This is our old blog. It hasn't been active since 2011. Please see the link above for our current blog or click the logo above to see all of the great data and content on this site.

The average number of pitches thrown per game is rising

Posted by Andy on July 23, 2010

These days, each team throws an average of 146 pitches per game, up about 11 pitches per game from 20 years ago. Click through for data and discussion.

This simple plot shows the average number of pitches thrown per team per game each year since 1988.

Complete pitch data goes back to just 1988 so that's why this graph doesn't go further back.

In 1988, teams averaged 136.2 pitches per game. In 2009, it was 147.4, an increase of 8% over the 21-year period.

I generated this data by multiplying two numbers together--the average pitches per plate appearance and the average number of plate appearances per game. (The links are to the B-R pages where those data can be found.)

The average pitchers per PA has gradually increased since 1988. (I showed some of the data in this post and will show the rest of it next week.) In 1988 it was 3.59 pitches/PA and in 2009 it was 3.83 pitches/PA. That's a 6.7% increase itself.

The number of plate appearances per game generally follows overall offensive levels in MLB. That makes a lot of sense if you think about it--every time a guy scores a run, that guarantees another plate appearance in the game. While he's up at bat or on base, there is the potential that he will become an out, but once he scores that out will have to come against a future batter or baserunner. If you click on the link I provided above for average PAs per game, you can read down the column and note how it's changed. Just look at 1968 vs 1969. PAs per game had plummeted through the 1960s all the way down to 37.17 in 1968. Then the mound was dropped for the 1969 season, offense went up, and plate appearances rocketed up to 38.07 per game.

Some random thoughts about the number of pitches per game increasing:

• This affects the length of games, of course. With 8% more pitches, games will take at least 8% longer. Factor in the fact that more pitchers are used per game and the additional pitching changes mean that games are longer by more than 8%.
• In 2009, starting pitchers averaged 95 pitchers per game. In 1988, it was 96 per game.
• Relievers threw 40 pitches per game in 1988 and 52 pitches per game in 2009. (I got those numbers through simple subtraction of the other numbers in this post.) That's a 30% INCREASE.
• Using the PI Pitching Game Finder, I discovered that there were 7,331 relief appearances in 1988 and 14,239 relief appearances in 2009. Normalizing by the number of games in each season (6th column here) that works out to 1.75 relief appearances per game in 1988 and 2.93 relief appearances in 2009. So that's a 67% increase in relief appearances.
• Therefore, relief pitchers are actually throwing fewer pitches per appearance these days than they used to, even though relievers in total are throwing more pitches. (Simply put, there are more relievers these days but they throw fewer pitches per appearance--keep in mind this doesn't necessarily mean fewer pitches per plate appearance.)

It all smacks of a weird sort of inefficient efficiency. Sure, having pitchers throw fewer pitches per appearance more protects their arms, but that also means that you have more guys warming up and guys making more frequent appearances. In the end, I'm not sure pitchers actually have a lighter workload.

32 Responses to “The average number of pitches thrown per game is rising”

1. Johnny Twisto Says:

Brief caveat: I'm pretty sure that even since 1988 there are a few games without pitch data, or even worse, incomplete pitch data. I know searches have been run which have screwy results like a CG with 18 pitches, because for some reason the pitch data only exists for the 5th inning or something. I'm not sure if the number of games like this are enough to meaningfully affect the results.

2. Johnny Twisto Says:

In 2009, starting pitchers averaged 95 pitchers per game. In 1988, it was 96 per game.

I saw a great graph in the last few months exhibiting this. The average # of pitches per start has stayed fairly steady for a long time, but the variance has been decreasing rapidly. There used to be a lot 130-pitch starts, 150-pitch starts, but also some of 50 pitches or so. Now it's rare to top 120, but even if the starter is struggling, the manager will try to get him through a few innings, at least 80 pitches or so, to save the bullpen.

3. Chris Says:

In 2009, starting pitchers averaged 95 pitchers per game. In 1988, it was 96 per game.

With the overemphasis on pitch counts nowadays, I was surprised the difference wasn't greater.

4. Bryan Mueller Says:

Not that it would have too big an effect on pitches thrown per game, but the lack of "small ball" may increase pitch count just a little. Also, it would be interesting to see how many BBs and Ks there were in an avereage game in 1988 compared to 2009.

5. Charles Saeger Says:

There are fewer long and fewer short outings.

6. redsock Says:

In 2009, starting pitchers averaged 95 pitchers per game. In 1988, it was 96 per game.

Here is the proof: today's starters are wimps.

7. Mike Gaber Says:

Your Link for the Average number of plate appearances per game list has a big error in 1883 when it tries to compute the Average age for the 258 batters.

http://www.baseball-reference.com/leagues/MLB/bat.shtml

It shows 256589663296492.6

8. dukeofflatbush Says:

Also, the reemergence of the work horse is back. Guys not afraid to go deep into games multiple times.
CC two years ago on the BrewCrew, was a throw back, never mind his pitch count, but he was doing it on 3 days rest and going 8-9 IP.
This year after 16 starts, Cliff Lee is averaging 8.125 innings pitched. Not 8 1/3, but the true decimal unit based on tenths. I think the last guy to do so was Maddux in '94. 25 GS, 200+ IP. Lee has to pitch just a little over 7.67 starts over his next 9 starts to equal Maddux's '94 mark.

9. SJBlonger Says:

All it takes it one 256 trillion-year-old player to throw off the average.

I would love to find out how that happened.

10. Andy Says:

No, no, players just used to be a lot older back then. For just that one year. ðŸ™‚

11. Hartvig Says:

Geez, if Jamie Moyer played in 1883 he would actually lower the average players age by a little bit. Too funny.

My guess is that now steroids are pretty much cleaned up (at least for the moment) that pitch counts will drop a bit. Power numbers should go down & pitchers should no longer have to worry about someone like Brady Anderson going deep 50 times a year.

12. statboy Says:

How come it's about 137.5 (38.08 PAs/game) in 1990, but 147.5 (38.49 PAs/game) in 2009? If PAs/game only went up by .41, why does that equal 10 more pitches for the game?

FWIW, hardballtimes came up with pretty different numbers...

http://www.hardballtimes.com/main/article/what-pitch-counts-hath-wrought/
Estimated Pitches per MLB Team/Game:

2003 146
2000 149
1995 148
1990 144
1985 144
1980 143
1975 144
1970 145
1965 142
1960 144
1955 144
1950 146

13. jenifer dunn Says:

All it takes it one 256 trillion-year-old player to throw off the average.

I would love to find out how that happened.
thanks,
jenifer dunn
public figure biography

14. Richard Says:

@6 hahaha

@8 Your two examples are guys pitching their asses off for new contracts though. Just saying.

15. Richard Says:

Also, @ 12 I'd say this is proof that the hardball times estimates are incorrect. These are ACTUAL pitch counts; they had a formula for just estimating what the pitch count probably was based on walks, strikeouts, etc

16. Andy Says:

statboy, because the pitches per plate appearance went up to, by a larger fraction.

17. Andy Says:

I'm pretty sure that the data in this post is correct because I double-checked it by using a second method.

On this page you can get the total number of pitches thrown in a season. In 2009, it was 715,793. That came in 4860 games, as you can see here. Divide those two numbers and you get 147.28. I checked this for a number of the years in the data above and they all matched up.

18. Dan Franzen Says:

Isn't the higher overall pitch count due mostly to the increased propensity of managers to have a guy come in and face only a couple of batters? So instead of having two or three pitchers in a game, you now have four or five, each fresh when he comes in and perhaps more likely to throw more pitches than if the starter had remained in. Or maybe I'm just making stuff up.

19. Andy Says:

Dan, why would a fresh pitcher be more likely to throw more or fewer pitches than the starter? Maybe the starter's fastball is a little slower late in the game, and with fewer swings and misses, balls get put into play more often. I don't think the number of pitchers used in the game has much affect.

20. dukeofflatbush Says:

I'm sure this has been talked about before, but I just realized CC Sabatia led the AL & NL in shutouts in 2008. I can't imagine any other circumstance where anyone could lead both leagues, in any major, non percentage category in the same year.

21. Jim in NC Says:

Good article. Thanks for all the data. What impressed me is how consistent the numbers are. Plate appearences per game have stayed pretty close to 38 for the past century. If the estimations in one of the other replies for pitch count since 1950 are reasonably close then pitch count looks pretty stable too. Even in your chart the significant break looks like 93 to 94, and stable since 1994.

22. MichaelPat Says:

Makes sense to me that pitch counts are on the rise... Not so sure it's about the use of pitchers, though.

25 years ago, walks and OBA were considered fringe stats. (And most figured a walk was something the pitcher did.) Today, OBA and walk rates are right at the centre when a hitter's skills are considered, even in the GM's office.

Listen to a ball game today, and you will often hear announcers extol the virtues of the selective hitter. None of them were doing that in the 1970s and 1980s. (Mike Hargrove, the Human Rain Delay, would not be noticed in today's game.) I almost hate to bring up Boston - New York, but it's always there in those games. These teams have succeeded with what is clearly a conscious strategy to make the opposition pitcher work. Success gets imitated.

The old ethic of the aggressive hitter... get up there and hack... is rare today.

23. Proof: Starting Pitchers Are Wimps These Days | Boston Red Sox Shop Says:

[...] pitchers simply do not last as long as they used to 20+ years [...]

24. Tellin_It_Like_It_Is Says:

No offense, but this article is about 10-15 years too late. Since 1994, the pitch count hasn't moved much. This article would have meant something in the mid to late 90s but comparing 2010 to 1988 and focusing on the drastic change, most all of which occurred by 1994, is pretty lame.

25. Andy Says:

Well...firstly, this is not an article. It's just a post showing some of the data available on the site. Secondly, the graph above does not agree with your conclusion.

26. Ken Says:

25 years ago, walks and OBA were considered fringe stats. (And most figured a walk was something the pitcher did.) Today, OBA and walk rates are right at the centre when a hitter's skills are considered, even in the GM's office.

The BB rate has had some effect, but what struck me is how much the K/9 has jumped. From '88 to '10, the K rate increased from 5.69 to 7.23. Meanwhile, the BB went up just a bit, from 2.91 to 3.31. Since power is much more emphasized now, and the sting of the K has lessened, players are much less willing to go up there are simply make contact. More Ks and scoring --> higher total pitch counts.

27. JeffW Says:

Pitchers may simply be wasting more pitches. I seem to see a lot of pitches so far out of the zone, as to make me wonder what's going on? Messing more with a hitter's patience (or lack of...).

One guy going the distance is going to try to conserve his pitches more. As we have advancd into the era of more and more relievers getting involved per game, this is less necessary.

This would also allow starters to make more blatant attempts at getting less-patient hitters to go fishing. Starters can throw more pitches early, because they know there is a long line of relievers ready to pick them up. It's assembly-line pitching, where it no longer matters who wins.

Relievers may also be throwing more pitches because they can work the zone more, given their short-stint status. There are no more Goose Gossage/Bruce Sutter/Rollie Fingers-type three-inning closers. Like the starters, they can work each hitter more.

Thus, the counts rise.

28. purple Says:

any idea how to look up diceks japanese numbers?

29. Raphy Says:

@28 Very basic numbers can be found here: http://japaneseballplayers.com/en/player.php?id=matsuzaka
and here: http://sports.espn.go.com/mlb/news/story?id=2669791

30. JeffW Says:

Raphy,

Got it. Sorry I called you Ralphy last time.

Honest mistake.

31. Raphy Says:

JeffW - no reason to apologize. You were one of many. I found it funny.

32. purple Says:

thanks #29 (ralphy)