Toronto and the One Run Fundamental

Tuesday, July 12 2016 @ 06:21 PM EDT

Contributed by: Magpie

Once more unto the breach, dear friends, once more!

Yup, I've dusted off the Big Honking Database and propose once more to look at One-Run Games in general and the hometown heroes in particular.

This one's for BlueJayWay, obviously.

Before we get to the Blue Jays, we're going to review a few theoretical and data issues. For our purposes here, we're confining ourselves to the years 1900-2015. These are seasons for which I happen to have every team's record in one-run games in handy, easy-to-manipulate form. We're looking at 2414 seasons, which seems like a reasonably large sample. It includes 376,963 games. A lot of baseball. Just how many of those games were decided by a single run? Glad you asked - it's 114,242, which is just over 30 per cent.

The crucial point to understand before looking at one-run games is this: the effect of these games is always, always, always to pull teams towards .500. One run games drag the better teams down towards .500. They lift the bad teams up towards .500. This is the One Run Fundamental. It is a Law of the Game.

Let me break those 2414 seasons into digestible chunks, that you might see this for yourselves.

Since 1900, there have been ten - that's just 10 - teams that have played .700 baseball or better over the course of a season, from the 1902 Pirates to the 2001 Mariners. Those teams played a total of 1530 games, and won 1100 of them. Their overall winning percentage was .719. They played 393 one-run games (just 25.7%), and in those games they went 256-137 - which is a .651 winning percentage. Which is really, really good. But obviously, it's nowhere near .719, and in games that were decided by more than one run, those same ten teams went 844-293. And that's a .742 winning percentage. The overall record was dragged down - and dragged down quite a bit - by their record in close games. They were .091 worse in one-run games than in the rest of their games.

It happens at the top, and it happens at the bottom. There have been twenty teams whose winning percentage couldn't even crack .300, from the 1904 Senators to the 2003 Tigers. Those teams played .276 ball (843-2214) - but they did much better in one run games (331-578, .364) than in the rest of them (512-1636, .238). They were .126 better in one run games. My next group (.300-.349) improved by .095, the next group by .075, and so on and so forth.

And that's the basic pattern. The farther away a team is from .500, the larger the pull effect of one-run games in dragging them back towards .500. Which makes sense on an intuitive level as well. A team as awful as the 2003 Tigers - if by some weird chance they were to win a game, odds are they'd just barely win it. (They actually had a winning record, 19-18, in one-run games. While going a gruesome 24-101 in their other contests.) And similarly, a team as awesome as the 1927 Yankees - if you do manage to beat them, it's probably not going to be by a lot. And Ruth's Yankees went 24-19 in one-run games. They were 86-25 the rest of the time.

Behold! A Data Table!

                                  OVERALL                             ONE-RUN GAMES                            OTHER GAMES        
                                                                            
Record    No. of Teams        GPL      W        L      PCT            GPL       W        L      PCT          GPL        W         L     PCT
                                                                            
.700 plus     10             1,530   1,100      430    .719            393      256      137    .651        1,137      844       293    .742
.650-.699     43             6,556   4,394    2,162    .670          1,931    1,158      773    .600        4,760    3,236     1,389    .700
.625-.649     68            10,576   6,723    3,853    .636          3,106    1,837    1,269    .591        7,470    4,886     2,584    .654
.600-.624    133            20,686  12,651    8,035    .612          6,195    3,488    2,707    .563       14,491    9,163     5,328    .632
.575-.599    199            30,987  18,235   12,752    .588          9,305    5,230    4,075    .562       21,682   13,005     8,677    .600
.550-.574    233            36,253  20,401   15,852    .563         10,983    5,898    5,085    .537       25,270   14,503    10,767    .574
.525-.549    325            51,352  27,572   23,780    .537         15,589    8,103    7,486    .520       35,763   19,469    16,294    .544
.500-.524    258            40,532  20,677   19,855    .510         12,445    6,252    6,193    .502       28,085   14,425    13,662    .514
.475-.499    238            37,403  18,166   19,237    .486         11,472    5,655    5,817    .493       25,931   12,511    13,420    .482
.450-.474    256            39,927  18,439   21,488    .462         12,149    5,792    6,357    .477       27,778   12,647    15,131    .455
.425-.449    202            31,600  13,823   17,777    .437          9,746    4,529    5,217    .465       21,854    9,294    12,560    .425
.400-.424    173            27,129  11,166   15,963    .412          8,157    3,687    4,470    .452       18,972    7,479    11,493    .394
.375-.399    113            17,554   6,800   10,754    .387          5,320    2,253    3,067    .423       12,234    4,547     7,687    .372
.350-.374     62             9,520   3,446    6,074    .362          2,828    1,172    1,656    .414        6,692    2,274     4,418    .340
.300-.349     81            12,500   4,145    8,355    .332          3,712    1,479    2,233    .398        8,788    2,666     6,122    .303
.000-.299     20             3,057     843    2,214    .276            909      331      578    .364        2,148      512     1,636    .238
                                                                            
            2414           377,162 188,581  188,581    .500        114,240   57,120   57,120    .500      263,055  131,461   131,461    0.500

As you can see, the pattern is pretty dependable. The only group that doesn't fit perfectly includes the 133 teams who played between .600 and .624 ball. Those teams had results in one-run games that were just a little bit worse than all this would lead you to expect. They played 6195 one run games, and I would have expected them to win about 3560 of them (roughly .575 ball) - they actually went 3488-2707, which is just .563 ball- in other words, it's exactly the same as the difference between going 93-69 and going 91-71.

This, by the way, is why all the game's greatest and most successful managers have worse records in one-run games than they do the rest of the time. Having a worse record in one-run games than you do in the rest of your games is actually a characteristic of any group of good teams, and those are the types of team that were managed by McCarthy and McGraw and Weaver.

Anyway, that's the pattern, and let me run off a small tangent for the rest of this paragraph. Here's my question - how many games do we need before the sample size becomes large enough for the One Run Fundamental to show itself. It's true that all of these samples behave roughly as you would expect them to. The smallest sample here consists of 10 seasons, 1530 overall games, 393 one-run games. But I still suspect that even that number of games is just a little on the small size.

I do think we can be pretty damn sure that a single season, and the 30 or 40 one-run games it might include, is nowhere near a large enough sample for the One-Run Fundamental to assert itself - it's about the same as... oh, 30 or 40 at bats over the course of the year. In which anything can happen. It's a little like Josh Donaldson going 5-35 back in May. He's still Josh Donaldson.

So what's the biggest factor in one-run games?

Well, obviously the quality of the team is a pretty large factor. The best teams really do have the best records in one-run games.

But the closer a game gets, the greater the chance that Random Circumstance can determine the outcome. And Luck doesn't know anything about quality. Luck is a flip of a coin, and the odds are the same for everyone. Heads or tails, win or lose, it's 50-50.

Let's look at the teams that played .700 ball or better. They played .651 ball in one-run games. In an average season, they'd have played about 42 such games, so it's as if they went 27-15 in those games. But we would have expected them to go 31-11, had they played one-run games the same way they played the rest of them. It's almost as if half the games were decided by the team's quality (their .742 winning percentage) and the other half were decided by a coin flip (which ought to be .500).

Half of the one-run games are effectively decided by a coin flip?

Gulp. Because if you flip that coin 3000 times, you'll surely see heads come up 1500 times. But if you only flip it 30 or 40 times? Weird, weird things can happen.

Which brings us, at last to the Toronto Blue Jays.

From that snowy April day in 1977 through this year's All-Star Break, the Jays have played 6,284 games and they've had a losing record for almost all of that time. (They were briefly above .500 from September 1993 through May 1994, and intemittently after that for the next year. But the franchise has had a losing record every day since May 1995. Not to mention the sixteen years before the second championship.) At this particular moment, they got 3129 wins and 3155 losses, a .498 percentage. We would expect them to have a record in one-run games that's very close to that. Maybe a game or two better.

They don't. They're 874-926 (.486) in one-run games. It's not that big a deal - one more loss, one fewer win every other year or so. Here are the raw numbers:

        ONE-RUN                  OVERALL                    OTHER GAMES

Year    W      L    PCT          W     L     PCT        W     L     PCT         Manager(s)
                                                            
1977    17    27    .386        54    107   .335        37    80    .316        Hartsfield
1978    23    30    .434        59    102   .366        36    72    .333        Hartsfield
1979    19    28    .404        53    109   .327        34    81    .296        Hartsfield
1980    23    21    .523        67    95    .414        44    74    .373        Mattick
1981    10    17    .370        37    69    .349        27    52    .342        Mattick
1982    28    30    .483        78    84    .481        50    54    .481        Cox
1983    25    20    .556        89    73    .549        64    53    .547        Cox
1984    34    25    .576        89    73    .549        55    48    .534        Cox
1985    26    21    .553        99    62    .615        73    41    .640        Cox
1986    22    25    .468        86    76    .531        64    51    .557        Williams
1987    27    24    .529        96    66    .593        69    42    .622        Williams
1988    21    17    .553        87    75    .537        66    58    .532        Williams
1989    25    22    .532        89    73    .549        64    51    .557        Williams-Gaston
1990    24    27    .471        86    76    .531        62    49    .559        Gaston
1991    28    20    .583        91    71    .562        63    51    .553        Gaston-Tenace-Gaston
1992    28    20    .583        96    66    .593        68    46    .596        Gaston
1993    23    22    .511        95    67    .586        72    45    .615        Gaston
1994    13    15    .464        55    60    .478        42    45    .483        Gaston
1995    16    23    .410        56    88    .389        40    65    .381        Gaston
1996    19    22    .463        74    88    .457        55    66    .455        Gaston
1997    29    30    .492        76    86    .469        47    56    .456        Gaston-Queen
1998    28    17    .622        88    74    .543        60    57    .513        Johnson
1999    26    18    .591        84    78    .519        58    60    .492        Fregosi
2000    21    19    .525        83    79    .512        62    60    .508        Fregosi
2001    28    21    .571        80    82    .494        52    61    .460        Martinez
2002    23    21    .523        78    84    .481        55    63    .466        Martinez-Tosca
2003    14    23    .378        86    76    .531        72    53    .576        Tosca
2004    17    22    .436        67    94    .416        50    72    .410        Tosca-Gibbons
2005    16    31    .340        80    82    .494        64    51    .557        Gibbons
2006    20    10    .667        87    75    .537        67    65    .508        Gibbons
2007    29    25    .537        83    79    .512        54    54    .500        Gibbons
2008    24    32    .429        86    76    .531        62    44    .585        Gibbons-Gaston
2009    21    28    .429        75    87    .463        54    59    .478        Gaston
2010    24    28    .462        85    77    .525        61    49    .555        Gaston
2011    29    28    .509        81    81    .500        52    53    .495        Farrell
2012    15    25    .375        73    89    .451        58    64    .475        Farrell
2013    20    29    .408        74    88    .457        54    59    .478        Gibbons
2014    15    20    .429        83    79    .512        68    59    .535        Gibbons
2015    15    28    .349        93    69    .574        78    41    .655        Gibbons
2016     9    15    .375        51    40    .560        42    25    .627        Gibbons
                                                            
       874   926    .486      3129  3155    .498      2255  2229    .503

So what was the most disturbing experience, when did the team lose way more one-run games than you would expect from a team of that general quality? That would be the 2015 team, who played .349 ball (15-28) in one-run games and .655 ball (78-41) the rest of the time. (Using my Coin Flip logic, they went roughly 15-6 on Quality, but the coin came up Tails 20 times in a row when they were calling Heads. Yikes.

The next worse? Meet this year's team, playing .375 ball (9-15) in one-run games, and .627 ball (42-25 the rest of the time.

Next worse? The 2005 squad, who played the one-run games at a .340 clip (16-31), while playing .557 ball.

What do these teams have in common? Well, there's ownership. And batting practice pitcher Jesus Figueroa. And... well, let's look at the managers, shall we?

This requires a bit of digging through the Game Logs, as the Jays made mid-season managerial changes in 1989, 1997, 2002, 2004, and 2008. And while Cito Gaston was the manager of the 1991 team that went 91-71, Gaston went into Mt Sinai Hospital with a herniated disk on August 21 of that year and was away from the ballpark for just over a month. Gene Tenace managed the games in his absence.

This time, instead of running through the data in chronological order, we'll rank the teams by their proficiency in one-run games. And by proficiency I don't mean mere winning percentage - I mean winning percentage in one-run games compared to winning percentage of the other games. Which Toronto team had the most good fortune in the close ones?


          ONE-RUN                  OVERALL              OTHER GAMES

Year    W   L    PCT            W     L   PCT           W       L     PCT      DIFF   Manager(s)								
															
1997     2    0	 1.000		4     1	 .800	 	 2	 1   .667	.333		Queen
2006    20   10   .667		87   75	 .537		67	65   .508	.159		Gibbons
1980    23   21   .523		67   95	 .414		44	74   .373	.150		Mattick
1989    23   10	  .697		77   49	 .611		54	39   .581	.116		Gaston
2001    28   21	  .571		80   82	 .494		52	61   .460	.111		Martinez
1998    28   17	  .622		88   74	 .543		60	57   .513	.109		Johnson
1979    19   28	  .404		53  109	 .327		34	81   .296	.109		Hartsfield
1991    24   14	  .632		72   57	 .558		48	43   .527	.104		Gaston
1978    23   30	  .434		59  102	 .366		36	72   .333	.101		Hartsfield
1999    26   18	  .591		84   78	 .519		58	60   .492	.099		Fregosi
2002     5    6	  .455		20   33	 .377		15	27   .357	.097		Martinez
2004    13   14	  .481		47   64	 .423		34	50   .405	.077		Tosca
1977    17   27	  .386		54  107	 .335		37	80   .316	.070		Hartsfield
1984    34   25	  .576		89   73	 .549		55	48   .534	.042		Cox
2007    29   25	  .537		83   79	 .512		54	54   .500	.037		Gibbons
1995    16   23   .410		56   88	 .389		40	65   .381	.029		Gaston
1981    10   17   .370		37   69	 .349		27	52   .342	.029		Mattick
1997    27   30	  .474		72   85	 .459		45	55   .450	.024		Gaston
1988    21   17	  .553		87   75	 .537		66	58   .532	.020		Williams
2002    18   15	  .545		58   51	 .532		40	36   .526	.019		Tosca
2000    21   19	  .525		83   79	 .512		62	60   .508	.017		Fregosi
2011    29   28	  .509		81   81	 .500		52	53   .495	.014		Farrell
1996    19   22	  .463		74   88	 .457		55	66   .455	.009		Gaston
1983    25   20	  .556		89   73	 .549		64	53   .547	.009		Cox
1982    28   30	  .483		78   84	 .481		50	54   .481	.002		Cox
1992    28   20	  .583		96   66	 .593		68	46   .596      -.013		Gaston
1994    13   15	  .464		55   60	 .478		42	45   .483      -.018		Gaston
2009    21   28	  .429		75   87	 .463		54	59   .478      -.049		Gaston
2013    20   29	  .408		74   88	 .457		54	59   .478      -.070		Gibbons
1985    26   21	  .553		99   62	 .615		73	41   .640      -.087		Cox
2004     4    8	  .333		20   30	 .400		16	22   .421      -.088		Gibbons
1990    24   27   .471		86   76	 .531		62	49   .559      -.088		Gaston
1986    22   25	  .468		86   76	 .531		64	51   .557      -.088		Williams
1987    27   24	  .529		96   66	 .593		69	42   .622      -.092		Williams
2010    24   28	  .462		85   77	 .525		61	49   .555      -.093		Gaston
2012    15   25	  .375		73   89	 .451		58	64   .475      -.100		Farrell
1993    23   22	  .511		95   67	 .586		72	45   .615      -.104		Gaston
2014    15   20	  .429		83   79	 .512		68	59   .535      -.107		Gibbons
2008    11   17	  .393		35   39	 .473		24	22   .522      -.129		Gibbons
2008    13   15	  .464		51   37	 .580		38	22   .633      -.169		Gaston
2003    14   23	  .378		86   76	 .531		72	53   .576      -.198		Tosca
2005    16   31	  .340		80   82	 .494		64	51   .557      -.216		Gibbons
2016     9   15	  .375		51   40	 .560		42	25   .627      -.252		Gibbons
1991     4    6	  .400		19   14	 .576		15	8    .652      -.252		Tenace
2015    15   28	  .349		93   69	 .574		78	41   .655      -.307		Gibbons
1989     2   12	  .143		12   24	 .333		10	12   .455      -.312		Williams

We're going to pass over Mel Queen's five game stint at the end of 1997, and try to remember the good fortune enjoyed by John Gibbons's 2006 team. They played .667 ball (20-10) in the one-run games, while being little more than a .500 team (67-65, .508) the rest of the time? Go figure. Cito Gaston's 1989 team was even better in the one-run games (23-10, .697) but that team was quite a bit better than the 2006 squad - they played .581 ball (54-39) for Gaston the rest of the time. And the most unfortunate manager of all was not Gibbons, although heaven knows his name does show up often enough as the bottom of this table. It was Jimy Williams in early 1989. The 12-24 start that cost him his job was built on a horrific 2-12 mark in the one-run games. Something I hadn't known!

And finally, let's just add up the manager's records and see how they've done in one-run games.

	                 ONE RUN GAMES	         OTHER GAMES		OVERALL				
Manager	                  W     L     PCT         W    L    PCT	        W     L	   PCT		DIFF
													
Hartsfield               59    85    .410       107   233   .315       166   318   .343	  .095
Mattick	                 33    38    .465        71   126   .360       104   164   .388	  .104
Cox	                113    96    .541       242   196   .553       355   292   .549	 -.012
Williams                 72    78    .480       209   163   .562       281   241   .538	 -.082
Gaston I (1989-2007)	197   183    .518       486   453   .518       683   636   .518	  .001
Tenace	                  4     6    .400        15     8   .652        19    14   .576	 -.252
Queen	                  2     0   1.000         2     1   .667         4     1   .800	  .333
Johnson	                 28    17    .622        60    57   .513        88    74   .543	  .109
Fregosi	                 47    37    .560       120   120   .500       167   157   .515	  .060
Martinez	         33    27    .550        67    88   .432       100   115   .465	  .118
Tosca	                 45    52    .464       146   139   .512       191   191   .500	 -.048
Gibbons I (2004-2008)	 80    91    .468       225   214   .513       305   305   .500	 -.045
Gaston II (2008-2010)	 58    71    .450       153   130   .541       211   201   .512	 -.091
Farrell	                 44    53    .454       110   117   .485       154   170   .475	 -.031
Gibbons II (2013-2016)	 59    92    .391       242   184   .568       301   276   .522	 -.177
													
Gaston TOTAL	        255   254    .501       639   583   .523       894   837   .516	 -.022
Gibbons TOTAL	        139   183    .432       467   398   .540       606   581   .511	 -.108

That final column just compares the manager's winning percentage in one-run games to his percentage in the other games.

In my view, all of these represent Small Samples. A few tidbits, nonetheless...

The terrible teams managed by Hartsfield and Mattick both posted significantly better results in one-run games, exactly as you would expect them to do. Mattick's teams helped themselves a little more than you might expect.

Bobby Cox's excellent teams didn't lose much to the coin flip at all...

Whereas Jimy Williams' teams were pretty bad in close games. His teams were even better than Cox's teams in the other games, but everything went to hell in one-run games.

Gaston's first tour is remarkable - the exact same winning percentage (.518) in all kinds of games..

Jim Fregosi, Tim Johnson, and Buck Martinez all did quite well in one-run games, in their brief times in charge.

But since Martinez was fired, everything has gone to hell. All five managers (counting Gibbons twice) have seen significantly worse results in one-run games than you would expect. Tosca, Gibbons Mark I, and Farrell were all disappointing. Gaston Mark II was awful, twice as bad as that Group of Three. And Gibbons Mark II - great gosh almighty, he's been twice as bad as that, some twenty-seven shades of godawful.

Something is clearly wrong with the coins around here. These last ten years or so have been weird. Really, really weird.

Anyway - BlueJayWay.... we feel your pain.

40 comments

https://www.battersbox.ca/article.php?story=20160712142135301