Odds and Ends
“The only correct actions are those that demand no explanation and no apology.” – Red Auerbach
Over the last few weeks, I’ve received plenty of feedback related to my articles on sacrifice flies and the collective bargaining agreement: To Go or Not to Go, Scoring on the Sacrifice Fly, and Competitive Balance and the CBA. Because of that feedback, or maybe because I’m just not that original, this week I’ll answer a few of most frequently asked questions about both sets of articles.
Competitive Balance
First, several readers commented about my use of payrolls in order to perform the correlation. I should mention that the salaries that I used to calculate the payrolls in the Lahman database are those from the start of the season, and do not include in-season signings and trades. While this isn’t ideal, it does avoid the problem of using end of the season salaries, as these tend to overestimate the payroll differences between good and bad teams, due to bad teams shedding payroll to the good teams around the trading deadline.
In any case, the main point is that in my study, payrolls were acting as a proxy for market size, or market revenue potential. In other words, the link between payroll and winning percentage is not that surprising, since better athletes on the whole command higher salaries. But payroll has an implied link with the characteristics of the market the team plays in as well, and that is a variable that teams cannot control. The two are definitely linked, but they are not the same thing. As the Blue Ribbon Commission of 1999 noted:
“Although a high payroll is not always sufficient to produce a club capable of reaching post season play … a high payroll has become an increasingly necessary ingredient of on the field success …”
A more proper study, like that conducted by Donald F. Leypoldt, Jr several years ago, would include the revenue potential of the market, since payroll is under the discretion of management, so even a team with plenty of money can have a fire sale or squirrel away their revenue sharing money. I certainly concede the point, but I chose to use payroll, since that was the data that was on hand, and overall it serves as a good stand-in for the size of the market.
Second, reader Lou Poulas offered a second way of looking at competitive balance, and that is simpy through team records. Using this approach takes payroll and market out of the equation, and as a result illustrates the pure spread of the teams over time. The following table shows the number of teams with the specified number of wins during the last three CBAs (the number of wins were adjusted for 1994 and 1995 using the team’s winning percentage based on 162 games).
Category 1990-1996 1997-2001 2002+ 40-49 0 0 1 50-59 3 1 6 60-69 19 31 19 70-79 71 43 27 80-89 55 36 31 90-99 35 29 27 100+ 7 8 9
The table can also be converted into percentages (all rounded up, so each column does not total to 1.0), since the total number of teams went from 190, to 148, to 120 in the three successive periods.
Category 1990-1996 1997-2001 2002+ 40-49 .00 .00 .01 50-59 .02 .01 .05 60-69 .10 .21 .16 70-79 .37 .29 .23 80-89 .29 .24 .26 90-99 .18 .20 .23 100+ .04 .05 .08
As you can see, the percentage of teams with 90 or more wins went from 22%, to 25%, to 31% in the three periods, with the percentage of teams with more than 100 wins also increasing. This is reflected in the standard deviation of winning percentage, which increased from .281, to .329, to .331 over the three periods. At the same time, however, the percentage of teams with fewer than 69 wins has basically remained the same since 1997, although those with fewer than 60 wins has increased six-fold.
Third, our own Dave Studeman posted a variant of the graph I constructed on his site. In his analysis, he took out the Yankees and the four expansion teams from the 1990s, and finds that the coefficient of variation declined during the most recent CBA.
Although I hesitate to disagree with Dave, I generally dislike throwing out data points, since it can be argued that by doing so you’re influencing the results of your study. In this case, some might argue that the Yankees should be excluded, since they’ve made it clear that they’re not going to let little things like revenue sharing and a luxury tax get in their way. At the very least, however, the Yankees, one of the two teams playing in the largest market, should be included, but have their payroll capped at just above the maximum of other teams.
On the contrary, one could argue that the Yankees are a big part of the impetus for creating the revenue sharing and luxury tax system in the current CBA in the first place. As a result, excluding them would render any results fairly meaningless.
With that said, there are signs that the times are a changing. After all, the Yankees did not end up signing Carlos Beltran last year, in part at least, because of the luxury tax as reported by the Washington Post:
“Yankees officials acknowledge that they were constrained by two of the changes adopted three years ago—revenue-sharing and a penalty against high-spending clubs known as the luxury tax. ‘We had priorities this winter—primarily, improving our starting pitching—and we feel we met those priorities,” Yankees President Randy Levine said. ‘We’re like every other team, even though our revenues are larger than other teams. We’re conscious of revenue sharing and the luxury tax.'”
And it has been widely reported that the Yankees lost $50 to $85 million in 2005 and $37 million in 2004, totals that are less than the amount they contributed via revenue sharing and luxury tax—a bill that came to $110 million last year.
Finally, several readers asked about solutions to the problem of competitive balance. While in the article I showed that competitive balance hasn’t really increased in the last four years, I do think that the current system has helped, in the same way that even a partially clogged sink holds back some of the water. Without the provisions of the last CBA, can you imagine what the Yankees, Red Sox, and Angels teams would look like with the spigot turned on full?
But I do agree with Dayn Perry at Fox Sports, who argues that more needs to be done, and it seems to me that a good start is upping the percentage of local revenues shared to 50%. Equally important, however, is that teams should be held accountable for the way they use the money they receive. The late Doug Pappas, in reviewing the 2002 CBA, wrote that:
“Each club receiving revenue sharing is required to ‘use its revenue-sharing payments in an effort to improve its performance on the field,’ subject to unspecified penalties from the Commissioner if it doesn’t.”
So although Mr. Selig has the power, it seems he either doesn’t have the mechanism or the will to hold teams accountable.
I don’t agree with proposals that would force a minimum payroll, as that could just as likely cause a team like the Royals to spend money on third tier free agents, rather than invest in the future by signing international players or expanding their scouting and minor league programs. Wait … maybe that’s a bad example.
Sacrifice Flies
After my two articles looking at sacrifice flies—an interesting but admittedly small part of the game—several readers wrote to ask how teams stacked up against each other. Their reasoning for wanting to analyze this information was to see if some teams, perhaps via their third base coaching, were more risky and therefore enjoyed better success rates than others. Sounds reasonable, and so I totalled up the same set of sacrifice fly statistics I looked at in the previous articles, but did so for each team from 2000 to 2005 and for each individual team season.
First, at an aggregate level, here are the teams sorted by the percentage of time they were successful when the runner was sent.
Opp Scores OA Hold% Succ%
TEX 356 280 7 0.194 0.976
SFN 386 304 8 0.192 0.974
NYA 320 256 7 0.178 0.973
LAN 275 223 7 0.164 0.970
DET 361 293 10 0.161 0.967
NYN 300 230 8 0.207 0.966
TOR 340 282 10 0.141 0.966
SDN 353 283 11 0.167 0.963
CHN 329 257 10 0.188 0.963
KCA 385 307 12 0.171 0.962
PHI 343 263 11 0.201 0.960
ARI 334 276 12 0.138 0.958
ANA 392 288 13 0.232 0.957
ATL 339 282 13 0.130 0.956
MIL 293 233 11 0.167 0.955
SEA 409 332 16 0.149 0.954
TBA 332 253 13 0.199 0.951
FLO 354 271 14 0.195 0.951
OAK 340 271 14 0.162 0.951
CHA 357 290 15 0.146 0.951
BAL 359 287 15 0.159 0.950
CLE 350 281 15 0.154 0.949
BOS 396 316 17 0.159 0.949
MON/WAS 308 236 13 0.192 0.948
MIN 335 269 15 0.152 0.947
CIN 274 227 13 0.124 0.946
COL 345 278 16 0.148 0.946
SLN 376 307 18 0.136 0.945
PIT 314 236 14 0.204 0.944
HOU 360 282 17 0.169 0.943
As you can see the spread from the team that was most successful, the Rangers, and that which was the least successful, the Astros, varies from 2.4% to 5.7% over the period. The average for all teams, as I mentioned previously, was 95.6%. From an overall perspective, you don’t really see major differences in success rate, largely because the overall success rate is so high and also because in a larger sample like this, variation tends to decrease.
It is mildly interesting as well that Seattle had over 400 opportunities to score on sacrifice flies, while Cincinnati had just 274. I wouldn’t have expected such a large disparity.
Now let’s take a look at the same list but sorted by hold percentage, defined as the percentage of time the runner did not try and score on a sacrifice fly opportunity.
Opp Scores OA Hold% Succ%
ANA 392 288 13 0.232 0.957
NYN 300 230 8 0.207 0.966
PIT 314 236 14 0.204 0.944
PHI 343 263 11 0.201 0.960
TBA 332 253 13 0.199 0.951
FLO 354 271 14 0.195 0.951
TEX 356 280 7 0.194 0.976
SFN 386 304 8 0.192 0.974
MON/WAS 308 236 13 0.192 0.948
CHN 329 257 10 0.188 0.963
NYA 320 256 7 0.178 0.973
KCA 385 307 12 0.171 0.962
HOU 360 282 17 0.169 0.943
MIL 293 233 11 0.167 0.955
SDN 353 283 11 0.167 0.963
LAN 275 223 7 0.164 0.970
OAK 340 271 14 0.162 0.951
DET 361 293 10 0.161 0.967
BOS 396 316 17 0.159 0.949
BAL 359 287 15 0.159 0.950
CLE 350 281 15 0.154 0.949
MIN 335 269 15 0.152 0.947
SEA 409 332 16 0.149 0.954
COL 345 278 16 0.148 0.946
CHA 357 290 15 0.146 0.951
TOR 340 282 10 0.141 0.966
ARI 334 276 12 0.138 0.958
SLN 376 307 18 0.136 0.945
ATL 339 282 13 0.130 0.956
CIN 274 227 13 0.124 0.946
Here, you can see that there are some differences between teams. The Angels, who pride themselves on taking the extra base, held their runners fully 10% more often than did the Reds and the Braves. And while the bottom four teams in this list are from the National League, there doesn’t seem to be a strong tendency for one league or the other to either be more successful, or hold runners more frequently. The larger differences here may indeed point to personnel issues, either on the field or in the coaching box.
For individual team seasons, there were 20 teams that were never caught when sending the runner during the period, led by the 2000 Giants, who sent 66 runners successfully.
Year Team Opp Scores OA Hold% Succ% 2000 SFN 79 66 0 0.165 1.000 2002 ANA 89 64 0 0.281 1.000 2005 DET 61 52 0 0.148 1.000 2003 ARI 58 52 0 0.103 1.000 2002 MIN 61 52 0 0.148 1.000 2004 ATL 52 48 0 0.077 1.000 2000 TEX 64 47 0 0.266 1.000 2000 NYA 55 47 0 0.145 1.000 2003 NYN 53 44 0 0.170 1.000 2001 KCA 59 44 0 0.254 1.000 2005 BAL 50 42 0 0.160 1.000 2005 HOU 49 42 0 0.143 1.000 2000 SDN 50 42 0 0.160 1.000 2002 MON 50 41 0 0.180 1.000 2001 NYA 51 41 0 0.196 1.000 2003 TEX 50 40 0 0.200 1.000 2002 PIT 50 40 0 0.200 1.000 2003 FLO 47 39 0 0.170 1.000 2005 CHN 53 37 0 0.302 1.000 2004 NYN 47 34 0 0.277 1.000
Although this list features the 2000, 2002, and 2003 World Champions and the 2001 AL champs, those teams at the bottom include the 2001 World Champion Arizona Diamondbacks.
Year Team Opp Scores OA Hold% Succ% 2001 ARI 47 36 5 0.128 0.878 2003 CHA 53 39 5 0.170 0.886 2005 MIN 57 40 5 0.211 0.889 2000 CLE 68 49 6 0.191 0.891 2001 BOS 57 41 5 0.193 0.891 2001 TBA 33 25 3 0.152 0.893 2002 KCA 65 50 6 0.138 0.893 2001 PIT 51 34 4 0.255 0.895 2000 ANA 65 43 5 0.262 0.896 2001 NYN 54 35 4 0.278 0.897
In looking through I couldn’t really identify any year to year correlation in success percentage, indicating that the success rate is largely due to factors out of the team’s control or that the variety of runners on a team has the effect of evening out the percentages. The Royals which are pretty much all over the board, are a typical example.
Year Team Opp Scores OA Hold% Succ% 2000 KCA 84 68 1 0.179 0.986 2001 KCA 59 44 0 0.254 1.000 2002 KCA 65 50 6 0.138 0.893 2003 KCA 67 57 2 0.119 0.966 2004 KCA 54 38 2 0.259 0.950 2005 KCA 56 50 1 0.089 0.980
There were a few teams who were more conservative than most in sending runners, as you can see from the list below.
Year Team Opp Scores OA Hold% Succ% 2003 LAN 42 28 1 0.310 0.966 2004 CIN 36 24 1 0.306 0.960 2005 CHN 53 37 0 0.302 1.000 2000 DET 71 49 1 0.296 0.980 2002 MIL 51 34 2 0.294 0.944 2005 OAK 59 40 2 0.288 0.952 2002 ANA 89 64 0 0.281 1.000 2001 NYN 54 35 4 0.278 0.897 2005 PHI 65 46 1 0.277 0.979 2004 NYN 47 34 0 0.277 1.000
And generally, except for the 2001 Mets, that strategy paid off by getting few runners thrown out.