All things being equal, increased run-scoring would increase the time of the game. All things being equal, increased basestealing would also increase the time of the game. I don't know the extent to which the decreased basestealing in today's game is offset by the increased scoring. I do know we shouldn't pretend that 15-13 games are in any way commonplace.

Increased run-scoring by itself doesn't appreciably increase the time of the game. What increases the time of the game are (a) pitching changes and (b) men on base, especially but not limited to basestealers. Obviously those correlate with increased scoring -- but a game with a bunch of solo home runs and no other baserunners will be much shorter than a game with several rallies leading to those same number of runs.

Yankee-Red Sox Game Lengths, 2009 (9 innings, unless noted):

4:21 (11 innings, 5-4)
4:21
3:09
3:48
3:47
3:04
3:23
3:39
5:33 (15 innings, 2-0)
3:28
3:08
3:57
3:17
3:13
3:52
3:07
2:56.

One game of 19 under three hours; a game with no runs until the 15th inning played at a 9-inning equivalent pace of 3:19. The average eyeballed looks to exceed 3:30.

Having watched most of these games, it didn't bother me at all that they were generally long; actually, it kind of added to the drama. Most of these games were a lot of fun. Both teams had good offenses and took a lot of pitches, both managers ran the games as if they were playoff games and therefore (at least from observation) tended to change pitchers more than usual. The fact that the teams were evenly matched and a lot was at stake meant closer games and more pitching changes and more trips to the mound by managers and catchers and by Derek Jeter performing his Captain duties which I guess consisted of distracting his pitcher so that Jeter could perpetuate his image as a Team Leader.

Not sure how much of a factor time between pitches was, whether it was 12 seconds or 18. But there were a lot of other factors involved nevertheless. Hell, when Pedro pitched, the Yankees' main strategy was to force him to throw a lot of pitches so he'd have to leave the game earlier.

If they were all Yankees-Red Sox games, it might not bother me as much, either, since they're nearly always the dramatic equivalent of playoff games. Unfortunately that still leaves us another 2000+ games per season to contend with.

I am in the minority that cares more about overall length than pace. The pace does irritate me but I mostly watch at home on TV so I can pause and FF.

I work next to Rogers Centre (looking at it now) and, due to my life/work schedule, do not go to many games. This has more to do with time than money or anything else. If I could go to a game at 7:00 and be reasonably sure it would be done between 9:00 and 9:30 I would go to a lot more. I have no idea how many more people there are like me and whether or not that is worth anything to MLB but it is how I feel.

When my son is older and can come down to see a game with me (and I am at the point in my career where I am winding things down and don't mind coming in tired) I would probably enjoy the game being longer so I can soak it all in.

So I recognize my displeasure with the time is due to my own specific circumstances and that as they change my tastes change.

No game that the Yankees lose is ever too long.

Baseball is fine. Don't #### with it.

Baseball is great, but I don't know how anyone who saw some of the umpiring in the playoffs last year could think that it can't possibly be improved.

Just so we're all on the same page, the "OMG TEH PACE!" guys would be happy with a 1970s game where the TOG was 2:30-35? Today's game is what? 2:45? 2:52?

Without linear regressions, this thread is useless.

Srsy, doesn't retrosheet have time of game?

T = a*1b + b*2b + c*3b + d*hr + e*bb + f*k + g*sb + h*cs (or other out on base) + y*(N pitching changes) + z(N pinch hitters) + CCT (constant commerical time).

per IP.

I have timed the time between pitches and it is usually between 20-30 seconds. The situation has become so bad that you can almost use the skip ahead button and miss no pitches. Most pitchers, especially relievers, are closer to 30 seconds between pitches than 12 seconds.

Not only do we need to enforce the 12 second rule with no one on base. A rule needs to be invented for when there are players on base. In the late innings of an important game your average reliever will walk around the ####### mound after every pitch. In the playoffs I bet the average is a pitch every 30 or 35 seconds. That is unconscionable.

I like Gaelen's rule: limit of 30 seconds between pitches when a runner is on. Yes, if the pitcher gets close to 30 seconds, the runner can get a huge jump. All the more reason for the pitcher to throw the damned ball. Tell the umpires if >50% of the pitches thrown while they're behind the plate over the season exceed the limit, they'll be fired. And then do it.


I love baseball, even as it is - bad umpiring and all. But I'd watch more games if the pace was picked up a bit and/or the overall length shortened a bit.

I found those Yankee/Red Sox regular season games annoying as all hell because a) they were played at a very slow pace; b) they were treated as if they were playoff games. To me, b is annoying if it is May. In fairness, it may only have been that I watched on TV and the games were massively overproduced.

I like Gaelen's rule: limit of 30 seconds between pitches when a runner is on. Yes, if the pitcher gets close to 30 seconds, the runner can get a huge jump. All the more reason for the pitcher to throw the damned ball.

I think you guys are focusing on the pitchers too much. The batters share a role in this also, since almost every one of them takes a Sunday stroll around the field inbetween pitches.

Also, it does take _some_ time for the pitcher and catcher to get on the same page (particularly if someone is calling the pitches from the dugout, but the catcher sometimes has to look over into the dugout anyway), and it does take _some_ time for the batter to get the signs. Do these things need more than 12 seconds, typically? No, probably not.

They don't need another time limit rule for ROB situations. They just need to stop the freaking strolls around the park. I seriously doubt that any pitcher takes more than 12 seconds from the time that they actually toe the rubber (without ROB, that is).

#209, then we could throw in a some kind of a year variable (year - 1952 or whatever) and see if there's any kind of signal.

Yeah, I wouldn't change the rules with runners on base -- there would need to be too many exceptions (for holding a tricky runner, changing signs, etc.). I would probably instruct the umpire to grant time less in such situations, maybe even to the point of denying mound conferences if they became too frequent (particularly if there is no runner on second to relay signs). No clock with runners on, though.

The 12-second rule without men on base is the one that needs to be enforced. If the pitcher doesn't throw, it's a ball, or if the batter won't get in the box, it's a strike.... something like that.

Also Chris I honestly don't care about the time of the game. I've always hated the ritualistic getting ready whether it's Mike Hargrove or some guy kicking a penalty in rugby.

Mike Hargrove was pretty much unique in his day. I think anybody can name several guys as bad as him these days. Don't know what this adds to game time and I don't care. If it's as much as a minute I'd be surprised.

Similarly we all know guys who take forever to actually get around to throwing the pitch. That forever is maybe 5 seconds. Too bad, it annoys me.

I don't particularly mind pitching changes -- though I'd be totally fine with rules that restrict them to some extent (To those who worry about faked injuries, no problem. A player lifted when not normally permitted goes on the DL. Period)

[209]

Without linear regressions, this thread is useless.

Srsy, doesn't retrosheet have time of game?

T = a*1b + b*2b + c*3b + d*hr + e*bb + f*k + g*sb + h*cs (or other out on base) + y*(N pitching changes) + z(N pinch hitters) + CCT (constant commerical time).

per IP.

I think that games are probably taking longer due to the changing mix of plate appearances, in addition to a greater number of PA per game and the batter/pitcher procrastination. The impact of the procrastination is exacerbated by the shift in the mix of plate appearances towards walks and strikeouts (and away from contact).

I used retrosheet gamelog for 2009. It doesn't have pinch hitting, but otherwise a good dataset for this kind of analysis.

Dependent variable: Time of game, in minutes
Independent variables: E, 1B, 2B, 3B, HR, SH, SF, HBP, UBB (Unintentional BB), K, SB, CS, GIDP, IBB, RP, Other Outs (Outs - SH - SF - K - CS -2*GIDP)

Result: R-Squared and Adjusted R-Squared of 0.82 and all variables are significant at a 95% confidence level. This I read as a good explanatory model without a lot of extraneous variables.

Coefficients (in minutes, sorted)

Intercept  3.77
RP         3.23
UBB        2.86
IBB        2.78
K          2.43
HBP        2.37
SH         1.92
CS         1.80
Other Outs 1.76
GIDP       1.68
SF         1.61
SB         1.59
3B         1.58
1B         1.57
2B         1.57
HR         1.35
E          1.35


I would say that the coefficients don't really match the duration of each event, because the model blends the commercial breaks into each of the game events. (Since there is little variation in the number of commercial breaks from game to game, trying to model them explicitly just blows up the model.)

The high-coefficient, i.e. time intensive, events are the events which are on the rise: walks, strikeouts, and use of relief pitchers. No surprise there.
Other outs are less time intensive, followed by other ways of getting on base (hits of all types and errors). HR score as slightly less time intensive than other types of hits.

EDITED for format

god I love the internet.

I guess the question is:
1) do the coefficients change year-to-year?
2) in this model, does the increase in RP/BB/K account for the increase in minutes/game.

Actually, these should be the same question.

[218] I'll run another regression analysis on some earlier years, depending on the completeness of the annual gamelogs.

Even if the coefficients are stable, the increase in game may be accounted for by the increase in offense, use of RP and the decrease in pitching to contact, i.e., more K and BB per PA (or batter faced).

Adjusted R-squared for single-year regressions:

2009: .82
1999: .78
1989: .77
1979: .66
1969: .80


Regression coefficients (all but 1979 HBP are significant at the 95% confidence level)
<code>
Coefficients 2009  1999  1989  1979   1969 2009 vs 1969
Intercept    3.77 (0.79) 2.88 (7.02) (0.09)
RP           3.23  3.04  2.89  2.77   2.98      9%
UBB          2.86  2.80  2.68  2.20   2.36     21%
IBB          2.78  2.42  1.76  2.50   2.33     19%
K            2.43  2.53  2.53  2.62   2.46     -2%
HBP          2.37  2.01  2.00  0.57   1.15    107%
SH           1.92  1.66  3.06  2.44   2.16    -11%
CS           1.80  2.06  2.63  2.10   1.75      3%
Other Outs   1.76  1.97  1.93  1.95   1.74      1%
GIDP         1.68  2.40  2.96  2.57   2.17    -23%
SF           1.61  2.12  2.23  2.94   1.93    -16%
SB           1.59  1.35  1.37  1.60   1.20     33%
3B           1.58  1.03  0.80  1.26   1.00     58%
1B           1.57  1.62  1.61  1.36   1.51      4%
2B           1.57  1.67  1.23  1.66   0.94     67%
HR           1.35  0.48  0.64  1.01   1.04     30%
E            1.35  0.78  0.93  0.75   1.06     28%

Regression coefficients, sorted by greatest % change from 1969 to 2009

HBP      107%
2B        67%
3B        58%
SB        33%
HR        30%
E         28%
UBB       21%
IBB       19%
RP         9%
1B         4%
CS         3%
Other Outs 1%
K         -2%
SH       -11%
SF       -16%
GIDP     -23%


Count of events, per game for each season, and change from 1969 to 2009

<u>Year        2009 1999 1989 1979 1969 09 vs '69</u>
RP          5.86 5.12 3.75 3.04 3.20  83%
2B          3.60 3.60 2.99 3.06 2.49  45%
HBP         0.65 0.65 0.38 0.36 0.45  44%
HR          2.07 2.28 1.46 1.64 1.60  29%
SB          1.22 1.41 1.48 1.42 0.95  29%
K          13.8 12.8 11.2  9;.6 11.6  20%
SF          0.56 0.60 0.59 0.63 0.47  20%
GIDP        1.56 1.58 1.45 1.59 1.50   4%
UBB         6.35 6.91 5.74 5.83 6.17   3%
1B         11.9 12.4 12.4 12.9 12.2   -3%
Other Outs 34.8 35.5 37.5 38.6 37.3   -7%
3B          0.39 0.38 0.41 0.51 0.44 -10%
CS          0.47 0.63 0.68 0.76 0.58 -19%
SH          0.67 0.66 0.77 0.90 0.86 -22%
E           1.17 1.44 1.59 1.72 1.76 -33%
<u>IBB         0.49 0.46 0.69 0.65 0.74 -34%</u>
Time (min) 175  177  170  154  152    15%