Good as always, Chris.

I was amused to see Claude Osteen's name come up. I mentioned Osteen's career several years ago as being a sort of a worst-case scenario for (a healthy) Buehrle's career (no, it's not on record unless my brother keeps all his e-mail). I'm pleased to see that's still holding true.

That is definitely an improved comp list. Still, you do need to do some adjustment for era. Friend, for instance, was striking out batters at about league average rates during his pre-age 28 years in the late 50s. I would suggest using IP greater than 1300, ERA+ between 115 and 131 and K rate less than 5.2 (in light of the current high K rate environment) and using 1920 as the baseline year. W rates are almost invariably low for the era because of the ERA/K requirements. Pitchers such as Ned Garver, Waite Hoyt, Wes Ferrell and Mel Harder appear on the list. It is strange that when these latter search criteria are used (with minimum innings of 1300 innings) Buehrle has in fact pitched the fewest innings of any of the qualifying pitchers, despite having thrown almost 1500.

Good point, but if you go back to 1920, even 5.2 K's/9 IP becomes problematic. In the 1924 NL, for instance, only two guys struck out more than 86 batters on the season.

I know Dr. Memory once adjusted all pitchers' K & W rates for era, but I don't know if I have that anymore.

I know Dr. Memory once adjusted all pitchers' K & W rates for era, but I don't know if I have that anymore.

You can actually find the ol' RSI blog at archive.org.

Chris, I still have that if you want it. Just let me know.

I think the list is still off. It has to recognize the difference in usage from pitch counts. And the treatment and medicine since then.

I mean, are Charlie Buffinton and Jim McCormick really good comps for Mark Buehrle?

Good point. All my projections were improperly lowering his production due to the risks of consumption, dropsy, catarrh, and ague.

I did NOT use all of baseball history for my "study" (study is in quotes because it was just a cursory inquiry really). In fact, I only looked at pitchers since around 1980. I don't know why Chris says that I did. I guess he only perused my comments on the blog. No big deal though.

I was really lost trying to follow Chris' explanation of how this pitcher or that was similar or not to Beuhrle and I think that he was cherry-picking some pretty irrelevant patterns and stats to make some kind of point which I think was that Buehrle is more similar to the haves than the have-nots.

I think the bottom line is that no matter what you put in as a pitcher's characteristics, you will find that ALL pitchers have a high risk of injury (we knew that already) regardless of their history and regardless of how they pitch (yes, some pitchers are a higher risk than others). In fact, I think you will find that the best you can hope for 4 years down the road is 150 IP or so. And yes, that is an average that is likely based on a combination of pitchers who do not get injured at all and continue to rack up lots of innings and those who get hurt and lose a year or two or even their whole careers.

MGL, I based it on DSG's article. He quoted you there and used it in his article. I checked b-ref's PI. There's only 37 pitchers in baseball history who had 200 IP every year from ages 22-27. I gave the list in the thread about DSG's piece last week.

I don't think I did any search for 200 IP 7 years in a row. I think I looked at 2 or 3 years in a row. Anyway, I used my retrosheet database which only goes back 25 years or so. In any case, as I said, it is no big deal because I really didn't do any "reaearch," only, as I said, a cursory inquiry.

Comps are nice but if you can't show that similarities have to do with what you are trying to ascertain, I don't see the point. In this case, we are trying to ascertain Buehrle's expected IP over the course of his new contract. Other than age and injury history (any good pitcher who is not injured will get around 200 IP, so it is not really the IP we are looking for), I don't really see what else you need to throw in the mix. And even if other things were important (in ascertaining future injury), determine them BEFOFE the fact, not AFTER the fact. Doing the latter is bad research! You don't come up with 134 IP in the 4th year, then say, well, that is comprised of pitchers who were healthy and pitchers who were not, and guess what, I think my pitcher in question (Beuhrle) is similar to the ones who were NOT injured. Come on! You want to impress me (or anyone else), come up with your comps first (and as I said, convince me with evidence or common sense that you are using pitchers who are similar AND that the similarities might be or are related to chance of injury or longevity in general, such as K rate) and THEN let's see what the next 4 years looks like. Doing it your way is cherry picking, data mining, or whatever you want to call it - not good research. Granted, your inquiry was not meant to be rigorous, but your initial thesis was that you want to use modern pitchers rather than pitchers going back to the early days of basbeall. You did and came up with basically the same results as everyone else - that 4 years down the road, 28 yo pitchers with a long history of health and lots of IP only throw 140 IP on the average. That is what I expected. As I said, no matter what pitchers you look at - how you slice it - pitchers are simply not going to throw many innings 4 years down the road no matter what, simply because of the significant chance of injury that all pitchers have (and the volatility of their true talent level, independent of physical health). Trying to compare Buehrle to the "healthy ones" in your sample after you came up with results you didn't like - well, I don't buy it, as I said, especially since you did that after the fact.

Doing the latter is bad research! You don't come up with 134 IP in the 4th year, then say, well, that is comprised of pitchers who were healthy and pitchers who were not, and guess what, I think my pitcher in question (Beuhrle) is similar to the ones who were NOT injured.

I didn't say injured. The problem was that a lot of those who cropped up on my list clearly were on their way out by age 28. Dick Ellsworth had been healthy. He also hadn't been very good since age 23. Bill Monbouquette had been healthy. He's also fallen down in his mid-20s. Ross Grimsely was healthy. He had one good year in the previous four. It's tough to pick up 1400+ innings without being healthy.

I noted that the pitchers with similar wear on their arms as Buehrle, with similar W/9 and K/9 rates as Buehrle, who were still performing at a fairly high level in their late 20s, as Buehrle is, generally did well through age 32 aside from almost all of them having one bum season shorting before hitting 30.

You constantly harp on me for excluding healthy ones. In plain point of fact I didn't do that at the end of the article. Guys who were still pretty damn good at Buehrle's age stayed pretty good.

I don't think I did any search for 200 IP 7 years in a row. I think I looked at 2 or 3 years in a row. Anyway, I used my retrosheet database which only goes back 25 years or so. In any case, as I said, it is no big deal because I really didn't do any "reaearch," only, as I said, a cursory inquiry.

In that case, our beef resides solely with Gassko:

Take, for example, the comments provided by Mitchel Lichtman, author of The Book: Playing the Percentages in Baseball, and a former consultant to the St. Louis Cardinals:

More than 25% [of pitchers comparable to Buehrle] are not pitching at all in 4 years… you expect that on average Buehrle will pitch only 100 IP or so in 4 years…including the times he does not pitch at all… even with good, historically durable pitchers, you still can expect to lose 10-15% in quality/quantity per year.


Lichtman’s analysis foresees a bleak future for Buehrle. A one-in-four chance of being out of the major leagues by the end of his contract, double-digit decreases in innings pitched every year…it’s no wonder that stat-heads would be condemning this contract. But I can’t help but wonder if we’re digging deep enough in our analysis. Buehrle has stayed remarkably healthy throughout his major league career; I wonder if that trait portends continued success.

Luckily, that’s a question that can easily be answered. Buehrle is 28, and he has thrown at least 200 innings in each of the past six seasons. There have been 37 pitchers in baseball history who likewise from the ages of 22 to 27 pitched at least 200 innings each year, and 32 of those threw at least 115.7 innings at 28, which is how many Buehrle has at this point in the season.

We can look at these 32 pitchers’ career paths over the next four seasons to figure a reasonable expectation for Buehrle. In percentage terms relative to this year, here is how we expect Buehrle’s numbers to progress from next year to the end of his contract:

Year &#xIP; R;A
2008 0.81 1.08
2009 0.73 1.01
2010 0.64 1.03
2011 0.50 1.09


In other words, we expect Buehrle’s innings to drop 19% next year, and his ERA to rise 8%. By the end of his contract, we expect Buehrle to pitch only half as many innings as he does this season, and with an ERA that is 9% higher to boot.

I and others, including Chris, inferred that he was using your data for his article. I wonder if he would like to comment, as he did in the other thread?

How could you infer I was using Mickey's data? The whole point of my article was that I was using a different data set (more years, more stringent requirements).

Seriously. There was a link to the book blog thread in the article, and at least one more that i know of (provided by me) in the comments on Dave's article. Before impugning MGL, did anyone bother to read what he actually wrote? The thread is very excellent and merits a read. Tango has done several more comp lists with varying results. Here is the link.

As I said, it is no big deal that Chris misquoted what I did. No big deal at all. It is not like he was doing a thorough examination of the research done in this area. Plus, as I also said, I only did a quick look at some historical numbers. I wish that no one had actually quoted me, perhaps just linked to The Book blog thread.

I was too harsh on Chris' analysis and conclusions. However, if in fact some of those pitchers were "already on their way out," I would have liked to have seen him put some more conditions on the comps BEFORE he looked at the performance of those comps, and not AFTER.

If it were me and I were doing some more rigorous research, I probably would look at all pitchers who were good (some ERA+ threshold), who pitched lots of innings for several years, who were about the same age, and had around the same K rate. In order to increase my sample size, presumably without impacting the comparisons too much, I would not limit the comps to those who had 7 or 8 years of pitching in the majors with the "good" and "lots of IP" requirements, and I would not have any BB rate requirement, since I doubt that that has anything to do with longevity (whereas K rate might) and I don't think it matters all that much if the pitchers are 27 or 29, rather than 28 (maybe it does).

In any case, I would be real comfortable putting an over/under on IP in the last year of Buehrle's contract at something like 140 or 150.

See where Bill Gullickson is? Well, he missed his age 29 and 30 seasons with injury. And he's still ahead of a handful.

No, Gullickson spent his age 29 and 30 seasons pitching for the Yomiuri Giants in Japan, pitching 203 innings in 1988 and 111 in 1989.

He returned to MLB in 1990, and pitched until 1994.