Non-Linear Cost-Per-Win Part II

by Sky Kalkman on Jan 21, 2009 1:00 PM EST

After publishing my original post, I came up with another way of looking at the data. Instead of comparing a team's expected wins (based on payroll) to their actual wins, we can compare how much money a team actually spent to win their games to how much we would have expected they needed to spend. This will show us which teams saved money, which teams wasted money, and to what extent. The table should be pretty self-explanatory. All dollars are in millions:

Lg	Team	AdjWins	Exp Pay	Act Pay	Savings
A	Tampa Bay Rays	99	$243	$44	$199
A	Los Angeles Angels	102	$278	$119	$158
A	Minnesota Twins	90	$152	$57	$95
N	Chicago Cubs	95	$208	$118	$89
A	Boston Red Sox	97	$222	$133	$88
N	Florida Marlins	82	$100	$22	$78
N	Milwaukee Brewers	88	$140	$81	$59
N	Philadelphia Phillies	90	$157	$98	$58
A	Toronto Blue Jays	88	$140	$98	$43
A	Chicago White Sox	91	$160	$121	$39
N	Houston Astros	84	$114	$89	$25
A	Cleveland Indians	83	$103	$79	$24
A	Texas Rangers	81	$90	$68	$22
A	Oakland Athletics	77	$67	$48	$19
N	Arizona Diamondbacks	80	$83	$66	$17
N	St. Louis Cardinals	84	$110	$100	$11
A	Kansas City Royals	77	$64	$58	$6
N	New York Mets	87	$133	$138	-$5
N	Los Angeles Dodgers	82	$96	$119	-$22
N	Colorado Rockies	72	$36	$69	-$33
N	Cincinnati Reds	72	$36	$74	-$38
A	Baltimore Orioles	70	$28	$67	-$40
A	New York Yankees	91	$165	$209	-$44
N	Pittsburgh Pirates	65	$1	$49	-$48
N	San Francisco Giants	70	$25	$77	-$51
N	Atlanta Braves	70	$25	$102	-$77
A	Detroit Tigers	76	$58	$138	-$79
N	Washington Nationals	57	-$32	$55	-$87
N	San Diego Padres	61	-$17	$74	-$91
A	Seattle Mariners	63	-$8	$118	-$126

Some notes:

In order to expect to win as many games as the Rays and Angels did takes more money than the Yankees' payroll. And the Rays saved more money than any team but the Yankees spent.
Speaking of the Yankees, yes, they certainly wasted a fair amount of money ($44MM), but seven teams managed to waste more.
The Blue Jays ranked ninth in payroll efficiency. More fire-power for the anti-Ricciardi crowd (err, sarcasm).
Mr. Bowden, we understand that organizations put together a truly awful season every now and then, but did you really have to $55MM during a crappy season? The Twins spent $57MM and, well, were better.
I did not realize the Mets had the second-highest payroll last year. Where they're at, however, it's almost a no-brainer to spend a bit more. Wins in the 85 to 90 range are HUGE.

Tweet Comment 13 comments | 0 recs |

Do you like this story?

Comments

Dodgers MUST be wrong

ESPN assures us that Jeff Kent has “Hall of Fame credentials,” so they couldn’t have been wasting money there.

by klhoughton on Jan 21, 2009 1:57 PM EST reply actions

Sylvan

I’m not sure I understand how you arrive at the Expected Payroll figure. Does that “-$32” for the Nats mean that a team that bad should have a payroll of less than zero? That can’t be right, because even a replacement-level team would cost $10M or so, and wouldn’t win 57 games.

by psiogen on Jan 21, 2009 3:34 PM EST reply actions

The model is in the original post.

And yes, you’re right about the -$32MM meaning negative payroll. But we’re not talking about a replacement-level team. We’re talking about a team of average competence that only spent $13MM on it’s payroll. Those teams should still win like 65 games because of a decent farm system and trading all their arb-eligible players for more prospects. An organization with zero talent in the organization is just poorly run.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Jan 21, 2009 3:40 PM EST up reply actions

Ah, okay

So, I guess the real story of this chart is that the Nats have a below-replacement-level front office!

Your “original post” link doesn’t go anywhere, by the way. I had to scroll down the frontpage to find it.

by psiogen on Jan 21, 2009 3:56 PM EST up reply actions

Err, thanks. Fixed.

So, I guess the real story of this chart is that the Nats have a below-replacement-level front office!

Does that make Bill Bavasi Erubiel Durazo?

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Jan 21, 2009 4:02 PM EST up reply actions

I think an extra column that'd be interesting

would be the over / underpay as a fraction of the overall payroll – a team with more money to spend can afford to be less efficient with the the dollars they pay (ie, they can over pay to ensure they get the players they perceive to be the best on the market).

and so forth.

when did we stop using adverbs proper?

by alea iacta est on Jan 21, 2009 3:43 PM EST reply actions

I concur

That’s a very interesting way of thinking about.

Either way, kudos to you Sky, this is a really innovative idea.

The Crawfishboxes
A good friend of mine used to say, "This is a very simple game. You throw the ball, you catch the ball, you hit the ball. Sometimes you win, sometimes you lose, sometimes it rains." Think about that for a while.

by Stephen Higdon on Jan 21, 2009 3:58 PM EST up reply actions

Well, the fact that teams with higher payrolls spend more per marginal win is already built into the model.

Payroll $MM, ExpW, Cost Per Win
$13 67.5 $0.0
$20 69.0 $0.3
$30 70.9 $0.7
$40 72.8 $1.1
$50 74.6 $1.4
$60 76.3 $1.7
$70 77.9 $1.9
$80 79.5 $2.1
$90 81.1 $2.3
$100 82.6 $2.5
$110 84.0 $2.7
$120 85.4 $2.9
$130 86.7 $3.0
$140 88.0 $3.2
$150 89.2 $3.3
$160 90.4 $3.5
$170 91.6 $3.6
$180 92.7 $3.7
$190 93.8 $3.9
$200 94.8 $4.0

I’m not sure we need anything above and beyond that. It’s actually tougher for a low payroll team to waste money, because they tend to be worse. Under-performing by five games 75 to 70 doesn’t cost as much as under-performing 95 to 90, and that’s reflected in the chart.

Still an interesting thing to look at, perhaps. I may add it to a future post. Thanks.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Jan 21, 2009 4:06 PM EST up reply actions

What a tottally different way of determing if the GM knows what he is doing. Of course this assumes that the entire team stays off of the DL.

by Heap on Jan 23, 2009 12:24 AM EST reply actions

I do have a question though. I might be reading into this incorrectly but using the payroll break down as it is puts all players on the team on the field at the same time. It makes everybody a position player, where as starters typically go every 5 days. It also doesn’t take into effect salary being spent on players spending the entire season on the DL (Mike Hampton for the Atlanta Braves in recent years). We paid him but got nothing in return. Doesn’t that give an inaccurate projection?

If so is there a way to breakdown the formula to show at that each player will play only a certain amount of games so their salary should only be accounted for those games? After the season is over adjust the salary to reflect to show what portion of the salary played XX amount of games. Does that make sense?

Am I digging too deep?

by Heap on Jan 23, 2009 9:51 AM EST reply actions

Well...

It doesn’t really matter how often a player plays. Teams should pay their players based on how many games they’ll help them win. If that’s a starter providing a whole lot of value every fifth day or a position player provide a bit of value every day, it’s irrelevant.

I’m not sure what your point on Hampton is, but I agree(?) his salary should count, as that’s a player the Braves are paying, but aren’t getting production from.

The USA Salary Database definitely has some errors. That’s one thing I’m looking into. If anyone knows of better salary information, please let me know.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Jan 23, 2009 5:28 PM EST up reply actions