The first time you find yourself immersed in a group of sabermetricians is a magical experience. For me, it was the SABR Analytics Conference in March 2013. I have vivid memories of just about everything that happened that weekend: the roar of the crowd when the U.S. played Mexico in the World Baseball Classic, the patience of the numerous baseball luminaries who gave me the time of day even after talking to umpteen other kids, how lost I got on the first day when I assumed the streets of Phoenix were laid out in chronological order of the Presidents for which they were named.

Another particularly strong memory is of a question (I think originally proposed by Brian Kenny) that became the go-to conversation starter among us networking students over lunch and between presentations: “Over/under on 95 wins for the Astros in 2018?”

Considering the question again now, taking the under would have been a no-brainer. Projecting 96 wins or better from any team in any season takes a pretty big leap of faith, let alone five years in the future for a team that was about to lose 111 games. But at the time it felt like a tough call. And in the nearly two years since the conference I have remained fascinated both by the question and by the fact that I had initially found it so difficult.

After a bit of recent introspection, I think I’ve figured out why I had such faith in the Astros that I would have considered taking the over on that hypothetical bet. And this line of thought has led me to a theory about straying from the pack and trying new ideas in baseball: So long as it comes with the prospect of a substantial reward, riskiness in a long-term strategy is not a vice but a virtue.

The Astros, Experimenting, and You

Looking back at my younger self, I don’t think it was the specifics of what the Astros were doing that made me such a strong believer in their front office. The mere fact that Jeff Luhnow & Co. had the guts to blow up and reshape the organization in the most dramatic rebuilding process in my memory was enough for me to be predisposed to think highly of them. That I liked most of the specific moves the Astros had made was of less importance in shaping my opinion of the franchise than the appearance that they were absolutely devoted to the bold plan of totally tearing down the team and building a better one from the bottom up.

Broadly speaking, in a game that’s surprisingly (and often frustratingly) resistant to new ideas, I greatly admire those who are willing to forge new paths. Other recent examples of bold trailblazing could include appreciating and teaching the skill of catcher framing, implementing defensive shifts, and investing in baseball operations talent. And when the smart people who run a team think highly enough of a new philosophy to challenge the industry orthodoxy, I almost always agree that it’s worth a shot.

The response I typically hear when I express these sentiments is something along the lines of: “Experimenting for the sake of experimenting isn’t a good thing.” I’ve always thought this was missing the point of what I was saying. But as I’ve thought about it over the last few weeks, I’ve actually come to agree with this straw man. So long as the change has a plausible chance of helping the team, I think going with a bold new strategy is worth it just for the sake of experimentation. Here’s why.

Understanding Teams’ Goals

One of the central themes of my writing over the last year has been the distinction between the market price of a win and the utility-based value of a win to a team. When discussing the wisdom of a transaction, only the former is of real importance. Unless the player in question adds a substantial amount of personal name-brand value to the organization (in which case you’re not really paying for his wins), it is never rational to pay $8 million for a win if you could instead get one for $6 million, even if theoretically you’d be willing to pay $10 million to win an extra game. (Yes, this is an oversimplification of how the free agent market works.) No matter how much intrinsic value an asset has, in a competitive market your maximum bid should be the market value of the good, not what you would give up to get it in a vacuum.

But that doesn’t mean these subjective preferences are irrelevant to evaluating organizational strategies. Yes, every team wants to make money. But most MLB teams are owned not by corporations but by people. Owners care about making profits, but non-corporate purse-string-pullers also care about other things — chief among them being winning.

Having a successful team means different things to different people. Some owners might see “success” as winning 90 games at least once every two years; others might be happy to settle for good-not-greatness if it means minimizing seasons spent below .500; and still more might think the whole year was a waste if it doesn’t end with a World Series trophy. (Before you ask, I just made these categories up and they are not meant to describe anyone in particular.) None of these approaches are necessarily wrong; the point is that differences can exist.

Perhaps this would be different if I were actually in that position (“heavy lies the crown”) but my goal as an owner would be to make the playoffs as often as possible. What happens from Game 163 on might not be totally random, but it sure as hell isn’t predictable with any reliability. I wouldn’t be happy if my team made the playoffs 10 times but never won so much as a pennant, but I’d at least be satisfied in knowing that we did as well as could practicably be expected.

For the purposes of this article I’m going to use making the playoffs as our standard for a successful season. Quibble with it if you like, but either way I don’t think it takes away from the point of the forthcoming thought experiment.

More to Gain than to Lose

It’s been said that the goal of the social sciences is to restate the obvious in fancy prose. Like a good economist, then, allow me to point out two blatantly apparent facts that will be important to keep in the front of your mind. The first is that, without knowing any details about any of the 30 MLB teams, each organization would have a 33-percent chance (1 in 3) of being one of the 10 teams that make it to at least the play-in game. The in-a-vacuum odds of getting to at least the divisional series round would be 27 percent (4 in 15). Making it to October thus means beating your odds much more than failing to gain a playoff berth means not meeting expectations.

The second, related point is that a league-average team would have very little chance of making the postseason. According to Tom Tango’s quick playoff odds calculator, we would expect a .500 true-talent team to make the playoffs just 21 percent of the time. Even making the extremely generous assumption that every team is average in talent, our team would be bound by the same odds given above — 33 percent to make the play-in game, 27 percent to make (what is now) the traditional postseason.

I offer these seemingly unnecessary explanations because they lead to what I think is an important, less-obvious (though probably not novel) truism: There is a difference between how much improving or weakening a team by the same magnitude affects its playoff odds. This goes beyond the concept of the win curve — the point isn’t that making the team better does more good at different initial talent levels but that the effects of enhancements and subtractions are asymmetrical. And generally speaking, adding an extra projected win matters more than taking one away, which means that the average team has more to gain than it has to lose.

To get a quick idea of just how much win swings of different magnitudes matter to a team’s playoff odds, I picked a few plus-or-minus win totals from 81 and simulated 1,000 162-coin-flip “seasons” against 29 .500 true-talent teams. I then gave the experimental team credit for a playoff appearance each time it finished with one of the six best records in the group and half-credit for earning a seed between seventh and 10th to reflect the uncertainty of the play-in game. (Yes, this simulation is overly simplistic, but I don’t think it’s too far off the mark and it reflects the fact that we have no idea how good any individual team will be several years down the line.) I then compared each team’s odds of getting to the divisional series to the baseline of 4 in 15 for a true-talent 81-win team:

Starting with a delta of one win, the 82nd projected win does more than four times as much good to a team’s playoff odds as the 82nd projected loss does harm. Beyond that, the greater the magnitude of the risk, the more likely you are to make the postseason — you’ve got a lot more to gain from the gambit working out than you have to lose from things going sour. In other words, if you were given the choice between standing pat or taking a 50/50 gamble on a long-term strategy that would hurt or help you by the same number of wins, you want to take that chance. In fact, a risk might be worthwhile even if the potential detriment is of a greater magnitude than the possible gain, or if the odds of it working out favorably are less than even.

Risks Worth Taking

If you look far enough into the future, you’d have to project every team to be average. Ignoring a few wrinkles (which we’ll get to in a moment) the playing field will be level. Now think about your long-term strategy. Do you want to take the safe, do-no-harm route? Or do you want to shake things up? If you know that basic probability says the odds are stacked against you, trying something risky and new isn’t just justifiable — it’s probably in your best interest.

Think of all the strategic gains that could be had from putting ideas that have been deemed interesting but unrealistic into practice: Giving better nutrition to minor leaguers. Training young pitchers in effective velocity. Opening knuckleball academies. Restructuring pitcher usage, meaning anything from adapting to four- or six-man rotations to doing away with the concept of a pitcher throwing more than two or three innings per game. Perhaps investing more heavily in front office talent. I’m not saying all these ideas are shoo-ins to work or even that each would be worth the trouble to find out. But these are the kinds of things that team executives should be thinking about trying if gaining a long-term edge is more important than avoiding falling behind.

Especially when you consider that even in the long run, things aren’t fully equal for one big reason: money. Hold everything else constant and you’ll always be better off betting on the Yankees or Dodgers or whoever steps up next as the game’s biggest spender. The smallest-budget teams aren’t starting out as equals in the long-run projections — absent any other information I’d project the Athletics and Rays for maybe 78 projected wins in 2025, while I’d peg Los Angeles teams at 84. And if you’re starting at a disadvantage, you have even more to gain and even less to lose by taking a risk on a new way of doing things.

“If we do what the Yankees do,” Billy Beane is quoted as saying in Moneyball, “we lose every time, because they’re doing it with three times more money than we are.” In the context of the book, Beane is talking about paying market value to players in free agency and acquiring pricy established stars via trade. But the statement also rings true as a general statement about how MLB clubs do business. For low-budget teams to beat substantially richer opponents, they have three choices: get lucky, do it better or do it differently. And if and when the former two aren’t practical strategies, the only thing left is to take innovative risks.

Looking Ahead

“We know there are repercussions from straying from the herd,” Brian Kenny also said at the 2013 SABR Analytics Conference. Kenny’s point (as I remember it) was that teams should worry less about the inherent risk (among other deterring factors) in trying out new strategies. I’ll go a step further: Given the right combination of potential gains and probability of success, teams should actually embrace risk in organizational philosophies.

Thinking realistically, I’m not holding my breath for teams to start making riskier moves. After all, planning for long-term uncertainty probably isn’t a priority when GMs and their advisors are making multimillion-dollar investments on behalf of high-profile companies whose investors and clients are eager for quick success — not to mention that risk-averseness is basic human nature. But if the goal is to put a potential championship team on the field as often as possible, it’s worth the risk of falling below mediocrity for a chance at gaining the strategic upper hand.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
I recently finished reading Nate Silver's The Signal and The Noise and the latter chapters of his book go a bit into your line of thought. If you have a strategy that is high risk/high reward, then, barring initial conditions, it should be tested out. Then again, initial conditions cannot be ignored and have to be factored in. That's were probability kicks in and determines whether the risk is worth the reward. In baseball, once money rolls into the picture, then teams with low payrolls (and eventually those with deep pockets) will have to try and beat the odds every now and then. Great Article!
Isn't your conclusion simply a result of the fact that less than half the teams in MLB make the playoffs? If you ran the same exercise for the NBA or the NHL, where more than half the teams make the playoffs, I'm guessing you'd find that a more risk-averse approach was the best strategy.

When every team is average, and you need to finish in the top 10 of 30 to be considered successful, then your highest expected return will result from chasing a win while risking a loss. When every team is average, and you need to finish in the top 16 of 30 to be considered successful, then your highest expected return will result from avoiding a loss while risking a win.
Yeah, I thought I had made that clear in the calculations. Average isn't good enough in baseball, though it is in other sports. Though you could argue that, in other sports, making it to the playoffs isn't a lofty enough goal when it's relatively easy to do.
Blatant self promotion: I wrote about related things a month or so ago, based on simulating random seasons, and came to the same conclusion about risk. Unless your team has a talent level around 89 wins or higher, risk is good.

You know what they say about great minds. Thanks for the link, that was a good read!
Thanks. Of course they also say something about fools, but I can't remember what and I'm sure it isn't relevant.
Owners are incentivized to try risky things because odds are they'll be around for multiple anomalies (good and bad).

GM's, on the other hand...
This would seem to give merit to AJ Preller's offseason strategy.