Is Baseball’s Age of Parity Over?
If the postseason started today, five teams in the top half of major-league payrolls at the beginning of the year would qualify for the playoffs: the Boston Red Sox, Chicago Cubs, Los Angeles Dodgers, New York Yankees, and Washington Nationals*. That means that five teams in the bottom half of Opening Day payrolls would make the playoffs as well — in this case, the Arizona Diamondbacks, Cleveland Indians, Colorado Rockies, Houston Astros, and Minnesota Twins.
*Numbers current as of yesterday.
Presenting the standings in this way might give one the impression that we remain in an age of great baseball parity. An age in which the Kansas City Royals can win the World Series, Cleveland can get there, too, and teams like the Pittsburgh Pirates can sustain multiple years of playoff contention.
That isn’t quite the case, however.
Of the clubs that feature top-six payrolls this season, three have playoff chances of at least 96% (Dodgers, Red Sox, Cubs). A fourth, the Yankees, aren’t too far behind. If the Twins can’t hold on to a playoff spot and are overtaken by anyone but the Rays, the only team in the bottom 12 of payrolls this season to make the playoffs will be the Arizona Diamondbacks, and even their spot isn’t a guarantee. Money buys players, and those players rack up wins for their ball clubs. Last season, at around this time, I took a look at the relationship between payroll and wins, and noted that the relationship was one of the strongest we had seen in a while. This is what it looked like at the end of last season.
Last season saw one of the strongest relationships between payroll and wins to exist in several decades. Here’s how the relationship has developed since 1990, with help from data courtesy Brian MacPherson.
In the early 90s, Major League Baseball was coming off an era of collusion and lack of expansion. That, combined with a new influx of talent from outside the United States, meant that simply paying for major-league talent wasn’t the only solution to winning major-league games. (To track back further, read Dave Studeman’s piece in Hardball Times on the subject.)
At the height of the steroid era, when expensive players still performed at a high level, there was a strong relationship between money and payroll, but as teams became more savvy regarding the value of young players — and as PED testing ushered in a downturn in the performance of older ones — the relationship between payroll and wins on a single-season basis declined, becoming almost non-existent from 2011 to 2015.
Back in 2015, despite the weak correlation between payroll and wins on a single-season level, I expressed some concern about the relationship between a team’s overall wealth and win totals. Even if single-season spending didn’t correlate strongly with victories, multi-year spending did seem to have a positive effect, even then. Last season helped to validate those concerns, as money bought wins in a manner unprecedented in recent seasons. So far this year, the relationship hasn’t been as strong as it was last, but it is still exceeds the period between 2011 and -15.
The graph below shows this year’s numbers through August 22, with payroll figures from Cot’s Contracts.
This season hasn’t been as extreme as last year, and last season’s relationship did increase from this point in the season, but there’s still a pretty strong relationship. The biggest correlation with wins is always going to be talent, and there are a lot of ways to amass talent, but it certainly seems possible that, as teams reach a sort of analytical parity, that money can help be a deciding factor. Of the teams with a payroll above $130 million, only the Detroit Tigers, New York Mets, and San Francisco Giants have fewer than 60 wins; meanwhile, six of 10 teams with payrolls of $100 million or less are at that level. Some of those teams aren’t actively trying to compete. That said, all but Philadelphia are in smaller markets, have a really old stadium, or both. We can see some of this more clearly when we look at franchise values.
The chart below takes wins from 2014-2017 and compares that to the franchise valuations from Forbes.
The 11 teams above the average of $1.536 billion dollars have averaged 86 wins per season while the other 19 franchises have averaged 79 wins per year. For the bottom 10 teams, the average drops to 77 wins, and if it weren’t for Cleveland and Kansas City, the average would be down at 75.
At the other end of the spectrum, remove the Philadelphia Phillies, and the top-10 teams average 88 wins per season over a nearly four-year period. When we do the same exercise from 2014-2017 with payroll, we see an even stronger relationship.
There are certainly teams that can outperform their payroll over time, but having a higher payroll makes things a bit easier. Every team with an average payroll above $110 million over the last four years has a winning record except for the Philadelphia Phillies and Detroit Tigers, who are just one game under .500. Of the 14 teams with an average payroll below $110 million, only three — Cleveland, Houston, and Pittsburgh – -have winning records over the last four years. Some of this isn’t just big market versus small market. Teams that win tend to increase payroll to capitalize on their recent success, continue fan momentum and keep winning. We’ve seen the Mets, Royals, and (most recently) the Indians do this, while bigger market teams like the Astros, Cubs, and (more recently) the Phillies all carry low salaries as they rebuild. That doesn’t really change the fact that money is more tied to winning than it has been of late.
The relationship between money and winning is cyclical. We recently came out of a period where it didn’t matter as much, and we are currently in a period where it matters more. If we assume teams are on a level playing field when it comes to analytics, and we know they are on a level playing field when it comes to spending on amateurs, it could be that the difference on the field will be spending at the major-league level. However, we don’t yet know how the new Collective Bargaining Agreement is going to influence this trend. If the richest teams stay at or near the tax levels, we should see a bunch of teams in the middle rise up to meet them as revenues across baseball continue to grow. If many teams all have payrolls at close to the same level, other factors will come in to play when it comes to fielding winning teams. If the new tax levels are more of a mild deterrence than hard spending levels we could see the current relationship between wins and payroll continue, at least until small-market teams can figure out the next great advantage, whatever it might be.
Craig Edwards can be found on twitter @craigjedwards.
Maybe because it’s so much harder to “beat” other teams with trades than it used to be? Earlier, smaller markets could build effective teams by utilizing savvier baseball knowledge, but these days the gaps of market inefficiencies have closed up?
Not all the large-market teams are smart, but you now have big-spending teams who are also spending big on having the best front offices or front office teams (the Dodgers have what, 5 previous ML Gms in baseball operations?). When the Cubs and the Dodgers are out there using their wealth to not only go get the best available talent on the field, but the best available talent for the front office, it makes it hard to compete by just having a great FO.
You can criticize Theo or Andrew Friedman, but they’re certainly not definitively dumber than the smart small-market GMs the way that Ned Colletti was outclassed by Billy Beane.
Any evidence for using Beane to rip Colletti? Andre Ethier for Milton Bradley trade?
I wasn’t speaking to any particular deal between the two, but Colletti made a lot of questionable trades and Beane was the class of the league among small market GMs back then, before Friedman got the Rays going.
The Dodgers were in pretty poor shape when McCourt sold them – very little major league or prospect depth. Partly that was because the Dodgers weren’t really a big spending team in McCourt’s later years, but it took the new ownership throwing tons of money at the problem to turn the team around.
We are still in range of normal ebb and flow of wins and payroll correlation. It has shot up not because teams bought winners but because we just happen to have many of the top spenders going through unprecedented prospect success.
Yankees, Cubs, Dodgers and Boston have all had a ton of top prospects pan out lately and have started winning because of it. It’s not the money that is the cause.
Yankees, Dodgers, Cubs,and Boston’s success with prospects looks to be partly related to money. International signings, buying free agents without trading prospects as often, and front office spending all help.
I don’t think the front-office spending can be understated here. The edge that small market teams had for that period has mostly been flipped or equaled.
Part of the issue is that many of the low payroll teams are low payroll because they’re poorly run. Look at the Marlins, yes they’re not a huge market, but they’re a bigger market than they spent and have been horribly.
Also, being horrible leads to lower payrolls because being bad typically means that you win fewer games. Win fewer games, draw fewer fans, sell less stuff and most importantly get a smaller TV deal. So it’s a bit of a chicken and an egg thing.
That being said, there is a competitive imbalance and MLB as a business would be better off with more revenue sharing of TV contracts, along with a salary basement, meaning teams would be required to spend at least a certain amount. Or, and as a capitalist I love this one, they could blow up the draft and make everyone a free agent from day one. Can you imagine what Harper would’ve got on the open market or what would have happened to the GM who broke the bank for Todd Van Poppel? That would create some market inefficiencies to exploit and would stop exploiting young players.
Yeah, that’d be super-fun to watch the same five teams sign all of the first-round talent every year..
I think a big problem is that many big market teams now have smart managements, they don’t senselessly spend on aging veterans anymore but use the small market methods AND the big dollar.
The advantage of the small market teams was experimenting.the big market owners would stay with the proven old school way while the small market teams could try the numbers way. The big market owners however would wait until the modern way was really proven.
The small market advantage still exists to some degree as it is easier to experiment with shifts, bulpen roles and other stuff (epstein himself said 15 years ago that he preferred not to use a fixed closer but did not stick with it because the Boston media would be all over him if it backfired) but those advantages get fewer in between and also sticking shorter.
That is also a big difference.a discovery of the analytics department used to give you an edge for several years until the others would pick it up and buy into it. Now it is one season at best, coprorate secrets just stay behind closed doors as long anymore and even if everything stays tight the other clubs will pick it up via meta analysis.
That leads to a more level corporate quality playing field and thus to money dominating again. Late 00s to mid 2010s was a great age of parity and money is more even due to the local cable contracts but there are still differences and they are starting to matter more again.
A problem that can be completely solved by eliminating divisions.
and balancing the schedule
How much of this season’s r value of 0.58 is caused by the Dodgers? Largest payroll (by lots) and most wins (by lots), so they’re seperated from the other 29 teams and make the scatter plot look more linear.
Obviously you can’t just ignore what the Dodgers are doing, but one extreme team can really alter the r value.
Well, since correlation is a measure of how linearly-related two variables are, one outlier is not going to make it larger. In fact, a true outlier in either direction will reduce the overall observed correlation. So I would expect r to be higher than 0.58 if you removed the Dodgers.
So you think that if you removed everyone from the trend line who was above or below .500, r would be 1?
your statistics professor weeps
Googled around for an explanation. Go here:
http://www.theanalysisfactor.com/outliers-to-drop-or-not-to-drop/
Scroll down to #4–this is the problem that Phil is talking about.
Scroll up to the scatterplot above. It’s not really anything like the example in #4 (there does appear to be a relationship even if you drop the Dodgers), but it’s got some of that going on too.
Fun fact: You want to know a good way to identify outliers? Mahalanobis Distance! The same thing Chris uses to generate comps in KATOH. It is actually useful in the real world!
An outlier in one direction will reduce the correlation (to 0 in the limit). An outlier in both directions increases the (magnitude of) correlation (to +/- 1 in the limit). The Dodgers are clearly in outlier in both directions.
” If the Twins can’t hold on to a playoff spot and are overtaken by anyone but the Rays, the only team in the bottom 12 of payrolls this season to make the playoffs will be the Arizona Diamondbacks”
I don’t really think you can look at payroll in a discussion of parity, you’d just have to look at market size. The reason most bad teams have low payrolls … is because they are bad teams that know they aren’t going to be good and they are rebuilding, and not spending. It’s not like the Phillies are a small market and them not making the playoffs is evidence of a lack of parity, it’s just the Phillies know if they’d spent $40 million on free agents they’d just suck slightly less.
Bingo! The Cubs, Astros, etc. have proven that a low payroll/tank approach done correctly can lead to a quicker, lower cost turnaround (because you rebuild with young players under control for 6 years..See Houston still below average payroll 3 years in).
As recently as 3-4 year ago, teams like Philadelphia or White Sox would have signed a bunch of FA’s over the winter to try & get more wins, despite the evidence that they really needed to start over. Move forward to 2017 & that’s what they are doing.
There are variations but spending correlates very strongly to market size. A small market can spend more but they have an upper limit well. Those differences still drive outcomes.
I believe the new CBA will strongly aid parity in MLB. The tax and draft pick penalties are strong enough that they will function as a de facto salary cap. New CBA even reins in international spending.
Usually parity is discussed in terms of how many teams are successful. For instance, disregarding payroll, if the same 10 teams make the playoffs every year, there is no parity. If the same team wins the WS every year and a small group of teams never make the playoffs, you don’t have parity. The problem with focusing on payroll is you miss how the teams may alter their spending patterns in accordance with their talent. My favorite team, the White Sox, regularly places itself in the top 10 spenders when they are trying to compete (though they often don’t compete) but now with the team being torn down they have a payroll that is very close to the bottom of the league. In other words, there’s a chance teams spend because they win rather than just the other way around.
Have you considered if the inverse of your thesis is perhaps more true? Maybe teams are getting smarter in the opposite way, where the motto really is “if you’re not going to win, don’t spend.”
Money is always important, but it’s effect on wins in a particular season will vary.
We recognize that wins can deviate greatly from performance over a single season…that creates a LOT Of noise in analyses that evaluate single seasons spending and wins.
Plus, the way FA contracts are designed is to underpay at the beginning and overpay at the end…teams don’t expect to get the production they are paying for each year…so good teams get more expensive and worse as their players age on their FA contracts.
There are a LOT Of variables that contribute to win totals, the fact that money alone always explains at least 25% of the variation in win total makes it a VERY important variable, accounting for salary structures and the variation between performance and wins that regularly occurs over individual seasons show the true importance of money on being a winning team.
It is the single most important factor in baseball.
” the fact that money alone always explains at least 25% of the variation in win total makes it a VERY important variable”
actually the correlation only exceeds 0.5 (i.e. explaining >25% of variation) a few times. It’s important to remember that the correlation coefficient isn’t the proportion of explained variance.
Alternate title: “Why Salary Caps Are Good for Competition”
You must be new to Fangraphs 🙂
But with revenue sharing, TV and merch deals, shady stadium deals, draft and International caps, there is little incentive to get better for owners.
Loria drained the Expos, and his punishment was that he got to waste the prime years of a once in a generation hitter, using the Marlins and South Florida as his own piggy bank. Yet… nobody even cares, because the Marlins are just good enough not to make news.
Paris St. Germain is about to spend over €400 million just for the rights to negotiate and pay 2 players a combined ~€90 million per year for next n years. To put that in perspective, Barcelona received in compensation for a player that is maybe in the top 10 in his sport (probably not) the yearly payroll value of all but the Dodgers and Red Sox. Monaco is likely to get a similar deal. Yet if the Nats were to try in trade Harper, or the Angels Trout, they’d get a couple prospects and a bag of dicks.
I wonder how much of the single season data is skewed by the fact that teams who know that they are set up to lose do not spend in the offseason and then often shed salary at the deadline as opposed to teams that are already positioned to win who spend both in the offseason and at the deadline
None of the points on the scatter graphs are labeled with team names. It would be interesting to see connected graphs of each team over, say, a 15 year period. If that shows many different teams rising and falling in both $$ and wins over time that might indicate competitiveness is cyclical. Probably a few high spending teams would not show the rebuilding slumps but if most teams have a chance of winning for half their seasons but look really bad the other half, that is, in a way, competitive balance.
With so much technical information and statistics readily available to everyone, could scouting reports become more important?
What is this year’s correlation if you remove the Dodgers? They’re a pretty significant outlier both in spending and in wins.
R-squared would also be a better metric here for the importance of payroll
As a long suffering Padres fan (many decades), the relationship is very clear. I would also argue that there was not much of a “golden age” of parity. As the author noted, sustained spending (larger markets) tends to have good outcomes. I looked at share of spending vs. share of playoff appearances. I did 2003-2016 just because Baseball Cube had the payroll data for that range in a handy format. The correlation is .713 over that period. If you constrain the data to the “golden age” (I used 2008-2015, eyeballing the chart in the post), and even there the correlation is .525. Big and small spenders under/over perform year to year but spending is a big driver.
The data is here in case anyone is interested:
https://docs.google.com/spreadsheets/d/1GZOIwVPyyzV1NSsZZPFDLMlTwEXOc9Z0ECDKb73vM6Y/edit?usp=sharing
What about the WAR contributed to all of these high payroll teams by home grown players?
I don’t know if I buy that money is taking over the game again, one thing all of the big market playoff teams this year have going for them is that they’ve received huge contributions from cheap, young talent. If anything this season has been a demonstration that player development pays off, between these teams and the impending DET & LAA disasters.
Beyond that, it’s sequencing. The windows for these respective clubs in terms of that young talent have coincided. Yes, it helps that they have money to spend to fill in their rosters, but the Nationals are already on the backside of the youth movement curve, and the Cubs and Red Sox are entering it. I don’t expect this particular array of clubs to be all good at the same time for more than the next year or two.
More money makes it easier to recover from your mistakes. When San Francisco won its last World Series, it had more than $40M in nearly dead money on its roster and, even without those players, their playoff payroll was still about $110M. Many teams could not afford to replace that much payroll.
There is definitely data to indicate a correlation between wins and salary. That isn’t particularly new in professional baseball. But, I don’t think this necessarily means the end of parity. I would be interested in an article that looks at salaries and the current position of teams in their competition cycle. For example, the 2013 Astros were one of the worst teams in baseball and had one of the lowest payrolls. However, that was a team in the midst of a massive rebuild. Fast forward to the 2017 team which has a much higher payroll and is one of the best teams in Pro Baseball. More and more front offices are being run intelligently these days. Teams in the middle of a massive rebuild aren’t likely to spend much on salary as there is no point investing to get a few extra wins on the season until your core is in place. Once teams feel they are ready to compete, you tend to see payrolls increase accordingly. So, even with the correlation between wins and salary, I don’t think this necessarily spells the end of parity.