The Battle Between Payroll and Parity
Over the All-Star break, MLB Commissioner Rob Manfred addressed the Oakland A’s, their quest for a new ballpark, and their remarkably low payroll. With regard to the last of those items, Manfred exhibited little concern, suggesting there was almost no correlation between a club’s capacity to spend money and its ability to win games. John Shea reproduced and retransmitted the following comments, care of Manfred, at the San Francisco Chronicle.
“I categorically reject the notion that payroll should be the measure of whether somebody is trying to win in our game today. I reject that not because I prefer low payrolls to high payrolls. I reject that because I know that the correlation between payroll and winning in baseball is extraordinarily weak.
“You do not guarantee yourself wins by having a high payroll, and as the Oakland A’s have showed, you can win with a low payroll. So I really reject the premise of that question. Those are the economic facts.
“Falling into this notion that payroll is a measure of whether an owner is trying to win is literally sophistry.”
I’ve got good news and bad news for the Commissioner. The good news is that, in six out of the last seven individual seasons, the correlation between wins and payroll hasn’t been very strong, as the graph below suggests.
Among those individual seasons is last year’s campaign. A year ago at this time, the numbers from 2016 to 2017 indicated a trend in the opposite direction. By the end of season, though, the correlation had dropped even further.
Last year, we saw some high payroll teams doing very well and some low payroll teams doing very poorly. We also saw two clubs (Cleveland and Houston) qualify for the postseason while spending relatively little. That, combined with the failures of teams like Detroit and San Francisco, suggests there’s some logic to the argument that payroll doesn’t equal wins.
The correlation last year wasn’t much to speak of, nor has the relationship been much stronger this season.
Wins and payroll in 2018 aren’t entirely unrelated, but the correlation isn’t a strong one. Unless things change significantly before the season’s end, we are going to have witnessed a relatively weak — though not “incredibly” weak, Mr. Commissioner — mark in six out of the last seven years. If you want to see the sport with a glass half-full, then write off 2016 as an anomaly and raise your glass to parity. If you want to know the horrifying truth, however, then read on.
The relationship between wins and payroll can appear weak in a sample of one season. Clubs that spend a lot sometimes fare poorly; clubs that don’t, like the A’s, sometimes play well. It happens every year. The effect looks a little different, however, if we consider several seasons together, to see how prolonged spending (or lack thereof) impacts winning.
A year ago, I noted that the correlation between wins and payroll over the previous three and three-quarters seasons was pretty high, at .58, as the graph below shows.
So after a decrease in the correlation from August to the end of the season, plus another lower figure this season, it would be reasonable to expect a drop in the 2015-18 numbers. That’s not what the data reveal, however.
The correlation remains virtually unchanged from last year, and it’s pretty strong. If the chart is a bit of a slog, here’s a few bullet points.
- There are eight teams with sustained winning records — an 84-win pace or better, in this case. Six of those eight teams had average payrolls of at least $140 million and all three of the biggest spenders are included in that group.
- There are eight teams with sustained losing records — a 74-win pace or worse, in this case. Six of those eight teams had average payrolls of less than $110 million.
- Of the eight teams with an average payroll of $150 million, five of them were incredibly successful, with a minimum of an 89-win pace per season.
- Of the 11 teams with average payrolls under $110 million, none of the them had sustained success and only two were around .500.
- Of the 13 teams with an average payroll of at least $140 million, nine teams were above .500.
No matter how you slice up the data, winning and payroll are linked. Teams like Oakland this year or Cleveland and Houston the past few seasons represent the exception. They don’t demonstrate that payroll is immaterial to winning. They demonstrate that sometimes payroll doesn’t matter, and that’s not the same thing. Worse, financial might is also showing a stronger correlation with winning. A year ago, the relationship between team value and wins over the previous four seasons looked like this:
One year later, it looks like this:
Of the eight most successful franchises of the last four years, all but one are in the top 11 when it comes to the value of the franchise. Of the bottom 10 franchises in terms of value, only one has a winning record since the start of the 2015 season. Ultimately, talent wins out. Teams with leaner budgets have certainly found ways to acquire that talent by means of the draft and international market — and retain it by signing young players to extensions. That has always been the case. The value of a franchise and the size of its payroll, however, account for about one-third of a club’s wins. Included in the other two-thirds is scouting, development, injury prevention, analytic strategy, shrewd acquisitions, and maybe a decent amount of luck. All that stuff is important. But 33% is still a substantial portion of a team’s success to have determined by financial might. What happens if and when the competitive advantages in the other two-thirds begin to dissipate?
Logic dictates that, with caps placed on draft bonuses and international spending, small-market teams have fewer ways to leverage what spending capacity they do have to find talent outside of free agency. Larger-market clubs, meanwhile, continue adapting to the modern game, with front offices embracing the kind of analysis utilized mostly by smaller clubs in the past. And while the worst hasn’t occurred yet, what happens next season when the Cubs, Dodgers, and Yankees all go over the competitive balance tax threshold, while the large-market Astros and Phillies increase payroll? What happens when the Orioles, Royals, and Tigers cut spending again? Is the relationship between money and wins going to get weaker or stronger? Owners with low payrolls might be trying to win, as Manfred has insisted, but owners with large payrolls likely will win, and those teams that don’t spend in the future might find the task of competing even more difficult than they do right now.
Craig Edwards can be found on twitter @craigjedwards.
Awesome article. I hope someone puts this in front of Manfred’s dumb face.
The real problem is that the MLBPA will fight a salary cap tooth an nail, even if it means that 40% of its work force never plays for a playoff team and the sport continues to alienate small market fan bases.
What’s good for the players and what’s good for the game are, quite often, mutually exclusive.
If someone tried to cap your salary at less than what your employer’s willing to pay, you’d fight too.
“If someone tried to cap your salary at less than what your employer’s willing to pay, you’d fight too.”
No. Not if I already had $50M in the bank and I was wise enough to realize unregulated salaries are bad for the future of my incredibly elite industry. But your mentality does explain a few things.
A salary cap usually comes with a salary floor and a (loosely) set share of total revenue to players. Right now the players would be far better off if they had taken such a deal 10 years ago.
There is a de facto soft cap in place with the Luxury Tax. It has served as the now widely accepted rationale for the paucity of spending this past winter.
The owners have the players coming and going. No salary floor. And a tax that stopped the bulk of big market teams from exceeding the threshold.
Not to mention a service-time structure that excludes players from seeking market value for their services during the most productive years of their careers.
A salary cap only makes sense if the owners aren’t making money and need to cut costs to survive. This is manifestly not the case. With perhaps a handful of exceptions, player salaries could easily be higher than they are now, which would not only benefit players, it would benefit the fans of mid- and small-market teams where owners continue to rake in a handsome profit while crying poor.
A requirement that clubs spend a certain percentage of their income on player payroll would ensure that teams are truly investing in the on-field product and not simply pocketing billions for their owners.
You have access to teams’ financial statements?
Salary caps are designed to do one thing – depress wages. If the intent is parity, aggressive revenue sharing and salary floors makes more sense.
Do you understand how aggressive revenue sharing would have to be to put into place aggressive salary floors under which small market teams could survive? Either make two leagues, or install a hard cap like every other US sport. Until then, MLB is not a sport; it’s a ludicrous contest between large corps and small corps. To many of us, that’s simply not entertaining.
For small market teams, a player like Jose Ramirez appears to be the new ideal. A prospect that arrives in the league with less fanfare and while showing promise produces at around a league average rate for his first two seasons. Thus, he signs a long term team friendly deal before experiencing a massive break-out and becoming one of the best players in the league.
Isn’t that ideal for every team?
True. But it would appear to be more critical for mid-market to small-market teams to identify impact players before their breakout and then lock them into long-term deals. For example, suppose that the Oakland A’s had drafted Aaron Judge. Given Judge’s incredible rookie season, and impressive sophomore effort, it is highly unlikely that Oakland would have been able to sign him to a team friendly extension prior to his arbitration years. Oakland has already shown they are not adverse to trading super stars once arbitration makes them more expensive. So, he’s likely gone following his pre-arb years. Where as a team like the Yankees will be more than content to pay him to continue mashing in Yankee stadium until he hits free agency, and possibly beyond.
You know we did draft him right?
All the teams are trying to do that. You are correct that it is more critical for the smaller market teams, but they don’t have any structural advantage to do so, unless you believe desperation is the mother of defensible innovation. So what it means is that smaller teams, on average, just do worse than the big market teams. Fine for owners, sucky for their fans and, arguably, bad for broad base support of MLB.
Very happy with how this article went as I was thinking after first few paragraphs that multiple seasons would have a stronger correlation to winning than individual only to have my thoughts confirmed.
Commissioner must be grateful for the Mets throwing a wrench in the correlations.
Categorically rejecting ideas makes it hard to analyze those ideas.
Once you go into multiple seasons, you likely get causation running both ways. Payroll drives success, but success also drives payroll.
If a team is successful developing good young players, after a couple of years their salaries rise. Likewise, if that team is winning, its revenues rise, and so it can support a bigger payroll.
Over a longer period, payroll is a pretty good proxy for talent on the team. This would be true even if no team ever signed another team’s FA. Given that, I’m not sure a .55 R-squared is that high.
my bet is payroll correlates with a tighter confidence interval around wins. Paying people more is risk management- you increase returns by higher EV AND lower variance.
Great comment. I think Manfred was mostly correct, though his statement could be improved by saying instead, “I reject that because I know that the correlation between payroll and winning in baseball is extraordinarily weak, HOLDING FRANCHISE VALUE CONSTANT.” Trying to buy wins through Free Agents is economically stupid and isn’t clear that it works absolutely. Adding 30+ year-olds on the decline on expensive multi-year contracts doesn’t add a lot of wins.
Instead of proposing payroll redistribution to promote parity, talent redistribution would work better. Allow bad teams to have 2 picks in the 1st round of the Amateur Draft, create an international draft and giving bad teams 2 picks in the 1st round there, and create an expansion draft for bad teams to draft a player off of good teams. These re-distributive methods would speed a bad team’s rise perhaps enough to counter-balance the incentive to be bad.
True to a degree. Especially over medium term and looking at delta spending within clubs. But payroll over a longer term, where you’d expect teams to go through multiple, natural good and bad cycles, payroll is going to largely be connected to the revenue the team’s market can support in both good and bad years, just averaged out. If those averages are higher, teams are going to be able to scrape together better teams even in the ‘bad’ years, and monster teams in the good years. While a team like the A’s, for example, has to more or less suck something god awful while they rebuild, but once rebuilt, they can be good and crank the payroll up a little to help add to it and sustain it. But absent the ability to sign say a Miguel Tejada to a $200M deal, they are always going to have go back to the cupboard of player development and cheap talent acquisition.
I would be interested in seeing a more complicated model, that includes just average payroll for a team over time with an inflation adjustment , current divergence of that payroll, giving us essentially a team specific payroll Z-score, and looking at wins. I’d bet other than a few small market clubs that are catching the first few years of competitive window with pre-arb or early-arb players, it would do very well. Other teams that fall off this line are likely to be teams that just plain have especially good or especially bad front offices.
@snapper Huh? Teams that develop young players well pay them peanuts compared to mediocre FAs. Having a great farm system leads to lower average salaries on a team, not higher.
I have found spending is really strongly correlated with market size. Sure winning helps, but looking across teams the media value of the market supports a lot of their spending ability.
Spending vs. Market Size: https://imgur.com/06UdlvO
Same chart, without top 4 markets: https://imgur.com/vtZDUBY
No way the Brewers or my Padres can sustain the spending of teams from top 5 or 10 markets. This is not crying for those owners, but for those fans. Owners will make their money, big or small market, winning or losing. Those fans, on the other hand, do not share in those profits, so not much fun.
As others have suggested: create a spending floor coupled with revenue sharing. To make the sharing less crazy, probably split the league into two, with the bigs vs. smalls sharing the MLB revenue for salary.
Why is SD’s market size so small there? It’s the 8th largest city in the country, and while its MSA isn’t huge, it’s the same size as the Minneapolis MSA, which is much higher up on the list
I live in San Francisco now, a city with a population of ~700k. The larger metropolitan region is around 5m, I believe. Some actual cities are quite small relative to their core metro region, others large. SD is the latter. The region is bounded by Mexico to the south, desert to the east, ocean to the west and Camp Pendleton (USMC base) to the north. So it may be a populous city (defined by boundaries of the city) but a small metro region.
How does “success drive payroll?” The differences in arbitrated salaries — which are the only ones a team has to worry about increasing — are minute compared those of veteran players, which teams have complete knowledge of ahead of time.
Please explain to us how the top 5 payroll teams are paying that much because they are great at developing players. The evidence simply isn’t there.
Well Manfred represents the owners and their incentive is to spend as little as possible.
But of course the correlation between money and success is growing. It was high before and then it dropped with sabermetrics because now smart small market teams can outsmart richer teams. However over the years almost all teams got super analytical front offices. There are still edges to be gained but they are much smaller and more short lived than they used to be.
And if the knowledge field is more equal than money becomes more important again even through free agency is less dominant.
Imo really the best time for baseball was the early 00s when there was still a mix out of smart and dumb front offices.
Fangraphs often attacked him but for both sabermetric writers and fans it was good that Dave Stewart existed because his dumb trades would make it more interesting.
While the small front offices have largely lost the competitive advantage they held over larger franchises by employing advanced analytics, I believe there are still market inefficiencies to exploit. The writers of fangraphs have expressed several times that emphasis on player development is the next big market inefficiency. The major league success rate of players taken beyond the first round is just comically low. Teams that can figure out how to raise that percentage by several points (through a combination of scouting and development) are going to have a huge advantage.
Yeah, its the next big market inefficiency, and the teams viewed by FG writers as the best at it are the .. Dodgers and Yankees.
A salary cap is meaningless unless the owners also had complete revenue sharing, which they never will. The Dodgers and Yankees will continue to have more money and better profit margins for their investments, so they can continue to employ an army of 40 or 50 analysts when the Pirates have 4.
Agree. I do think there is a case for the league to split between Large and Small and do revenue sharing and spending cap/floor within those groups, at least within the small. BTW, the smaller teams, like the Pirates, provide excellent, uncompensated value to the larger teams. They provide them someone to play and generally beat, creating a lot of value to the big teams. Imagine if the bottom 15 disappeared, MLB would become much less interesting for fans and less fun for the big market teams, assuming the loss of the bottom 15 led to more parity among the remaining. If I were rich enough to buy a franchise, I would work on that. The MLB player market could stay a single market, so better talent would gravitate towards the Large, but since Small is their own league, with own playoffs, etc. still fun for the fans. Then the World Series becomes the ultimate of David vs. Goliath, also probably more interesting for non-fans of the two finalists than the current set up.
This is true but that hole is also closing quickly with things like driveline going viral.
There are always going to be advantages to be gained but information lead is lasting shorter than ever. If a team finds something others will catch up because now they are open for change.
One advantage small teams have is they can experiment and fail. Things like the opener will always come from small teams because big teams would get killed by media if they fail unconventional.
Also data science is getting more expensive.instead of a single crazy programming genius they now have expensive teams and super computers costing a ton using big data methods to find small edges.
Right, and the Dodgers and Yankees have the money to do that. The Dodgers are hiring tenured professors and machine learning specialists from Google. Good luck, Rockies.
Player development as a frontier is likely to favor the rich, not the reverse.
Finding market iniquities as a way of getting a competitive advantage only works for low-resource teams when high-resource teams arent. When they are, its just another place they can spend the money.
I remember last year someone asked Dave Cameron what an effective baseball salary cap would look like, and his response was that the only actually effective cap would be a complete organizational spending limit, including minor leagues, coaches, executives, and everything, along with complete revenue sharing. Anything else and there will always be a place to invest those larger dollars for better returns on investment than the small teams can get.
How much more are the Yankees and Dodgers spending on analytics than the average team? I suspect (based on some limited inside knowledge) it’s far less than you think.
I hope the Wilpons see this and actually spend some money on the team
The wilpons could spend more considering the market but they actually spent more than the astros last year. (143m)
They aren’t good owners and should spend more in that market but mets fans always act like the mets spend like the rays and As.
Fwiw, in the case of the giants i have the feeling that spending a lot of money hurts at least as much as it helps. Management feels compelled to give long-term expensive contracts to keep or acquire talent and then continue to play the players even if they aren’t working out or getting old because they need to justify the investment Plus they routinely add no-trade clauses. I don’t think many other rich teams spend money this stupidly, but this is a possibility.
The two teams (SF and Det) mentioned as high prices failures are coming down off sustained runs of success. It’s hard to sustain a run of success without signing talent to long term contracts for short term gain, even if you know those contracts will have negative value sooner or later.
Would probably require a deeper dive on this question, but how much of that higher payroll is wasted dollars that’s not even contributing to wins?
Take my Red Sox – they have the highest payroll and the most wins. But they’re spending almost 40 million for Hanley Ramirez and Pablo Sandoval to not play in Fenway Park. David Price, although improved of late, is making 30 million a year for what’s ultimately been average performance. Dustin Pedroia’s earned 10 million plus to recuperate his knees. And let’s not even talk about Drew Pomeranz and his 8 million.
I don’t mean to cherry pick bad deals that have gone wrong, but the BoSox in an alternative universe could have close to the same record with perhaps $100 million less payroll. I see the correlation – I’m just curious if it’s more good rich teams are able to take on the risk of bad contracts and still have resources to stay good.
This was definitely the case with the giants in 2010-2014. Among pitchers, for example, they paid zito, lincecum, and cain huge sums of money for contributing very little towards winning championships and got carried through by people like bumgarner, vogelsong, and the relievers, who were making comparatively little.
That’s the biggest advantage that the high payroll affords – the right to screw up. Teams like Cleveland or Oakland can’t afford to have 40M in dead money, whereas for Boston it’s affecting them sure, but they can still build around it without too much trouble.
Which is kind of the point – a high payroll doesn’t mean you’ll win every year, but it means you can fix your mistakes and stay relevant for longer.
From a psychological point of view. It isn’t necessarily as simple as the right to screw up. The giants have been consistently unwilling to cut bait on people they’ve invested in even when it would clearly be the best thing to do. For years they kept assuming that lincecum and cain would make brilliant comebacks and basing their plans on this, even though obviously this wasn’t going to happen. The same thing is happening now with pence. It’s dumb but there it is.
Yeah, following up on what Matt said – the Red Sox can have $50+ million in dead money to Hanley, Panda and Castillo this year and still go out and spend $24MM on JD Martinez. When Cleveland screwed up by signing Nick Swisher and Michael Bourn and they went south immediately, it hamstrung them until they could get that money off the books.
Great point. Rich teams can afford to take ridiculous chances (Hanley and Pablo), have them blow up in their faces, and move on like nothing happened. Smaller teams do not have this luxury. One or two moves like that could destroy them.
The correlation is the incentive to spend. KC, HOU previously have shown that there no difference between 55 and 75 wins in the short term toward longterm success. CWS is doing it this year. They are not bad because they have a low payroll; they have a low payroll because they are not ready to be good.
I was shocked when Philly spent this offseason, but it apparently is their time to emerge from under the Ryan Howard rock. Atlanta has been shedding payroll to make room for their talent and will have some interesting decisions the next couple years.
The problem is that there is no incentive to be good other than trophies, and no disincentive to being bad other than some bad press. The NBA is a constant comedy of errors where marginal players sign large deals with bad teams who continue to remain bad, while good players settle for bargain deals to play in situations where they have the best chance to succeed. But even crappy replacement NBA players make $5M.
The Nats are far from small market, but they spent a lot of time to analyze and deem it was worth letting 2016 WPA God Adam Lind walk in favor of Matt Adams, to save $1M. That is 1 Scherzer start. The move paid off, as Zimmerman got hurt and Adams murdered righties for 3 months. Then, come August, they sacrificed his bat to STL for cash… maybe to the account of 1 more Scherzer start.
Bottom line- Manfred knows the current CBA allows small market teams to be competitive if they scout, draft, develop, and trade well. The carrot of hope is all they need to for those owners to justify going with farm one more year, and if it doesn’t work, fire sale, save $, and aim for 3 years down the road. It’s not small/large market. It is ROI and incentive.
“They are not bad because they have a low payroll; they have a low payroll because they are not ready to be good.”
At the same time though. If their market could support higher payrolls, they wouldn’t have to wait to be good.
No one said a small team *can’t* win. It’s just that they’re at an incredible disadvantage — one that does not exist in other sports. Your pep talk is akin to telling someone in KC that if he cannot sell his house for as much as someone can sell the same spec house in NY, he’s simply not trying hard enough.
In the offseason we read countless articles about MLB ownership was colluding to keep salaries low.
This is another argument against that notion (with the main argument the performance of last year’s free agents).
Executives have the right to be savvy, even if baseball writers and fans think they should just sling money around.
The modern correlation is indeed very weak. You should have ended the article right there. Instead you dug back into time to try to find something to support your flawed thesis and avowed opposition to Manfred.
This looks waaay too much like advocacy journalism instead of analytics.
So “modern” = 2018 and “outdated” = 2014-2017? mkay
Make two leagues with different caps.
So make both leagues boring? The second league will just end up losing their talent to the higher capped league anyway. It would literally be the exact same as what people view the MLB as currently.
Well, “literally” how is that more boring if it’s exactly what we already have?
The bottom line is that the current system works well for only one group: veteran players and and handful of young superstars. Young players, smaller-market teams and fans seeking genuine competition in skill (not in dollars or geographical value) are all getting screwed.
I feel like a lot of these discussion are around how to obtain parity and more frequent success with the small market teams. Salary caps/floors are often suggested.
I think the problem is player pay and how it’s structured. Quality pre-FA players used to be the key for small market teams, but that advantage is harder to get given that the large market teams now invest even more in analytics.
What if we opened the Rule 5 up more? Say a team can claim 30 players in the first round, 5 more in the second, 5 more in the third. With the worst 10 teams totally protected. The Rule 5 was designed to keep teams from stockpiling young talent, why not tighten that and skew it a bit to help bad teams. Of course, this could lead to lazy player development on both ends.
and/or maybe have a franchise tag. 1 pitcher, 1 position player. Each team gets $50M to allocate to those two players through rev share. To be fair to the player, another team can try to sign the player if they offer more. This could, theoretically, help small market teams keep the superstars they would normally lose after 6 years.
Both of those ideas I just thought up without a lot of vetting.
FWIW, I’ve always thought it would be great to expand the Rule 5 draft into a mini “expansion draft” like you proposed. If nothing else, it could help baseball remain in the public eye during the offseason.
So dumb by the commish. Look at the cubs. Their pitching staff alone costs $130M this year. That’s higher than a third of teams’ ENTIRE payrolls. Imagine if they didn’t have that empire weight to throw around- they’d still be rebuilding or trying to make shrewd moves like 26 other teams have to do every year. GMAFB commish if you think there is no correlation.
People love to point to the cubs rebuild as an example. It’s an example of a large market team intentionally losing and then spending over a half billion dollars. 3 other teams can pull that off, the rest have to play by a different set of rules.
Rays and Dodgers both with 67 wins today.
100 million payroll gap and the Rays are in the tougher division. Id say parroty is just fine.
Hen picking does not prove your point
Yes they are linked. Long term of course big payrolls help for sustainment. That’s the point of teams spending. But teams also tank (keep their payroll down temporarily realizing they can’t win at that moment.). Philly has tanked for a while and now they’ll spend. Oakland will never spend to sustain. But small payroll teams can be successful short term. That’s your parity.
Teams willing to spend shouldn’t be penalized.