got there quite cheaply which is the other part of the Moneyball hypothesis”
Voros McCracken's claim that balls in play are largely random and that pitchers control only strikeouts, walks, and home runs is a good first approximation, but pitchers' ability on balls in play does matter somewhat; most pitcher ability shows up in strikeouts, walks, and home runs independent of defense.
“pitchers ability to convert balls in play matters somewhat but most of it is picked up by the strikeout walks and home runs measures”
The Moneyball edge was largely a matter of lost information rather than secret information: once Lewis's book described the approach and the statistic became widely published, rival front offices adopted it, and the advantage was dissipated by competition, so the book's disclosure of trade secrets did not itself cause the edge to vanish.
“by the time he was on to it I think other people were too and so it's just a matter of describing it for you know readers and not so much giving out information which would affect the marketplace”
Teams with the best players at each position may be following an order-statistic strategy, chasing the top of the talent distribution at each position, which causes them to overweight big, fast, power-hitting players and skip over solid, productive players who fill the roster gaps.
“they select what appear to be in the flesh you know really talented players ... trying to be uh the order statistic”
Owners and general managers may want flashy players who hit home runs and steal bases because fans enjoy them and they bring people into the seats, even though those skills cost some winning; undervaluing walks may therefore be a rational response to entertainment demand.
“they hire flashy players who say hit homeruns but strike out a lot or steal bases because the fans enjoy it”
Baseball is not truly competitive in the economic sense: owners face low costs for not winning, so teams can persistently underperform and still earn substantial money, which explains why a sport that looks competitive can sustain inefficiency for decades.
“I'm an argue baseball's not competitive and it looks competitive every team's trying to win supposedly”
Russ Roberts challenged Sauer's claim that owner approval is the main constraint on sales: he argued that owners might like a mediocre, non-threatening owner because the league's voting members benefit, so the approval process may protect incumbents rather than block bad owners.
“don't the owners kind of like that I mean why would as long as it doesn't degrade the quality of the league”
Sports leagues face a tension between competition and protecting the monopoly: if all owners are sleepy and avoid competing for talent, the bidding for talent stays suppressed and owners benefit, but if the league is too sleepy it becomes boring and attendance falls across the league.
“there's this weird tension in the sports leagues between competition and and protecting a monopoly”
Changing owners in a sports franchise is hard because owners are in an exclusive club with non-monetary benefits, and owners must approve any sale, which restricts entry and reduces competitive pressure compared with normal businesses where a bad performer is driven out by a buyer.
“that's a feature that retards competition in a serious way because if you have someone say fit well”
Owners such as Bill Bidwell are not forced out of their teams, because ownership yields large non-monetary benefits like luxury boxes and social status, and other owners must approve any sale, which reduces competition for ownership.
“he's in an exclusive club that has an immense amount of non-monetary right benefits for in terms of handing out goodies box you know luxury box seats to his friends”
Teams that pay for the single best player at each position overpay, whereas assembling a team through several small increments of productivity can produce similar output more cheaply, because offensive value is distributed across the whole lineup.
“if you avoid the syndrome of overpaying for the best by you know getting small increments that the the seventh best person or potentially the eighth best person then with increments and productivity”
The A's repeatedly lost productive free agents such as Jason Giambi, Miguel Tejada, Mark Mulder and Keith Foulke, and replaced them with cheaper alternatives by reconfiguring the roster holistically, making several small upgrades across positions rather than chasing the single best player at each one.
“they didn't replace them perfectly I didn't find somebody as good as Jason Giambi because they couldn't afford to”
Fan awareness of advanced statistics has changed dramatically because USA Today and daily newspapers now list on-base percentage among league leaders, which Sauer thinks has put pressure on general managers to pay attention to it.
“USA Today and other publications now list on-base percentage among the leaders which is really a revolution”
Baseball's clash of cultures between statistical analysts and traditional scouts reflects a persistent problem: when evaluation methods have little common ground, a person who is skilled with models may be poor at judging players in person and vice versa, making integration hard.
“the kid who's good at tweaking models in pulling out information from a laptop may not be too skilled and evaluating a player in the flesh and vice versa”
Baseball's competitive market should have eliminated the mispricing of on-base percentage quickly because teams compete for talent, but the persistence of the anomaly for a decade suggests the market was less competitive than it appears, a puzzle Sauer flags as an interesting question he is skeptical about.
“the market is one that you you think would be quite competitive as these teams try and crawl over each other to get to the top of the heap yet somehow this strategy appears to have been under exploited until the 8th came along”
Economists should not default to stupidity explanations for persistent market mispricing; instead the competing rational explanation (entertainment value or a different objective) should be treated as a horse race and tested empirically, such as by examining attendance patterns for teams that steal bases.
“I'm loath to adopt stupidity theories that's not the way I do business I'm an economist”
The A's edge is likely to erode because the Moneyball insight is now widely known, and unless the A's have a comparative advantage in discovering new strategies, the on-base trick will not work anymore as other teams adopt it.
“everyone knows that that problem now so unless it's the case that you know there's some other objective”
Sauer argues the Moneyball thesis was testable and refutable, which is why he and Hakes tested it empirically; he treats Lewis's claim that on-base percentage was underpriced as a falsifiable proposition rather than a story to accept.
“that's a testable proposition and capable of being refuted and so Hakes and I what about that task”
Lewis's claim that the Oakland A's gained an edge by focusing on an underappreciated aspect of offense is that avoiding outs is critical, and on-base percentage measures the ability to avoid making outs and reach base.
“an underappreciated aspect of offensive productivity was simply the ability to avoid making it out and if you're going to avoid making it out you're going to get on base”
In the NFL most revenue comes from the national television contract rather than local revenue, so teams like the Cincinnati Bengals in the 1990s could collect a check without working hard, while baseball's larger local revenues weaken this incentive problem.
“most of the revenue comes from the NFL contract the television contract that they have and a minor share comes from local revenue”
Walks as valuable as hits is an old baseball saying that is not exactly right, since a walk advances a runner one base while a hit can advance two, but the skill of drawing walks has long been recognized as valuable at every level of baseball and yet was undervalued by the market.
“walks as good as hit yeah we know it's not exactly because a walk can only advance a player one base”
Standings and payroll are inversely related in baseball: teams that spend the most do not reliably win the most, and the A's record of winning with low payroll is evidence that there is probably more than luck involved, which even a non-statistical reader can see.
“you look at the standings in terms of wins and then you look at the standings in terms of payroll and they're inverted”
Sauer is reluctant to adopt the stupidity theory of front offices because economists work with alternative assumptions; the stupidity theory and the entertainment-value theory are a horse race, and the entertainment theory has not yet been tested.
“I'm loath to adopt stupidity theories that's not the way I do business I'm an economist and we make our living with alternative assumptions”
The A's showed that if everyone else adopts a high-investment strategy, a team can exploit the others' errors by applying Yankee-style ingenuity at low payroll, and winning raised attendance and ticket prices, so winning itself generates the entertainment value that drives revenue.
“they increase their attendances by winning there's a lot of entertainment in winning”
A possible explanation for the undervaluation is that teams are not trying to maximize their winning percentage each year but are trying to take a chance at being at the top of the table, which Sauer calls a behavioral explanation.
“teams are interested not in in in say maximizing their winning percentage per se this this year but in taking a chance at at being at the top of the table”
Kevin Youkilis led the majors in pitches seen per plate appearance, and batting first with that patient style may have been a positive externality for the Red Sox lineup because it made other hitters more productive.
“you make your other hitters in your lineup more productive and so you know he was batting first for that reason probably an externality”
Earned run average is a cloudy measure of pitching quality, and wins are heavily weighted in awards such as the Cy Young despite being a poor measure of individual pitching skill, which reflects an award process that prefers familiar statistics over pure skill measures.
“winning percentage is a cloudy statistic and yet for siyoung Awards its heavily weighted”
Economic statistics suffer from the same problem as baseball's old metrics: news outlets report wages but ignore fringe benefits in compensation, so the reported picture of pay is incomplete.
“be like asking America's newspapers to report details on fringe benefits when they're talking about compensation instead of just focusing on wages”
Broadcasters repeatedly criticizing players who swing at the first pitch shows that the idea that patient hitting matters has become a cliche, which indicates the Moneyball insight is now widely understood by the public.
“that's a sign that it's become cliche when they start harping on it”
Lewis's book did not give away a still-valuable trade secret: by the time it was written other teams were already on to the approach, so the book mainly described a known edge for readers rather than revealing information that would affect the market.
“by the time he was on to it I think other people were too and so it's just a matter of describing it for you know readers and not so much giving out information which would affect the marketplace”
Yogi Berra may have said that pitching and fielding account for most of baseball, but Sauer does not know the source of the quote.
“who was it that said that I don't know it might have been Yogi Berra”