This week's question comes from J.A., who asks:
Does anybody ever plug in attendance figures to the great statistical slicer-and-dicer in the sky? Specifically, what percentage of fans who attend the actual game have ever witnessed the home team win? I bet it's different from the percentage predicted by simple win-loss record.
I was trying to think of some things that affect attendance:
- overall popularity of the franchise (more people have seen Yankees home games than any other club, I bet); this would tend to increase Home Victories Witnessed (HVW)
- strength of the opponent; in non-perpetually-sold-out stadiums, this would decrease HVW
- home-team playoff contention; this would increase HVW
- day of the week; no idea whether home teams play better in front of Sunday-afternoon crowds as opposed to on Tuesday nights
I bet 60% of fans in all the ballparks over all the years have witnessed the home team win, even if only 53% of games have ended with the home team victorious.
Thanks for the question, J.A..
I love this kind of question. While it might not be deeply meaningful or provide profound sabermetric insight, it shows a novel way to think about the game. At the same time, you might discover something unexpected, and heck, it's just fun.
Let's extend J.A.'s idea in a couple of ways. First, there's his original question: what percentage of fans get to see the home team win? Though it's presumptuous to assume that all fans root for the home team, let's describe fans who see the home team win the game they attend as "satisfied fans." What factors will influence the percentage of satisfied fans?
Given that there's a bit of a home-team advantage in baseball (though not as large as in some other sports), we'd expect that slightly over half the attendees across all games were "satisfied," i.e., saw the home team win. On the other hand, if visiting teams that draw well on the road are also those that win more often on the road, this would tend to push the percentage of satisfied fans down.
Another factor that may influence the observed percentages is the time of year. Attendance tends to rise in the summer months as the weather gets better and kids get out of school. Therefore, teams that start the year on a hot streak may have a lower percentage of satisfied fans than a team with a comparable record who happened to win more during the summer months.
There's the intangible angle: are there teams that tend to play better (and win more often) in front of large home crowds than small home crowds? Do they tap into the energy that a supportive crowd cheering for them provides? If so, the synergies between large crowds and team performance should drive up the rate of satisfaction.
Enough speculation. Let's get to the numbers. To investigate this further, let's define two metrics that aren't part of our usual arsenal:
HW_ATT%: The percentage of fans who saw the home team the game they attended, aka "satisfied fans"
H_ATT/WIN: The ratio of HW_ATT% to Home Winning Percentage. Values less than 1.0 indicate that fewer fans saw the team win than you'd expect from their winning percentage. Values greater than 1.0 indicate that more fans saw the team win that you'd expect from their winning percentage.
Here are the results through games of June 24:
TM G W L H_ATT HW_ATT WIN% HW_ATT% H_ATT/WIN PHI 32 21 11 619,195 374,151 .656 .604 .921 SDN 38 14 24 1,107,747 386,158 .368 .349 .946 NYN 38 18 20 1,308,830 587,912 .474 .449 .948 ATL 34 18 16 1,142,171 584,678 .529 .512 .967 ANA 33 20 13 825,774 485,883 .606 .588 .971 ARI 38 22 16 1,228,720 691,659 .579 .563 .972 CLE 35 20 15 1,341,924 746,612 .571 .556 .974 MIN 38 25 13 844,678 541,655 .658 .641 .975 FLO 36 22 14 568,808 339,812 .611 .597 .978 SEA 37 27 10 1,509,889 1,081,961 .730 .717 .982 CHA 36 22 14 776,169 465,923 .611 .600 .982 KCA 39 16 23 811,894 327,545 .410 .403 .983 BAL 41 20 21 1,464,873 706,720 .488 .482 .989 HOU 37 18 19 1,245,726 600,976 .486 .482 .992 COL 41 22 19 1,696,205 904,666 .537 .533 .994 SFN 39 25 14 1,572,095 1,002,297 .641 .638 .995 MIL 34 20 14 1,154,771 676,693 .588 .586 .996 LAN 40 25 15 1,354,494 844,477 .625 .623 .998 CHN 35 23 12 1,136,667 745,431 .657 .656 .998 BOS 39 23 16 1,258,672 742,353 .590 .590 1.000 CIN 34 8 26 824,647 196,776 .235 .239 1.014 SLN 40 26 14 1,559,857 1,030,770 .650 .661 1.017 TBA 40 14 26 624,074 222,837 .350 .357 1.020 DET 38 19 19 808,486 412,694 .500 .510 1.021 TOR 38 18 20 790,900 385,812 .474 .488 1.030 PIT 34 15 19 1,073,286 488,711 .441 .455 1.032 NYA 36 21 15 1,382,871 839,156 .583 .607 1.040 OAK 37 18 19 943,947 479,191 .486 .508 1.043 TEX 34 13 21 1,258,664 503,646 .382 .400 1.047 MON 34 14 20 300,085 154,007 .412 .513 1.246 587 518 32,536,119 17,551,162 .531 .539 1.015
Overall, the home team wins about 53% of the time, while a slightly higher percentage of fans actually see the team win. Naturally, the percentage of satisfied fans varies from team to team, as good teams tend to win a higher percentage of their home games, and thus would have higher percentages of satisfied fans even if attendance for every game was identical.
The only real outlier in the table above are the Montreal Expos, who have treated more than half their fans to wins, despite having lost nearly 60% of their home games. However, the sabermetician's favorite excuse, "sample size" can probably be invoked here as well. With relatively small numbers of fans attending each game, a single well-attended series coinciding with a sweep can skew the totals much more easily than a team who sells out every game.
It's not too surprising that the teams on each end of the list are also two of the three teams with the lowest home attendance totals, while conversely, no team that has drawn 1.5 million people at home so far is more than 2% away from the expected value of 1.00.
For completeness, here's the counterpart to the table above–each team's road "satisfaction" figures–keeping in mind that "satisfaction" still refers to the team indicated winning, that is, the percentage of fans on the road who saw the team win (i.e., beat the home team). Maybe this is better termed "dissatisfaction" or "anti-satisfaction," since we're talking about spoiling the fans' chance to see the home team win.
TEA G W L V_ATT VW_ATT WIN% VW_ATT_% V_ATT/WIN OAK 37 17 20 993,582 371,876 .459 .374 .815 BOS 34 20 14 1,092,419 550,395 .588 .504 .857 COL 34 15 19 979,705 395,675 .441 .404 .915 MIN 35 18 17 1,019,766 484,040 .514 .475 .923 BAL 33 14 19 875,334 345,878 .424 .395 .931 CHN 38 20 18 1,326,714 651,287 .526 .491 .933 LAN 34 13 21 1,114,489 397,494 .382 .357 .933 PIT 38 10 28 1,003,414 247,752 .263 .247 .938 SDN 37 21 16 1,182,414 630,895 .568 .534 .940 DET 33 13 20 948,023 355,159 .394 .375 .951 SEA 37 28 9 1,080,639 779,563 .757 .721 .953 ANA 40 16 24 1,202,043 459,410 .400 .382 .955 CHA 36 14 22 981,359 364,906 .389 .372 .956 TBA 34 8 26 897,844 203,426 .235 .227 .963 NYN 38 15 23 1,083,080 413,077 .395 .381 .966 CLE 37 24 13 1,014,949 640,261 .649 .631 .973 SLN 35 14 21 1,201,258 470,790 .400 .392 .980 MIL 38 18 20 1,106,068 515,152 .474 .466 .983 ARI 37 24 13 1,137,336 728,398 .649 .640 .987 PHI 41 21 20 1,004,479 508,138 .512 .506 .988 HOU 36 20 16 967,214 533,804 .556 .552 .993 TEX 39 15 24 1,146,149 448,654 .385 .391 1.018 FLO 38 17 21 976,922 446,894 .447 .457 1.023 SFN 36 15 21 1,191,869 515,413 .417 .432 1.038 ATL 40 22 18 1,199,351 689,257 .550 .575 1.045 NYA 37 20 17 1,272,956 721,427 .541 .567 1.048 KCA 35 12 23 905,791 327,254 .343 .361 1.054 CIN 40 20 20 1,465,236 781,907 .500 .534 1.067 MON 42 16 26 1,244,366 508,503 .381 .409 1.073 TOR 36 18 18 921,350 498,272 .500 .541 1.082 518 587 32,536,119 14,984,957 .469 .461 0.982
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