Mariano Rivera's Nasty Statistics

When fiddling with some statistics of pitchers across baseball history, I accidentally noticed that Mariano Rivera ranks incredibly high on many critical stats. I had expected pitchers from the dead-ball era to dominate this list - and they did, with the exception of the Sandman.

This involved all pitchers who pitched at least 1000 innings since 1901, the beginning of the modern era of major league baseball. What I was looking for were indicators of pitchers who were nasty to hit against, be it due to stuff or deceptive motion or a trick pitch, and thus postulated that such pitchers might rank well in stats that measure the following:

• HRs allowed
• Hits allowed
• Runs allows
• Strikeouts
• Walks - which I included because if a pitcher is always walking men on base and getting into hitters' counts, his nastiness is limited by his inability to command.

Since this study involved walks and strikeouts, I excluded seasons earlier than 1901, because prior to 1901 the rules that defined what constituted strikeouts and walks were different.

I ranked the pitchers by these stats, and that is when I noticed Mariano kept appearing at the top. So for these stats, I averaged each pitcher's ranks in the categories, then ranked these averages. The following table shows the pitchers who ranked in the top 10:

Pitcher	Hits/9	ERA	HR/9	BB/9	K/9	Avg. of All Ranks	Overall Rank
Waddell, Rube	31	6	28	227	79	74.2	1
Rivera, Mariano	6	10	264	72	26	75.6	2
Walsh, Ed	12	1	6	40	405	92.8	3
Wood, Joe	13	4	5	313	203	107.6	4
Johnson, Walter	35	9	72	83	386	117.0	5
Bender, Chief	90	30	45	94	447	141.2	6
Mathewson, Christy	101	7	89	9	543	149.8	7
Young, Cy	83	8	46	1	647	157.0	8
Plank, Eddie	100	19	11	101	592	164.6	9
Joss, Addie	23	2	7	5	797	166.8	10

If it is not clear, the way to read the table is as follows:

• The 2nd through 5th columns list the pitchers' ranks in particular stats. For example, Mariano ranks 6th in H/9, 10th in ERA, etc.
• The 6th column  ranks the pitchers' average ranks. Mariano is 2nd.

Mariano is the only non-dead-ball era pitcher who makes the top 10, and he even outranks most of them. Later on this post will compare Mariano with only pitchers in the live ball era. But first, let's let's discuss the particular stats used and any considerations that provide context.





     Note: This exercises uses the stats in the latest version of the Lahman database. I posted a spreadsheet containing stats and rankings for all 1037 pitchers with at least 1000 IP to Google docs - sclick the following link, and download the spreadsheet if you want to sort/filter/edit it: https://docs.google.com/spreadsheet/ccc?key=0AkMu391-e5t0dGNLRmxmc2ZmNHNMY01YVkM0eHU2Zmc#gid=0

The first stat at which I looked was ERA, expecting the leaders to be a bunch of dead-ball era pitchers. This was mostly true, with one exception: Mariano Rivera. Mo's 2.21 ERA ranks him 10th, right behind Christy Mathewson, Cy Young, and Walter Johnson. To put this in context, Mariano is the only pitcher in the top 40 ERA leaders who did not make his debut in the dead-ball era; the next highest-ranking live-ball-era pitcher was Hoyt Wilhelm with a 2.52 ERA. (Mariano's ERA is even more off-the-chart because he pitched in the high-scoring 1990s and 2000s; I didn't expect anyone from this era to rank that high.)


Another stat examined is home runs allowed, expressed as HR/9. Mariano has allowed .47 HR/9, ranking a seemingly pedestrian 264th. However, this is deceiving in that Mariano has pitched in the highest HR environment of all time. This is worth a digression, so the following graph charts HR/9 annually. As you can see, Mariano's career has occurred in the only era that averaged over 1 HR per 9 innings:

To rank Mariano with his peers, I flooked at only pitchers who debuted after 1990 but before 2000. Mariano ranks first in this group, with no one really close to him; the next closest player is his old teammate Bob Wickman at .68, and he is then followed by a group of players such as Pedro Martinez, Derek Lowe, Tim Hudson, and Andy Pettitte. The difference between .48 and .68 may seem small. But what it means is that instead of giving up 65 HRs in his career, he would have given up 91; 26 more home runs in the high-leverage situations in which Mariano pitches would probably have caused an extra Yankee loss or two each season.

Another stat I looked at was WHIP. Mariano's .998 WHIP ranks second all-time.  I also measured Hits-per-9 innings (H/9) - Mariano's H/9 ranks 6th.

The final set of stats I examined related to strikeout dominance and control; specifically I looked at K/9, BB/9, and strikeout K/BB. Here is where Mariano ranks:

• K/9: 8.26; ranks 26th all-time. Strikeouts have increase throughout most of baseball history, so pitchers of recent vintage tend to rank highly.
• BB/9: 2.04, ranks 98th all-time.
• K/BB: 4.04, ranks 4th all-time.

When determining how to average all the various rankings, I excluded two of the stats just discussed:

• WHIP - H/9 and BB/9 cover hits and walks allowed as separate components; including WHIP would have double-counted those.
• K/BB:  K/9 and BB/9 cover strikeouts and walks as separate components, so K/BB would have double-counted those.

Now let's circle back to the listing promised earlier in the post, which shows how Mariano ranks among pitchers in the live ball era. The following table lists the top 20-ranking pitchers in the live ball era:

Pitcher	H/9	ERA	HR/9	BB/9	K/9	Avg of All Ranks	Overall Rank
Rivera, Mariano	6	10	264	72	26	75.6	2
Martinez, Pedro	11	122	575	207	3	183.6	14
Sutter, Bruce	42	93	449	330	64	195.6	18
Fingers, Rollie	67	116	428	288	113	202.4	19
Hoffman, Trevor	7	105	673	261	6	210.4	21
Seaver, Tom	34	102	516	295	114	212.2	22
Drysdale, Don	142	128	534	138	155	219.4	27
Clemens, Roger	47	177	447	458	18	229.4	31
Maddux, Greg	307	197	414	26	221	233.0	34
Wilhelm, Hoyt	8	41	389	566	167	234.2	37
Gibson, Bob	43	117	384	562	81	237.4	38
Chance, Dean	69	121	293	563	166	242.4	42
Marichal, Juan	145	109	668	32	259	242.6	43
Lincecum, Tim	27	135	368	690	4	244.8	48
Pena, Alejandro	165	167	418	416	86	250.4	54
Smoltz, John	109	271	552	298	33	252.6	57
Perry, Gaylord	227	166	455	165	254	253.4	58
Koufax, Sandy	2	79	618	586	7	258.4	61
Halladay, Roy	390	224	532	36	112	258.8	62
Santana, Johan	41	164	848	241	10	260.8	64

No one is even close to Mariano. Pedro Martinez's 2nd-place average ranking is 183.6; 108 behind Mariano's average of 75.6. This is a significant difference. You can see that by comparing the difference between Mariano and Martinez to the difference between Martinez and the 20th-place pitcher, Johan Santana, which is only 77.2. (From a statistical perspective, the standard deviation of the "Avg. of All Ranks" values is 163.5, meaning that Mariano is 2/3 of a standard deviation ahead of Pedro.)

While it is not earth-shattering news that Mariano Rivera is unique to the point of freakishness, I was startled to see Mariano ranking with pitchers from an era where the all-time HR leader had 138 HRs.

In closing, I want to make clear that this approach should not be taken for more than it is, which is to examine 'nastiness' using actual stats. Specifically, the results here do not pretend to define the greatest pitchers of all-time. This approach is not applicable to that question for a few reasons; for example, Walter Johnson was a much greater pitcher than Rube Waddell, in that the Big Train pitched twice as many innings as Waddell.

In addition, this approach does not account for era differences. Lefty Grove, for example, pitched in an era when strikeouts were relatively low, H/9 relatively high, ERA relatively high, and so does not get treated well here. This makes a nice segue to upcoming sequels; future posts will adjust these stats by era, and then we will see the implications and where Mariano ranks.