Using the new Batting Game Finder, we can search for games with the most pinch-hit homers since 1954.

As it turns out, there have been two games in which the teams combined for 3 pinch-hit homers:

The 1996 isn't all that interesting. The Dodgers went ahead in a slugfest, 14-4, after 5 innings. This was after Mitch Webster pinch-hit for Chris Gwynn and homered. In the bottom of the 6th, John Vander Wal hit a pinch 3-run homer to make the score 14-7 Dodgers. Then in the 7th, after two Rockies homers, Mike Kingery hit his own pinch-hit job to make the score 14-10. A few more runs got tacked on and the final score was17-11 Dodgers. None of the pinch homers were particularly meaningful.

But that game was high excitement compared to the 1999 snoozefest. The Padres were already up 12-2 when Jeff Barry pinch hit for Dante Bichette and homered to make it 12-3. In the top of the 9th, Phil Nevin and Carlos Baerga hit back-to-back pinch-hit homers, which is kind of interesting and must be a pretty rare feat.

A single team has had two pinch homers in the same game 75 times since 1954. The full list is here. It happened most recently in this game between the Cardinals and Diamondbacks, the only time in 2009 it occurred. Check out that box score, and you can see that the pinch homers made all the difference in the world. First, Mark Reynolds hit a 2-run pinch homer in the bottom of the 5th.  He batted for starting pitcher Max Scherzer, made the score 3-2 in favor of the Cardinals, and increased the Diamondbacks' chances of winning the game from 27% to 42%. Later, in the 8th inning and with 2 outs, Conor Jackson pinch hit with the score 4-3 Cardinals, two runners on, and the D-backs having just a 30% chance of winning the game. Jackson homered and put his team's chances up to 90%. The Cardinals ended up tying the game in the 9th but Arizona won it in the 10th on an Eric Brynes single. Say what you will, but those two pinch-hitting appearances were quite effective.

Just a quick hit from me today:

Oakland Rookie Pitcher Vin Mazzaro set a record this season for most PAs in a season without recording an AB. Over the course of 2 inter-league games Mazzaro came to the plate 4 times, each with a runner on first, and successfully sacrificed the runner over each time. Here are the most PAs without an AB:

Read the rest of this entry »

The new PI arithmetic function (which is my new favorite thing in the world, by the way) allows us to do lots of quick searches for things we always wanted to know.

How about this?

Here are the 130 times since 1954 in which a pitcher had more shutouts than losses in a season. This includes guys with the same number of shutouts and losses, and I did this by setting the arithmetic criteria to SHO > .999 * L (if you don't know what I'm talking about, click here to review my last post on the new arithmetic function.)

(As an aside, I am willing to bet this feat was MUCH more common before 1954 as complete games were much more common. Guys tended to pitch a lot more shutouts but not necessarily lose a lot more games.)

As you can see, many of those seasons involved a guy having 1 shutout and 1 loss and aren't that interesting. So let's limit it to seasons with at least 4 shutouts:

Here are the top 25 most recent such seasons. See what I mean about this new arithmetic tool finding hidden gems? Most of us could probably have guesses that this feat was achieved by Hershiser in 1988 or Fernando in 1981, but who had Knepper in 1980 or Reuss in 1980?

To have more shutouts than losses over a career is impossible. I checked and the most anybody since 1954 has is 1 shutout and 1 loss. However, some of the really great pitchers managed a shutout total as high as 30% of his career losses:

Note that Ford, Wilson, and Reynolds played before 1954 as well and therefore the numbers above don't actually cover their entire careers. (Karl Spooner pitched in only 1954 and 1955 so his numbers are complete.)

Data for complete games is available way back, so check out the leaders for pitchers, over their career, with at least 50% more complete games than losses:

As expected, these are just about all names from long ago, with only Spahn and Feller pitching in the latter half of the 20th century. Limiting the years to the last 50 seasons (1960-2009) gives only 3 guys with more than 2 complete games: Marichal, Koufax, and Fidrych. Again, these are guys with at least 50% more complete games than losses.

Among active pitchers, nobody comes even close to qualifying for the list. In fact, we need to drop the ratio way down and look at the only 3 active pitchers who have half as many complete games as losses:

Incidentally, the list above is minimum 37 IP. Two guys (Wade Davis and Devern Hansack) each have 1 complete game and 1 loss.

One of the interesting new features of the PI is the sortable summary provided when using the game finder. This summation provides the total statistics of the games found can be ranked by a variety of different stats. I find this very exciting because it represents the first step in searching splits. Let me explain with an example.

Suppose we wanted to find the player with the most hits on the road in a single season.

1. Use the Game Finder called Player Batting.
2. select "Find Players with Most Matching Games in a Season"
3. Limit your search to Visitor
4. Search for H>=1

That will provide you with this list:

Until now we had to stop at this point and say that "Lou Brock had the most road games with a hit in a season (since 1954)". However, the new version of PI lets us take this a step further. Assuming that it is logical that the player with the most hits on the road in a season would be among the top 300 in games with a hit, we can find that player by sorting for hits. After completing the above search, simply click on the column heading "H" and you will be presented with this list:

Brock drops down to #22 and Ichiro now tops the leader board!
Let's do the same thing for hits at home:

Clearly, this technique is limited. It can't be used for stats that usually occur in bunches (most pitching stats). It also can't be used for rate stats like batting average. However, for a lot of hitting stats it can be a fun and useful new tool.

In 2009 Dane Sardinha had a short, but freaky season. Sardinha played 12 games for the Tigers and barely hit, batting a whopping .097  in 31 AB. However, Sardinha did hit a couple of sacrifice flies. Combine that with the fact that he never walked and Sardinha posted an On Base Percentage that was actually lower than his batting average.  Sardinha's season was freaky, but hardly unique. In fact since 1901 (although for many of those years there were no SF recorded), 27 position players have had more PA in a season while posting a higher BA than OBP. Here are the top 10:

Incidentally, only one player has ever accomplished this feat after taking a walk:

This list was complied with the Baseball Reference Play Index Batting Season Finder. 
Sort by PA and search for BA>1*OBP

With the new additions to the Play Index, you can run a wider variety of queries than ever before. Here are just a few examples...

Read the rest of this entry »

One of the additions to the PI (rolled out this morning by Sean) is a neat arithmetic function that allows comparison of two stats as one of the search criteria for batting or pitching seasons. I've been hoping for something like this for a while.

Here are some of the neat things you can do with it.

First, go to the Batting Season Finder.

In the criteria selection area (the white area in the lower right), you'll see that the fifth line down contains an equation. I've highlighted it for you:

This entry allows the user to compare one stat to another stat, and optionally to add in a multiplying factor.

Here's an example. Let's take a look at guys who have driven in a lot of runs per plate appearance. In the box I've highlighted red above, I'm choosing RBI on the left and PA on the right. I'm also changing the multiplying factor from the default of 1.0 to a value of 0.2. Finally, I'm changing the sign from the default of "=" to a ">" sign. This, then, looks for seasons in which the players RBI total is greater than 20% of his plate appearance total.

I changed the years to search just in 2009 and here are the results:

I've sorted the results by plate appearance. So, again, this shows guys who batted in a run at least, on average, once every 5 plate appearances. It's not such an easy thing to do. One note is that the search uses a greater-than sign, and not a greater-than-or-equals sign. This means that guys with an RBI total that is exactly 20% of their PA total don't show up. There's an easy fix to this, and that's changing the multiplying factor from 0.2 to 0.1999. Here is the new list in that case:

This adds a bunch of players with exactly 5 times the PA as RBI, headed by new #8 Ryan Rohlinger.

I think we can all agree that many of these players are irrelevant for this particular search, so let's make two more changes. Let's limit the plate appearances to 100 or more and let's expand the search to the last 10 seasons:

Pretty neat, huh? While there are a lot of expected seasons in there (mainly big-HR total seasons such as Ortiz in 2005 and Sosa in 2001) we find a lot of hidden gems. How about Alex Ochoa in 2000 (#49)?He batted in 58 runs with just 275 plate appearances. A lot of us probably remember Tony Clark in 2005, who batted in 87 runs in just 393 plate appearances. Even more memorable is Clark's 2007, where he got 51 RBI in just 245 PAs. Once upon a time, Mike Jacobs looked pretty promising too, when he batted in 25 runs in just 112 plate appearances.

We can't yet sort the list by the arithmetic criteria, but this new feature does let us discover lots of new interesting things. I will be featuring lots of searches using the arithmetic function over the next week or so.

Baseball-Reference.com Play Index / Subscribe to the Play Index

We did a major reworking of nearly the entire website last April. One area that remained unchanged was the Play Index. It took seven months, but the Play Index has finally been updated to match the rest of the site including features like in table sorting, the Share Tool, and pre-formatted and csv text. I have also reworked a number of the tools allowing for more types searches.

The main advantage to this update besides those listed above is that these changes will speed the development of new features dramatically. I've basically been on hold for new Play Index features for 18 months as I worked to move the entire site over to our new platform. Thankfully, that is essentially done now, and we can focus on improvements and additions rather than refactoring old code.

Some of the new features.

I've added a fifth criteria for the game and season finders allowing searches like seasons with "HR > SO" or "CS > SB". This is still pretty crude, but it is a first step toward full arithmetic searches.

The season and game finders now have more matching options at the top of the forms. Most of these options were already in the tools, but were buried in the sort options. These are for searches like "Most multi-hit games in career, 1954-2009" or "Most 20-HR seasons".

For the game finders there are additional new features. When doing a matching search (like "most games with a home run in first 50 games"), you also get a summary line for all of the games matching that search, which allows you to definitively say that Wally Joyner has the most home runs in a player's first 50 games since 1954. This also gives you a way to do a reasonable good split search ("Most wins against the Blue Jays in the last ten years"). There are also more matching options for things like most players with a multiple hits in a single game for one team or both teams.

I've also re-organized the front page and now show only the last 30 days of saved searches by default.

We are also going to add a few more photos to Play Index, so look for those later this week.

Special thanks to our team of bloggers who put these updated tools through their paces. They will be highlighting some of the new things you can do in upcoming blog posts. There are definitely still bugs in there to be ironed out. There are millions of potential searches to do, so we just can't test every possible combination. You can report issues you find through our bug system. Suggestions are also welcome there. Please be as specific as you can.

Subscribe to the Play Index

To celebrate the update, we are offering a free trial for the play index through November 20th. Give it a whirl for free and we hope you'll find it worth your while to subscribe after the 20th. Also, it has been three years since we launched the Play Index and our prices have held constant during that time. We have increased the prices with this update, so that a full year subscriptions has gone from $29 to $36/year.

If something isn't working as before and you really, really can't wait for it to get fixed, the old version is still available.

Thank you for your support of Baseball-Reference.com and your feedback is always appreciated.

With all the attention that Roy Halladay is going to get this off-season (and probably into the 2010 season as well) I thought it might be interesting to guess how the rest of his career is going to turn out.

With the similarity scores recently updated to include 2009 stats, let's first look at the 10 most similar pitchers through age 32:

1. Tim Hudson (964)
2. Mike Mussina (939)
3. Bret Saberhagen (935)
4. Dizzy Dean (933) *
5. Don Newcombe (923)
6. John Candelaria (921)
7. Andy Pettitte (921)
8. Jim Bunning (914) *
9. Jimmy Key (913)
10. Dennis Leonard (911)

I found this list very surprising as I expected to see higher quality pitchers, on average. Mind you, I'm not suggesting that being similar to Mussina, Dean, or Key is a bad thing, but let's look a bit more closely.

Hudson, as one of two active players on this list, isn't much help. The point of this post to to try to guess how the rest of Halladay's career is going to go but Hudson is only one year older (and was injured most of that year) so he doesn't give us much data on the future for Halladay.

The two current HOFers on the list, Dean and Bunning, are interesting examples. Dean was essentially finished as a pitcher by age 30, pitching only 5 more innings in the majors after that. Dean had a great career but very short and also doesn't tell us anything about what to expect for Halladay in the future. Bunning is generally regarded as one of the weakest pitchers in the HOF (as is Dean, in fact, in his case based only on brevity of career.) Bunning had a few outstanding years, a handful of good years, and bunch of average years. He pitched 7 seasons after his Age 32 year. The first 3 of those (1965-1967) were all excellent, but he never pitched a full season (or a particularly good one) after that.

The fact that there are no other HOFers other than the two mentioned is, I think, a bad sign for Halladay's HOF chances. More on that below.

Mike Mussina was up and down after his Age 32 season (which came in 2001) and never pitched as many as 220 innings in any of those seasons. He averaged 189 IP over the 7 subsequent seasons.

Saberhagen was injured during his Age 32 season, joining the Red Sox the next year in 1997. Although he pitched fairly well (including a 0.8 BB/9 in 1998!) he totaled only 335 innings more in his career, spread over 4 injury-riddled seasons.

Newcombe's Age 32 season came in 1958, the year he was traded from the Dodgers to the Reds (or Redlegs as they were still officially known back then.) After that, he lasted only 2 more seasons, pitching 358 innings total.

Candelaria bounced around for 7 more seasons after age 32 (1986 with the Angels), eventually becoming a lefty relief specialist. He started only 44 games after 1986 but appeared in 206 others. His numbers were decidedly average.

Pettitte was injured in his Age 32 season and didn't pitch much. Since then, he's averaged 210 IP. His Age 33 season was superb but he's been average in his last 4. He'll probably be back with the Yankees next year but isn't likely to be much more than an average pitcher. (That being said, Pettitte hit a ton of milestone incentives this past season that boosted his salary significantly.)

Key follows a similar pattern to Pettitte, with a great season the year after his Age 32 season. In 1994 at age 33, Key led the league with 17 wins (remember that year was strike-shortened and could have cost Key his only real shot at 20 wins) and finished second in the Cy Young voting. After that, though, Key was slightly above-average pitcher who couldn't stay healthy. He pitched only 4 more years, averaging just 123 innings per season.

Finally, Leonard was injured in his Age 32 season and missed all of his Age 33 season. He came back to pitch just 194 innings more later in his career.

So, we see from this group of 10 that Halladay compares to a lot of guys who didn't pitch all that much or all that well after their Age 33 season (one year older than Halladay right now.) A few of the guys had very short careers after Age 32 and only Mussina, Pettitte, and Bunning had long careers as starters after that age.

This group is similar to Halladay, though, in that these guys were mostly very good pitchers who missed time due to injury early in their careers. That doesn't apply to everyone on the list--some of the guys compare to Halladay at Age 32 because they started their major-league careers a little later. But at this point, Halladay is a bit of a statistical oddity, being a top-tier pitcher over a long period of time despite having missed significant portions of 6 seasons (1998, 1999, 2000, 2001, 2004, and 2005.)

This sort of thing can go two ways: 1) He's prone to injury and therefore may continue to have a lot of missed time for the rest of his career or 2) He doesn't have a lot of innings on his arm despite his age and therefore, if he can stay healthy, can put up some big numbers in the rest of his career.

Injuries are impossible to predict so it's impossible to say for certain which path he'll follow. For my money, though, looking at his comparison group above, I think future injuries are more likely than future excellent seasons. Sadly, I see Saberhagen as an excellent comp for Halladay.

Halladay's injuries are also why his HOF chances are slimmed down somewhat. He is going to have a hard time reaching significant career numbers because of the missed time earlier in his career. He has only 148 career wins now, making 300 impossible unless he has a miraculous run. Even 250 wins is fairly well out of reach unless he can string together 7-8 more seasons averaging 13-14 wins. That's certainly possible but not all that likely based on the comparison group above.

Let's take a look at a different comparison group. Here are all the pitchers since 1901 to throw at least 700 innings over their Ages 30-32 seasons with a cumulative ERA+ of at least 140.

  Cnt Player            Year ERA+   IP   From  To   Ages   G   GS  CG SHO  GF  W   L   W-L%  SV   H    R   ER   BB   SO    ERA   HR   BF  HBP
+----+-----------------+----+----+------+----+----+-----+----+---+---+---+---+---+---+-----+---+----+----+----+----+----+------+---+-----+---+
    1 Roy Halladay      2009  142  710.1 2007 2009 30-32   97  96  25   7   0  53  28  .654   0  686  271  243  122  553   3.08  55  2877  20
    2 Tom Glavine       1998  151  704.2 1996 1998 30-32  102 102  10   5   0  49  23  .681   0  621  244  220  238  490   2.81  47  2898   6
    3 Greg Maddux       1998  178  728.2 1996 1998 30-32  102 102  19   8   0  52  24  .684   0  626  218  193   93  553   2.38  33  2858  16
    4 Tom Seaver        1977  141  812.2 1975 1977 30-32  104 103  47  17   0  57  26  .687   0  627  242  227  231  674   2.51  44  3225   8
    5 Bob Gibson        1968  159  760.1 1966 1968 30-32   93  93  58  20   0  56  28  .667   0  559  201  172  180  640   2.04  41  2983  15
    6 Bucky Walters     1941  149  926   1939 1941 30-32  112 107  87  10   4  68  36  .654   2  783  301  260  289  381   2.53  44  3742  13
    7 Carl Hubbell      1935  151  924.1 1933 1935 30-32  136 102  71  16  28  67  36  .650  13  856  294  247  133  424   2.40  50  3726   8
    8 Lefty Grove       1932  185  871.1 1930 1932 30-32  135  92  76  10  40  84  19  .816  21  791  286  240  201  572   2.48  31  3558   7
    9 Walter Johnson    1920  181  760   1918 1920 30-32   99  73  68  19  26  51  37  .580   8  611  212  144  148  387   1.71   7  3010  20
   10 Hippo Vaughn      1920  146  898   1918 1920 30-32  113 108  76  16   5  62  40  .608   1  781  271  202  219  420   2.02  15  3625  21
   11 Christy Mathewson 1913  160  923   1911 1913 30-32  128 106  81   9  20  74  36  .673   9  905  303  211   93  368   2.06  19  3696   3
   12 Ed Walsh          1913  144  859.1 1911 1913 30-32  134  92  72  12  39  62  38  .620  15  750  287  213  205  543   2.23  11  3470  13
   13 Mordecai Brown    1909  177  888   1907 1909 30-32  128  92  79  23  34  76  24  .760  15  640  193  137  142  402   1.39   4  3363  18

This reflects, then, the 3 most recent seasons for Halladay and it casts him in a much better light. The only guys on this list not in the Hall of Fame are Glavine, Maddux, Walters, and Vaughn. Maddux is going to be there in 4 years, and I think Glavine will get in easily too.

More importantly for Halladay, only 3 of the 13 comparables on this list failed to put up at least 3 really good seasons after the age of 32 (Vaughn, Mathewson, and Walsh.)

However, the quality of this comparison group is debatable. Halladay has just finished what was arguably the healthiest 3-year period of his career. Most of the guys on the list above consistently pitched a lot of innings. It would be foolish to assume that Halladay will continue with what he's done over the last 3 years and ignore the 6 problem years he had previously.

If I had to guess, here's what I would project for Halladay: an excellent Age 33 season coming up in 2010. All of the comparables on both lists above support that notion, plus Halladay will be in a contract year unless he signs an extension before then. After that, who knows? I'm inclined to predict some injuries and maybe only 1000-1200 IP in the rest of his career afterward. Taken together, that should be good for another 90 wins or so, putting him at 238 for his career.

We'll see...

Chain of Teammates from Bid McPhee to Elvis Andrus

Just updated the oracle with the 2009 player data. Enjoy.